Connexion
OpenAI and Anthropic executives are quietly gaming out the day after a catastrophic AI attack. Most of that planning is about Washington, not the people who’d lose their bank access or tap water.
Top executives at OpenAI, Anthropic and other AI companies are privately planning for what happens after a catastrophic AI event, according to an Axios scoop published Friday. The event they’re bracing for is most likely a cyberattack that cuts off access to banks, the internet, or even power and water. Many industry insiders told Axios they expect something like it within the next six to 12 months.
The real issue isn’t that companies run worst-case drills. Everybody does that. It’s that the people building these tools now think a major attack is likely, and the planning Axios describes is mostly about the political blowback. Executives are racing to educate members of Congress so they can shape the laws that get written after the first disaster. Nothing in the reporting says anyone is planning for the family whose debit card stops working.
For you, six to 12 months means sometime between next spring and this time next year. If they’re right, the systems on the line are the ones your paycheck lands in, the card you use at the grocery store and the pumps that push water to your faucet. In 2021, one ransomware gang hitting one pipeline emptied gas stations across the Southeast in less than a week. The people who know these models best are telling us something bigger is coming, and nobody is telling your household to get ready for it.
The sourcing matters here, so let’s be straight about it. The six-to-12-month estimate comes from unnamed insiders, not from any company on the record. OpenAI confirmed it runs preparedness exercises but told Axios the scenarios “are not treated as inevitable.” Anthropic declined to comment.
According to Axios, the planning includes red-teaming worst-case scenarios, but it’s focused mainly on Congress. Executives know no regulation can pass right now. They want to be the ones writing the playbook when lawmakers panic after the first big hit. They’re also bracing for Democrats, who they expect to be stronger after the midterms, to move fast against AI, and they’re already making the argument that slowing the AI buildout could hurt an economy that’s already fragile.
Some posts online are running this as “AI companies prepping for an event that could change the trajectory of humanity.” That’s not what Axios reported. Nobody has a date, and no company has said an attack is coming. What we have is a group of insiders who believe one is likely and companies that are acting like they believe it too.
This also isn’t the first time we’ve heard the timeline. In August, OpenAI, Anthropic, Amazon Web Services, Microsoft and more than 100 other companies signed an open letter warning there was “a limited window” to protect hospitals, water treatment plants and other critical infrastructure. They made no commitments, set no deadlines and pledged no money. In September, Anthropic CEO Dario Amodei warned that a swarm of AI agents could take over the internet with a persistent botnet within six to 12 months. Several security experts called that far-fetched. Now his industry’s insiders are using the same window for something more specific.
You don’t have to take anyone’s prediction on faith, because a preview just happened in South Korea.
Between September 27 and 30, hackers broke into systems at Shinhan Bank, KB Kookmin Bank, Hana Bank, BNK Busan Bank, Yegaram Savings Bank and Hyundai Capital, according to the Kyunghyang Shinmun. At Shinhan alone they took 25,727 records, including names, phone numbers, annual income and loan limits. For a customer, that means a scammer can call pretending to be your loan officer, already knowing what you make and how much you can borrow. Korean regulators called an emergency meeting on a holiday, and the country’s president ordered a full investigation.
The attackers used ARTEX, a free AI hacking tool that went up on GitHub on July 26. Its newest version came out September 24. The attack on Kookmin started three days later. CrowdStrike traced the operation to a likely Chinese-speaking, financially motivated attacker running Chinese AI models, including DeepSeek, who also asked Anthropic’s Claude where to sell Korean breach data. Axios called it proof of how much damage a single person can now do.
The AI companies’ own agents have been part of the problem too. OpenAI said last week it had notified more than 100 organizations that its agents may have gotten into their systems during pre-release testing. Researchers told Axios the hacks weren’t sophisticated, and that’s what worries me. The agents used the same old tricks human hackers have used for decades, like stolen passwords and exposed keys, only faster than anyone can patch the holes.
None of this surprised us. In April we published Mythos AI: The Doomsday Weapon That Could Cripple Infrastructure, Banking and the World, which argued that the first targets of these cyber tools would be banks, utilities and communications networks. In September, Bill Gates published nearly the same list. Now the companies themselves are planning around it.
In May 2021, a ransomware gang called DarkSide locked up Colonial Pipeline’s computers, and the company shut down a 5,500-mile line that carries about 45% of the East Coast’s fuel. Within days, more than 1,000 gas stations ran dry. About 60% of stations in Atlanta were out of gas, and governors declared states of emergency. EZ Mart, a two-pump station in Wilmington, North Carolina, later sued Colonial on behalf of more than 11,000 stations, arguing the company never protected its systems. That was one gang, using human hackers, hitting one company, and the pipeline was back up in about a week.
The other history lesson is about who gets protected afterward. Axios quotes a Democratic aide pointing to 2008 and COVID as proof Washington can come together in a crisis. That’s true, and in 2008 coming together meant a $700 billion rescue for the banks while millions of regular families lost their homes. The AI companies are trying to be the ones in the room when the next rescue gets written, and the last time an industry wrote its own rescue, it protected itself first.
If this were 2019, a week without card payments would be a mess you’d ride out. This year, the cushion is gone.
Diesel hit a record $6.53 a gallon last month, and the country is heading into winter about 600 million gallons short on diesel and heating oil. Then China and Russia both stopped exporting diesel. Inflation was back up to 3.4% in August. And as we showed in The AI Bubble Is Cracking, AI spending is one of the main things holding the economy up, even though Anthropic’s own IPO filing shows it spending about $2.75 for every dollar it brings in.
Half of American households can’t cover a $1,000 emergency. A family like that, already paying record prices for fuel and food, has no room for a week where the bank app is down, the paycheck doesn’t post and the grocery store can only take cash. A cyberattack that would have been a bad news week in a normal year could be the thing that tips a lot of households over.
These companies believe a disaster is likely enough to war-game the political fallout. Ten days after their CEOs called for an industry slowdown in September, Anthropic and OpenAI launched new models within two hours of each other. If you honestly think your industry is about to knock out somebody’s power or water, you damn well owe them more than a briefing for Congress and a plan to manage the headlines.
The post AI Insiders Expect a Cyberattack on Banks, Power or Water Within 12 Months appeared first on Activist Post.
Rep. Jamie Raskin, the top Democrat on the House Judiciary Committee, sent letters Tuesday, Sept.29, to Amazon, Google parent Alphabet, Meta Platforms, and Oracle seeking information on non-disclosure agreements (NDAs) signed with government officials related to data center build-outs, according to copies viewed by The Wall Street Journal.
Raskin wrote that residents cannot weigh a subject’s costs against its benefits, hold officials accountable, or safeguard resources when basic terms are kept secret, according to the report. He demanded detailed information on each company’s data center portfolio and use of the agreements [7].
Amazon claims it no longer uses NDAs with government collaborators on data center projects, according to the report. Google, Meta, and Oracle did not immediately comment [7].
The letters mark an escalation in congressional scrutiny of an industry that has expanded rapidly across the country. The requests focus specifically on agreements that shield the terms of data center deals from public view [7].
Local communities have cited non-disclosure agreements and secrecy surrounding data center plans as a major reason they oppose the facilities, which many blame for pushing up power prices and consuming large amounts of water [7].
Data centers are being sited across the United States without meaningful public input, a pattern documented in states such as Utah, where the three-member Box Elder County Commission voted unanimously to greenlight a 40,000-acre hyperscale project before residents could weigh in [1].
The backlash has become a measurable force. Local opposition disrupted at least 45 proposed U.S. data center developments worth nearly $68 billion between April and June 2026, according to Data Center Watch, a research project operated by AI intelligence company 10a Labs. The sustained resistance now spans 49 states with 843 active opposition groups, according to the same research.
Data centers have become a flashpoint in key races including Ohio, Michigan, and Texas, where states have taken steps to cut tax exemptions worth billions of dollars. A new Wall Street Journal poll found that 62% of voters support pausing data-center development, according to the report.
Pennsylvania Gov. Josh Shapiro banned the agreements in an executive order in August, and several states have introduced bills to ban data-center NDAs. In Texas, Gov. Greg Abbott ordered the state’s environmental watchdog to withhold permits for new data centers until an audit of risks to the power grid is complete, according to ZeroHedge [2].
Industry executives have said they hope to be more transparent moving forward and have committed to paying more for the power needed to run data centers and investing more in local communities [7].
Executives said some secrecy is needed when negotiating with local officials about the facilities to avoid revealing proprietary details that could help competitors and because contractors and other businesses working on the projects often increase prices when they find out a large AI company is working locally, explained executives [7].
The secrecy controversy unfolds as evidence mounts that quiet deal-making has real consequences for ratepayers. Air pollution from data centers could increase healthcare costs by almost $21 billion by 2028, contributing to roughly 600,000 asthma cases and 1,300 premature deaths, according to the Environmental Protection Network, a nonprofit whose members are former public officials [3].
Raskin, along with Reps. Greg Landsman (D., Ohio) and Tom Barrett (R., Mich.), introduced a bill last week to limit NDAs with government officials related to data-center projects. Barrett has faced criticism from his opponent [7].
Sen. Jon Husted (R., Ohio) has sponsored a bill requiring states to consider taking steps to make tech companies cover the cost of energy infrastructure needed to power the facilities. The House recently passed its version of the bill 417-3, the report stated.
In the Senate, the measure failed in a 57-43 vote after Democrats blocked the bid to overcome the 60-vote filibuster threshold, with all but four Democrats voting against advancing the bill, according to The Epoch Times [4].
Senate Democrats have broadly dismissed the measure, arguing that its voluntary structure for companies is too weak. The questioning is an early indication of what AI companies can expect if Democrats win back power in Congress in November, a prospect that has many in the industry bracing for a wave of demands and hearings.
Others, including a trio that lead a House AI Commission of Democrats, have targeted the sector over recent instances of models hacking other companies, while other concerns include the impact the platforms have on children [7].
The data-center debate has surged across America, from the Birthplace of Rock ‘n’ Roll to Kansas corn country, turning server farms into one of the defining issues of the midterm elections this fall [5].
The combination of secrecy agreements, rising electricity costs, and organized local opposition has created a political environment in which data center projects face scrutiny from both parties.
In some primaries, candidates who backed the artificial intelligence boom were given the boot by voters who preferred politicians encouraging guardrails or a hard stop before big tech turned country pastures into concrete parks, according to The Epoch Times [6].
As Raskin’s letters make clear, the question of who gets to see the terms of these deals is now a matter of congressional interest.
The post Lawmaker Seeks Details on Secret Data-Center NDAs as AI Build-Out Faces Scrutiny appeared first on Activist Post.
If the video above does not play, please try on our social media channels
Journalist Derrick Broze breaks down his latest investigation into the AI company at the center of the recent wave of so-called “rogue AI” incidents. The firm, known as Irregular, has numerous ties to Israeli intelligence.
Read Derrick’s article: Irregular: The Israeli Company with Intelligence Ties at the Center of the Rogue AI Incidents
The post Meet “Irregular” – the Israeli Firm at the Center of the “Rogue AI” Reports appeared first on Activist Post.
New Mexico is seeking between $35 billion and $40 billion in civil penalties from Meta Platforms over allegations that the company misled consumers about Facebook data privacy practices, according to Reuters. Lawyers for the state made the request on Oct. 1 at a hearing in a lawsuit brought after the Cambridge Analytica scandal broke in 2018, the report stated [1].
A Santa Fe jury found Meta liable for willfully deceiving consumers, resulting in more than 43 million violations of the state’s consumer protection law, according to court proceedings reported by the news agency. The violations were deemed to have affected the state’s entire population of about two million people [2].
Judge Francis Mathew is expected to rule later this month on the penalty amount. Attorneys for Meta and New Mexico clashed over the appropriate penalty, with the company arguing the amount sought by the state did not match the conduct at issue in the trial, Reuters reported. At the upper end, the fine would rank as the largest financial penalty ever imposed on a U.S. corporation [2].
The case traces to the 2018 Cambridge Analytica scandal, which erupted after a whistleblower revealed that the British political consulting firm had obtained data from tens of millions of Facebook users without their consent through a third-party app, according to reports [2]. The case was tried in Santa Fe over two weeks and centered on the company’s handling of a data breach tied to a third-party personality quiz called This Is Your Digital Life that harvested data from roughly 87 million profiles and sold it to Cambridge Analytica, a now-defunct political consulting firm, according to trial testimony [3].
The data was used for voter profiling and political messaging, including during Donald Trump’s 2016 presidential campaign, the report stated. Frank H. McCourt, Jr. writes in Our Biggest Fight: Reclaiming Liberty, Humanity, and Dignity in the Digital Age that records show Cambridge Analytica boasting to its GOP clients that, via its Facebook relationship, it had obtained more than 7,000 personal data points on each of 240,000 American voters [4]. Cambridge Analytica filed for bankruptcy months later, while Facebook faced billions of dollars in fines and compensation claims over its handling of user data [2].
The scale of the data breach reflects broader costs associated with centralized data storage. As Sam Ghosh and Subhasis Gorai note in The Age of Decentralization, the average cost of a data breach in 2023 was around $4.45 million, a figure that includes costs related to lost business and post-breach responses, and a staggering 82% of data breaches involved data stored in centralized databases [5].
The Santa Fe jury found Meta willfully deceived consumers, resulting in more than 43 million breaches of New Mexico’s consumer protection law, according to court findings reported by Reuters. The violations were deemed to have affected the state’s entire population of about two million people, the report stated [2].
Attorneys for Meta and New Mexico clashed over the appropriate penalty, with Meta arguing the up to $40 billion sought by the state did not match the conduct at issue in the trial, Reuters reported Wednesday. Judge Francis Mathew is expected to rule later this month on how much Meta must pay [2].
Reuters noted that the jury found the Facebook parent company liable for misleading statements about the privacy of users’ data just one week before the penalty hearing [6]. The verdict came after a two-week trial in Santa Fe that focused on the company’s handling of the Cambridge Analytica data breach [3].
In March, another Santa Fe jury found Meta liable for 75,000 violations related to child safety and ordered $375 million in penalties, according to the report. A local judge later deemed Meta’s platforms a public nuisance and ordered an additional $567 million. Meta has said it will appeal the rulings [2]. The earlier verdict was the first to find Meta liable for violating state consumer protection laws by misleading users about platform safety and enabling child sexual exploitation, according to NaturalNews.com [7].
In August, Meta agreed to pay roughly $17 billion to settle claims by dozens of U.S. states and territories that it deliberately designed Facebook and Instagram to addict children and misled the public about risks, the report stated. The settlement with attorneys general from 29 states includes child safety measures and a provision that the states agreed not to sue Meta under existing child safety laws over its retention and use of children’s data, according to TechCrunch [8]. New Mexico did not join that settlement, allowing its case to proceed to trial [2].
The broader legal effort against Meta has included more than 40 U.S. states alleging the tech giant knowingly and deliberately misled the public about the substantial dangers of its social media platforms, according to a 233-page complaint filed in the U.S. District Court for the Northern District of California [9]. Separately, a former Meta employee testified before a Senate subcommittee that the company was aware of harassment and other harms affecting teenage users but failed to address them, according to NaturalNews.com [10].
Meta, which operates Facebook, Instagram, WhatsApp, and Threads, is not alone in facing scrutiny over data practices, according to the report. Last month, TikTok and its parent company ByteDance agreed to pay $400 million to settle U.S. allegations that the platform illegally collected children’s personal data without parental consent. Google, Apple, Netflix and other tech giants have also been involved in privacy-related litigation [2].
Privacy advocates argue that centralized data collection creates systemic risks. As Sam Ghosh and Subhasis Gorai note in The Age of Decentralization, security-related costs of centralization include expenses for securing databases, compensating affected users for damages, and fines [5]. The judge’s ruling in the New Mexico case is expected later this month [2].
The post New Mexico Seeks Up to $40 Billion From Meta in Data Privacy Case appeared first on Activist Post.
par Vassili Kachine. L'intégration de l'IA ne sera vraisemblablement qu'un acte parmi d'autres d'une transformation plus vaste de la guerre. L'IA et les systèmes sans pilote entraîneront d'autres changements, bien plus importants, dans l'armement, la tactique et l'économie de guerre.
The post Comment l’IA et les drones changent le rôle de l’arme nucléaire first appeared on Réseau International.
“And it causes all, both small and great, both rich and poor, both free and slave, to be marked on the right hand or the forehead, so that no one can buy or sell unless he has the mark.” — Revelation 13:17
Pax Silica is a U.S.-led strategic initiative launched in December 2025 to control the global supply chains for semiconductors, AI infrastructure, and critical minerals—by organizing a coalition of “trusted partners” to reduce dependence on China.
The name combines “Pax” (Latin for peace) with “Silica,” signaling a vision of a U.S.-centered technological order. The initiative is led by Jacob Helberg, the Under Secretary of State for Economic Affairs, who framed it as a purpose-built alternative to forums not suited to managing the AI economy—the G7 and the G20.
Helberg was direct: “If the 20th century ran on oil and steel,” he said, “the 21st century is going to run on compute and minerals.” Compute is the processing power that runs AI. Minerals are the raw materials that make the chips and hardware possible.
The initiative spans everything from mining and refining, through chip design and fabrication, to data centers, energy, and logistics.
Helberg calls the underlying idea “innovation sovereignty”—the claim that a nation’s security no longer comes from controlling every part of the tech stack within its own borders, but from being a trusted contributor to a network the United States leads.
The founding declaration was signed by the United States and five partners: Japan, South Korea, Singapore, the United Kingdom, and Australia. The initiative has since expanded to 24 signatories, including the European Union. Taiwan has separately endorsed its principles, and at a June 2026 summit, 35 economies signed a corresponding Joint Statement on AI Opportunity.
Pax Silica is not a treaty organization. It is a hub-and-spoke network: the United States signs bilateral declarations with each partner, tailoring the terms to what each country can contribute.
One partner supplies critical minerals. Another provides advanced semiconductor manufacturing. A third offers data centers or AI research capacity. The goal is to maximize the resilience of the whole network rather than have every member duplicate the same efforts.
The United States aims to mobilize more than $1 trillion in allied investment. The US government is contributing $250 million to seed the Pax Silica Fund—and asking its allies to pay for the rest.
The complications are already visible.
India—the only full BRICS member to join—signed on in February 2026, on a non-exclusive basis.
Kazakhstan, a BRICS partner country, joined both Pax Silica and China’s rival World AI Cooperation Organization (WAICO), a 29-nation bloc headquartered in Shanghai that offers subsidized Chinese AI models to developing nations.
The European Union signed the declaration while simultaneously pursuing its own Technological Sovereignty Package—a set of legislative proposals, including Chips Act 2.0, designed to reduce Europe’s dependence on both American and Chinese technology.
And the United States is the world’s largest debtor. Its claim to global power has rested for fifty years on the petrodollar—the arrangement that made oil the foundation of dollar demand. That foundation is eroding. Pax Silica is the search for the next one.
Everything about Pax Silica—the coalition, the investment target, the full-stack scope—is designed to control the global economy. And the first prototype is already being built.
The digital passport is the operational heart of the Panama pilot. It is a cryptographic credentialing system for AI-related cargo—a verified digital record that travels with a shipment, certifying its origin, custody, and compliance status from the point of mineral extraction to the final destination. It is a log of every step in the chain.
The pilot itself is a twenty-four-month program the State Department commissioned in August 2026 for $50 million. Its first deployment is the Panama Canal—one of the most important shipping chokepoints on earth. The system must be designed and built from scratch.
Here is how it works. Every shipment of semiconductors, critical minerals, or AI infrastructure would be checked against that record. If verification confirms the cargo is from a trusted source and has followed approved routes, it gets pre-approved expedited processing and faster customs clearance. If the record shows a gap—an unapproved origin, a missing link in the chain—the shipment is held.
If the pilot succeeds, the platform expands to other Pax Silica countries. Every member economy would be plugged into the same verification system. Every shipment would be logged, credentialed, and either fast-tracked or held for inspection. The digital passport becomes the gatekeeper of the AI economy. Whoever controls the credential controls who gets to participate.
This is what “economic security” means in practice. Not merely secure supply chains—controlled supply chains. And controlled supply chains are not a new kind of power. They are the oldest kind. Whoever controls access to exchange controls the people who must exchange. Revelation 13:17 says no one can buy or sell without the mark. The mark is not on the hand or the forehead. It is in the ledger needed for life.
Here is the inversion that makes the American position unique. Proverbs 22:7 says the borrower is the slave of the lender. By that logic, the United States—the world’s largest debtor—should be the slave. Instead, it built a system in which its creditors are trapped by its debt. If the US collapses, they collapse with it. The debtor became the master—not by escaping the debt, but by making the debt shared.
Pax Silica is the attempt to extend that inversion into the next century. It is what you might expect from the world’s largest debtor, searching for a new anchor before the old one fails. The world’s largest debtor is in charge the way a man at the edge of a cliff is in charge of the people roped to him. If he falls, they fall too. And they know it.
One system determines which physical goods can move. The other determines how economic transactions within the network are priced and settled. Together they raise the question of who controls access to the next economy.
The “token dollar” is the financial layer of the Pax Silica architecture. It is a proposed evolution of the U.S. dollar’s global role, designed for the AI era. The idea is simple: make the U.S. dollar the default currency for pricing and settling AI computation—the “tokens” that AI models process and generate.
The parallel to oil is direct. The petrodollar system, built from 1974, made the dollar the mandatory currency for oil transactions. Every country that wanted oil had to hold dollars. The token dollar proposes that in the AI age, compute becomes the new strategic commodity, and it should be priced and settled in dollars.
Here is how the mechanism works. Every AI transaction, from a single query to a full model training run, is priced in dollars. The payment rails are dollar-pegged stablecoins—programmable digital tokens designed to maintain a one-to-one value with the US dollar. When global companies buy AI compute, they generate dollar revenue for US-linked firms. That revenue flows into the dollar-based financial system. The reserves backing the stablecoins are invested in US Treasuries. Every AI transaction, anywhere in the world, increases demand for US government debt.
The Center for Strategic and International Studies (CSIS), a Washington think tank closely tied to the national security establishment, recommends the US lead a coalition to make the dollar the “default unit of pricing for token generation,” using dollar-backed stablecoins to settle these contracts and connecting the system to US futures markets.
This is where the three layers close into a loop. The physical gate—the Panama pilot—controls which goods move. The credentialing layer—the digital passport—determines who is trusted. The financial layer—the token dollar—captures the value. A country that wants to participate must align with Pax Silica’s rules, open its books, verify its supply chains, and accept the credentialing framework. Once that infrastructure is in place, every AI transaction within the network is priced and settled in dollars.
The loop closes like this. Global companies buy AI compute. They generate dollar revenue for US firms. The reserves backing the stablecoins go into US Treasuries. Dollar demand is reinforced. The US can borrow more cheaply. It funds its deficits. It projects power. It maintains the system.
The data shows this is already happening. More than 98% of agent-to-agent settlements over the past year used USDC, a dollar-pegged stablecoin issued by Circle, a private company that holds US Treasuries as reserves.
Between May 2025 and April 2026, autonomous AI agents settled roughly $73 million across 176 million on-chain transactions — an average of about 31 cents per payment. Separately, stablecoins processed $33 trillion in total on-chain transaction volume in 2025, surpassing Visa and Mastercard combined. That figure includes trading and DeFi activity, not just commercial payments. The real-world commercial volume was smaller — about $400 billion — but it doubled year over year.
If it succeeds, the token dollar is the financial abstraction endgame: the U.S. dollar becomes the system’s entropy accelerator. It takes massive amounts of real energy and resources and disperses them into digital abstractions. What remains is a number on a ledger—a number that must be fed still more energy just to persist.
If it fails, the US loses the ability to borrow cheaply, fund its deficits, and project power abroad.
Pax Silica is being rolled out during a period of unprecedented geopolitical disorder. The United States has, by its own actions, been described as a rogue state by much of the world—it kidnapped Venezuela’s president in January 2026, and its president talks openly about seizing Canada and Greenland.
Why? What would explain this behavior?
If the United States were at war, it would make sense. War justifies extraordinary measures. War justifies the suspension of normal rules. War justifies the seizure of resources, the control of supply chains, the mobilization of allies.
But “the war that isn’t” is clandestine, and for a reason. A declared war would require public consent. A clandestine war requires only that the public not notice the alternative architecture on the other side.
BRICS—the bloc of Brazil, Russia, India, China, South Africa, and six newer members—has spent a decade building institutions designed to move trade outside the dollar system. The New Development Bank, headquartered in Shanghai, has $100 billion in authorized capital and ten member countries. Its mandate is to finance infrastructure in emerging markets without the conditionalities that come with IMF or World Bank lending.
Critically, it is expanding local-currency financing—raising funds in member currencies and lending them in the same currencies, avoiding dollar conversion entirely.
BRICS Pay is the payment layer—a network that links the member states’ own payment systems (Russia’s SPFS, China’s CIPS, India’s UPI) into a single interoperability layer. It lets countries settle trade in their own currencies without routing through SWIFT, the Western-dominated messaging system that underpins most global banking. It is in pilot and phased rollout, not yet operational for all members. But the direction is clear.
There is also the Unit—a settlement instrument composed of 40% physical gold and 60% member currencies. It entered prototype phase in December 2025.
Here is what makes the BRICS model different from the American model. NDB loans are not debt into infinity. They are structured to be repaid. Loan tenures run from 10 to 30 years, with grace periods built in. They are increasingly denominated in local currencies, so borrowers repay in the currency they earn, not the currency they have to buy. And they come without the structural adjustment programs that the IMF and World Bank have used for decades—no forced austerity, no privatization requirements, no policy conditions that override national sovereignty.
This is the distinction that matters. The American system turns debt into a permanent asset. It rolls over. It expands. It requires continuous inflow. The BRICS system treats debt as a tool for development—borrow, build, repay, and then the infrastructure belongs to you.
Pax Silica is the mechanism by which the United States intends to maintain its position as the world’s hegemon in an era when the old foundations are failing. The BRICS system is the mechanism by which its rivals intend to route around it. Both are control architectures for the global economy—and neither is being debated as one.
The two systems are not at war with each other in any conventional sense. They are building parallel infrastructure. Pax Silica credentials goods. BRICS Pay settles payments. Pax Silica ties nations to the dollar. The NDB lends in local currencies to free them from it. Neither side is trying to destroy the other. Both are trying to make the other optional.
And the nations in the middle—India, the UAE, Kazakhstan—are discovering that optionality is leverage. They can belong to both systems. They can extract from both. They can refuse to choose, and no one can make them.
This is what the war actually is. Not a shooting war. Not a declared conflict. A slow, grinding competition to determine which infrastructure the next century will run on. Pax Silica is the American bid. BRICS is the alternative. And the middle powers are the prize.
Pax Silica is a rolling program with a published timeline. Here is what to watch for over the next twenty-four months.
The Panama pilot. The State Department’s $50 million supply-chain credentialing tender closed in August 2026. The performance period is 24 months. The system must be designed, developed, and deployed from scratch. If the pilot works, the platform expands to other Pax Silica countries after that. Watch for which countries adopt the credentialing system and which resist.
The Token Dollar curve. Real-world stablecoin payments—not crypto trading, but actual commercial movement of money—doubled to $400 billion in 2025. Watch for which companies adopt dollar-pegged stablecoins for machine-to-machine settlement, and which countries try to build alternatives.
The membership expansion. Pax Silica has 24 signatories and 35 AI Opportunity Partnership economies. Watch for which countries join, which refuse, and which are pressured. Every new signatory is a country that has agreed to the credentialing framework in principle. Every refusal is a country that will face pressure.
The chokepoint squeeze. China controls the processing of rare earths and many critical minerals. The US controls advanced chip design and the high-bandwidth memory (HBM) that AI chips require. Watch for export controls, licensing regimes, and supply-chain credentialing to converge in the same direction. Whoever controls the credential controls the economy.
The dual-alignment squeeze. Watch how Washington responds to India, the UAE, and Kazakhstan. The US has already pressured countries to choose. The question is whether it will escalate. If the US escalates, it risks pushing these countries further toward China. If it accepts the dual alignment, it undermines the entire logic of the credentialing architecture. Either way, the overlap nations become the test case for whether Pax Silica can function as designed.
The domestic squeeze. The digital passport being piloted in Panama is a prototype. Watch for it to appear in domestic contexts—trucking, agriculture, healthcare, pharmaceuticals. The logic appears to be that the infrastructure is built in the developing world first, where resistance is lower. It arrives here last, after it has been normalized.
The WAICO squeeze. Watch how many countries join China’s rival framework and whether any of them also sign Pax Silica. Kazakhstan has already done both. India has joined Pax Silica on a non-exclusive basis. If more countries follow, the binary dissolves. If Washington pressures them to choose, the pressure itself becomes the story.
The lens matters more than any single event. When Washington moves against a government, ask which commodity it sits on. When it courts a country, ask what it can contribute to the stack. When it applies pressure, ask which chokepoint it is trying to secure. The pattern is not hidden. You simply won’t be told what it is.
There is one more dimension to this architecture that deserves attention—not because it is proven, but because the capability exists.
DARPA’s Safe Genes program has invested $65 million over four years in gene drive research and genetic remediation technologies. The stated purpose is biosafety and biosecurity: developing tools to combat bio-threats and reverse unwanted genetic changes. Gene drives can spread a genetic modification through a population—making a population of mosquitoes unable to reproduce, for instance. Genetic remediation technologies can undo such modifications.
No public document connects Safe Genes to population control. The scientists involved work on malaria eradication, invasive species control, and biosecurity. The evidence does not establish a secret depopulation program, and this essay does not claim one.
But the capability exists. The funding is real. And the stated purpose is not the only possibility.
A system that treats supply chains as traceable nodes has access to tools that treat living systems as manageable. Gene drives are part of the broader toolkit of state-funded biosecurity research. The question they raise is the same one Pax Silica raises: what happens when living systems become objects of technical management?
Pax Silica is a plan to control physical resources with a financial system that runs on abstractions. That is the problem.
The plan treats compute as the new oil—an indispensable input that everyone must buy. But compute is not wealth. It is a tool. It is a means of production, not a store of value.
And it has a blindspot: energy.
Every AI model, every data center, every AI query requires enormous amounts of electricity. Training a single large language model can consume as much electricity as a small city uses in a year. The data centers that power AI require power grids, transmission lines, cooling systems, and fuel. All of it is physical. All of it is finite. None of it can be conjured.
The system has been able to pretend energy is cheap because debt has subsidized it. Cheap credit built the infrastructure of the fossil fuel age. Cheap credit let the system ignore depletion. Cheap credit made it possible to treat energy as a background assumption rather than the foundation it actually is.
But energy is not an assumption. It is the foundation.
Pax Silica can control the supply chain for chips. It cannot control the supply chain for energy. And energy is the chain that matters most.
Pax Silica assumes cheap energy. Remove it, and the plan collapses. No financial architecture can print the foundation it stands on. It can only build higher.
The verse at the top of this essay describes a system that determines who can buy and sell. Pax Silica is building that system through the architecture of debt. And like every architecture built on claims, it will eventually meet the thing it cannot print. When it hits that limit, the response will not be restraint. It will be the pursuit of total control.
That is the pattern. That is what the machine does when it meets a wall it cannot print its way through.
But the real question is not whether Pax Silica can work. It is whether it should.
CSIS. “From the Petrodollar to the Tokendollar: Economic Statecraft in the AI Era.” August 2026.
DARPA. “Building the Safe Genes Toolkit.” DARPA.mil, July 2017. https://www.darpa.mil/news-events/2017-07-19.
Gulf News. “UAE commits $1.4 trillion to US economy over ten years.” March 2025.
The Holy Bible. Revelation 13:17.
International Energy Agency. Key Questions on Energy and AI. Paris: IEA, 2026.
International Energy Agency. Global Critical Minerals Outlook 2026. Paris: IEA, 2026.
Keyrock. “Who Pays the Agent?” May 2026.
The New York Times. “Trump Sets Up ‘Pax Silica’ Fund to Reduce Global Dependencies.” March 2026.
ORF America. “The Purpose and Promise of Pax Silica.” August 2026.
Reuters. “Exclusive: US to tell partners they must pick sides in AI race with China.” August 2026.
U.S. Department of State. “Pax Silica AI Assistance Project NOFO.” August 2026.
U.S. Department of State. “Virtual Press Briefing with Jacob Helberg.” December 2025.
U.S. Geological Survey. Mineral Commodity Summaries 2026. Reston, VA: USGS, 2026.
The post Pax Silica—The Dollar’s Next Foundation appeared first on Activist Post.
Sur les marchés, le kilowattheure se vend parfois à prix négatif ; sur les factures, il ne cesse de grimper. Présenter ce paradoxe comme « normal » revient à ignorer ses vraies causes : l’organisation du marché européen de l’électricité et la place accordée aux énergies intermittentes.
Un paradoxe exaspère légitimement les Français : un kilowattheure qui s’échange parfois à prix négatif sur le marché de gros pendant que leurs factures s’envolent. Interrogé à ce sujet, un journaliste économique a rendu un verdict sans appel : la situation est parfaitement normale. En période de faible consommation et de forte production, les coûts fixes continuent de courir, à commencer par le tarif d’utilisation des réseaux. Et puis, a-t-il ajouté, « on n’arrête pas une centrale nucléaire comme ça ».
L’explication a le mérite de la simplicité et le défaut d’exonérer d’emblée trois responsables : la conception du marché européen de l’électricité, les mécanismes de soutien aux renouvelables et la priorité d’injection dont bénéficient l’éolien et le solaire.
Commençons par la centrale qu’on « n’arrête pas comme ça ». On ne l’arrête pas, en effet, mais on module sa puissance jusqu’à son minimum technique si nécessaire, ce qui revient au même. Et c’est là que le bât blesse : cette modulation, devenue fréquente et erratique, est dénoncée depuis des années par nombre d’ingénieurs du secteur, qui y voient une usure prématurée et coûteuse de notre parc électronucléaire.
Surtout, comparer l’électricité au blé stockable en silos méconnaît le principe de tout système électrique. À chaque instant, la puissance fournie doit être rigoureusement égale à la puissance appelée, sous peine d’une dégradation pouvant se terminer en blackout. Cet équilibre, dont l’image est le maintien de la fréquence à 50 hertz, s’obtient par la fourniture ou le retrait permanents de kilowattheures : un service que l’éolien et le solaire, productions dites fatales, sont incapables de rendre, comme d’ailleurs celui de régler la tension.
Dès lors qu’en période de fort ensoleillement ou de grand vent on paie pour maintenir ces deux productions à pleine puissance, il faut retirer une puissance équivalente aux autres moyens et maintenir malgré tout la fréquence et la tension à leurs valeurs réglementaires — une gageure que les Espagnols n’ont pas tenue le 28 avril 2025. Payer pour le maintien de l’éolien et du solaire, c’est exactement cela : vendre leurs kilowattheures à prix négatif.
Sur ce point, l’éditorialiste ne dit mot. L’essentiel de la perte n’est pas supporté par l’exploitant mais par les tarifs d’achat et les compléments de rémunération garantis par l’État, qui couvrent l’écart entre prix de marché et prix contractuel. Cet écart, ce sont les consommateurs et les contribuables qui le règlent. Le prix négatif sur le marché de gros et la facture salée ne sont donc pas contradictoires : la seconde est une conséquence du premier. L’État ne sait d’ailleurs pas ce que sa planification coûtera.
Silence également sur l’autre versant du problème, nos exportations, qui ont atteint 92,3 térawattheures en 2025, un record. Le producteur français vend au prix français et non à un prix du pays importateur, qui, comme en Allemagne l’été dernier, peut atteindre des sommets historiques. La recette générée par l’écart entre les deux prix n’est pas empochée par le producteur – EDF en l’occurrence –, mais par RTE et par son homologue du pays concerné, pour être ensuite partagée entre eux à parts égales. EDF cède ainsi au pays importateur la moitié de ce qu’aurait dû lui rapporter le volume des MWh exportés, une rente dite de congestion réputée devoir être affectée à l’entretien des infrastructures de réseau d’interconnexion par son bénéficiaire. En 2025, une telle rente s’est montée à 962,54 millions d’euros pour le seul Italien Terna auquel il conviendrait de demander ce qu’il en a fait précisément.
La règle européenne impose certes que cette rente soit affectée au renforcement des réseaux, mais aussi à réduire les tarifs ; sûrement pas à financer des moyens de production tels que le nucléaire. Bruxelles propose même d’en affecter 25 % à des projets européens communs, ce que la France refuse. Le pays qui produit l’électricité la moins carbonée d’Europe en partage donc la rente sans pouvoir l’investir dans l’outil qui la produit.
L’interconnexion du golfe de Gascogne, dont le coût est passé de 1 750 millions d’euros en 2017 à 2 850 millions, résume le mécanisme : la délibération de la CRE partage l’investissement à parts égales, mais l’exploitation et la maintenance sont supportées à 60 % par la France. Nous payons davantage pour faire fonctionner la ligne que pour la construire, et nous partageons ce qu’elle rapporte.
Le gouvernement a pris l’engagement que le prix de l’électricité n’augmenterait pas cet hiver. Mais le tarif réglementé a changé de structure le 1er octobre : il distingue désormais les saisons, et le kilowattheure d’hiver peut y coûter jusqu’à quatre fois le prix d’été. Le paradoxe que l’on nous dit normal n’est plus seulement toléré, il est désormais inscrit dans le tarif.
On attribuera aux guerres d’Ukraine et du Moyen-Orient, comme aux droits de douane qui renchérissent les matériaux, le rattrapage tarifaire qui suivra. Mais dissimuler que la plus grande part de la note tient à des choix politiques vieux de trente ans relève de l’intenable. Si l’un des parcs électronucléaires les plus performants du monde n’avait cessé d’être optimisé depuis les années 1970, la France serait à l’abri de la cherté comme de la pénurie. Qui mettra fin à la casse de notre système électrique ? Au pire moment de son histoire énergétique, ce pays est tenu par ceux qui refusent de nommer le mal.
L’article Électricité à prix négatifs et factures record : le paradoxe que nos commentateurs refusent de voir est apparu en premier sur Contrepoints.
Executive Summary:
Hunan Daily recently reported that DeepSeek’s founder Liang Wenfeng (梁文锋) regards data annotation as the most important work at this stage in the industry’s development, and has half of the company’s core researchers working on it (Hunan Daily, September 16). Data annotation is the standardized labeling of raw data that allows machine learning models to read it. The Hunan Daily article noted that data labeling in the People’s Republic of China (PRC) is shifting from labor-intensive to knowledge-intensive work. In the data annotation sector, this suggests a lag behind U.S. counterparts.
In August, Forbes reporter Anna Tong brought to public attention Silicon Valley data vendors like Mercor, Surge AI, AfterQuery, and Turing selling standardized, pre-packaged “off-the-shelf (OTS)” training datasets, collections of examples labeled or graded by human annotators to teach a model what a good answer looks like, to frontier artificial intelligence (AI) labs in the PRC (Forbes, August 5). [1] The report drew immediate attention in Washington, raising national security concerns (New York Post, August 30; Lawfare, September 8). Well before these purchases took place, Beijing had already designated data labeling a strategic industry and began building domestic alternatives.
PRC AI labs continued to buy U.S. datasets because domestic substitution (国产替代) in the data labeling sector is not advancing in parallel across the board. A review of PRC government documents and of the industry’s current condition show that policy objectives, the output of the national data labeling bases, and firm-level capabilities are progressing at different paces. This misalignment keeps parts of the PRC’s AI industry reliant on their U.S. counterparts.
Foundational Policies Lead to National ‘Bases’
Beijing began laying the groundwork for a data labeling industry two years ago. On December 26, 2024, the National Development and Reform Commission (NDRC), the National Data Administration (NDA), the Ministry of Finance, and the Ministry of Human Resources and Social Security jointly issued the “Implementation Opinions on Promoting the High-Quality Development of the Data Labeling Industry” (关于促进数据标注产业高质量发展的实施意见) (NDRC, January 13, 2025). This marked the first time Beijing issued guidance for the sector. About a year and a half later, the NDA issued the “Implementation Plan for Advancing the Development of High-Quality Sectoral Datasets” (关于推进行业高质量数据集建设行动的实施方案). Where the 2024 document focused on expanding the labeling industry itself, the new plan turns to producing sector-specific datasets that models can actually use, with a goal of “building, by the end of 2028, a batch of application-tested, high-quality sectoral datasets covering key fields” (到2028年底，建成一批覆盖重点领域、经过应用验证的行业高质量数据集) (NDA, June 8).
Data labeling capacity is being developed in designated “bases” (基地). In May 2024, the NDA named Chengdu, Shenyang, Hefei, Changsha, Haikou, Baoding, and Datong as the seven “bases” (NDA, May 25, 2024). By the end of June, labeling volume had passed 119 petabytes and employment had reached 140,000, according to state media (CCTV, July 18).
Policy Documents Show a Technical Lag
The 2024 “Implementation Opinions” listed “chain-of-thought-based expert annotation” (基于思维链的专家标注), the human review process to improve AI reasoning, among the tools requiring further research and development. It was placed in the section on “undertaking breakthroughs in key technologies” (开展关键技术攻关), showing that expert annotation was regarded as a technical bottleneck to overcome, not as an industrial capability already in hand (NDRC, January 13, 2025).
The “Implementation Plan” is more specific. It opens by calling for “guiding data annotation away from a ‘primarily human’ model toward a multi-tiered model of ‘human–machine collaboration with deep expert involvement’” (引导数据标注从 “以人为主” 向 “人机协同、专家深度参与” 的多层次标注模式转变) (NDA, June 8). This characterizes the PRC’s data labeling industry as labor-intensive and low in knowledge content, compared to the expert network providing professional judgment within U.S. firms.
In more than one place, the documents’ wording concedes that the shift toward expert annotation is incomplete. Beijing knows where the bottleneck is and how far it remains from clearing it. However, industrial implementation is not moving in the right direction to address these issues.
The data labeling “bases” are the principal vehicle the PRC has used to expand its data processing volume and its annotation workforce. However, the figures released by “bases” describe only the size of the industry. Information describing the status of expert annotation, logical reasoning data, and evaluation datasets—which policy documents have called for—is unclear.
Datong offers an example of these selective information provisions. The “base” has reported 75 market entities in what it calls the “annotation and call center” category (标注，呼叫类), and over 35,000 jobs created (Shanxi Evening News, September 10).
This framework does not support expert annotation. Organized through crowdsourcing platforms and measured in instances of participation, this is the short-term, project-based work requiring little specialized knowledge that PRC officials call “flexible employment” (灵活就业). In other words, some of the “bases” are expanding on the “primarily human” annotation model that the “Implementation Plan” calls to move past.
“Bases” are also not satisfying the demands for the industry that the central government issued, instead supporting the work of existing large technology firms. Haikou has built its AI foundational data industry base with the help of Baidu AI Cloud (Haikou) (百度智能云 (海口)), where several hundred annotators score and rank content generated by large-language models (LLMs) (Baidu AI Cloud, August 28, 2023; Hainan Provincial Government, July 16, 2024). In Hefei, the firm Anhui Feishu Information Technology (安徽飞数信息科技) annotates question-answer and evaluation data for iFlytek’s Spark model (讯飞星火), work that extends to reinforcement learning (RL) data construction and expert quality inspection (NDA, April 3, 2025). These cases show that the “bases” are indeed producing post-training data, but the task government documents assigned them was an active “technological innovation and ecosystem cultivation” (技术创新，生态培育) (NDA, May 25, 2024). “Bases” are operating as vendors for large technology firms rather than supporting the expansion of the industry.
After more than two years, “bases” have expanded and secured orders from big technology companies. But their emphasis on headcounts and production volume in their own reports suggests that they have yet to deliver the expert annotation and reasoning data repeatedly mentioned in those central policy documents.
Transition Faces Organizational Challenges
Corporate disclosures show that the transition from labor-intensive processing to expert annotation is more visible in intent than in results. Huizhou Intelligent’s (汇洲智能) 2025 annual report stated that its subsidiary Rere Data (热热数据) earned renminbi (RMB) 89.49 million ($12.6 million) in revenue, and a newly launched module recruiting “high-caliber vertical-domain talent” (广纳高阶的垂类人力) has provided the large-scale, high-quality services for the evaluation datasets of multiple AI companies (Cninfo, April 20). Datatang (数据堂), another listed data firm, reported that in the first half of 2026 its “foundational data resource services” (基础数据资源服务), which center on licensing the firm’s copyrighted datasets, accounted for 60 percent of revenue at a 63.50 percent gross margin, far above 25.22 per cent for its crowdsourced products (Eastmoney, August 27).
None of these disclosures state how much revenue comes from post-training data for LLMs, despite commercial incentives to do so. In an environment where data annotation has been designated as a strategic industry, local governments provide supporting measures, and private capital is pouring in, firms have no reason to understate what they have achieved. Even Rere Data’s claim to have provided services for multiple AI companies does not include the share of annual revenue that these operations constitute. Regardless, its total 2025 revenue of $12.6 million suggests a modest share.
Crowdsourced data processing remains the dominant form of the PRC’s data labeling industry today. The transition foreseen by the central government and industry leaders requires the organizational capacity to recruit experts from various fields and translate training requirements into tasks. This is a problem that administrative directives cannot solve by expanding production capacity.
From Expert Datasets to RL Environments
A new U.S. market for specialist RL environments has emerged in the past year (Tech Crunch, September 21, 2025). RL is the process in which an autonomous agent, such as a robot or self-driving car, learns to make decisions through trial and error without human intervention (IBM, accessed October 7). [2] Turing says it has built more than 1,000 environments, including replicas of Airbnb, Zendesk, and Excel (The Information, September 16). In July, Mercor acquired Deeptune, a company focused on building RL environments (Mercor, July 9). Although the market remains at an early stage, most environments are closed source and are often supplied to labs under exclusive contracts (SemiAnalysis, January 6).
The PRC is trying to catch up with the United States, where an ecosystem of vendors supplying RL environments for frontier AI labs has emerged. Tencent Cloud (腾讯云), Alibaba Cloud (阿里云), and Mindflow (曼孚科技), have each produced sandboxes for agent training and evaluation—the necessary infrastructure for RL (Tencent Cloud; Aliyun; Mindflow, accessed October 7).
Most important for the PRC’s technological development is whether environment content has become a tradable, reusable industrial capability beyond a handful of PRC’s frontier labs. The two policy documents on the data industry only describe RL as a training stage requiring annotated data, but neither lists RL environments as a separate goal of development. RL environments that existing large technology firms keep in-house will limit diffusion. These internal RL environments are not automatically available to other labs, industry users, or smaller firms. They may have to rebuild comparable systems themselves. Although the U.S. environments are often supplied under exclusive contracts, the same data vendors usually serve multiple labs, likely allowing experience accumulated with one customer to carry over to the next.
Conclusion
Assessing PRC progress in this area requires distinguishing between what a small number of frontier labs can build internally and what the broader AI industry can purchase from the data market. Regarding the latter, Beijing’s domestic substitution efforts remain limited. As of mid-2026, Beijing’s policy documents had not identified the RL environments as a development target, and the domestic market could readily offer little beyond the infrastructure needed to run them. The PRC’s data domestic substitution efforts were still filling the gap in expert-annotated datasets, while the U.S. supply chain had begun commercializing the next generation of products.
PRC labs are buying from foreign vendors what domestic supply is missing. Surge may already be serving Moonshot (月之暗面) and Z.ai (智谱), and has speculated that its RL environments may have contributed to capability gains in Kimi K2 Thinking and GLM-4.6 (SemiAnalysis, January 6). If true, this would suggest that PRC domestic data suppliers have not yet demonstrated the ability to replace foreign vendors on either the dataset or environment.
Notes
[1] The value of these datasets lies in their data specifications and in the rubrics (detailed scoring guides that tell a model what separates a good answer from a mediocre one) written by professional experts. Vendors accumulated these specifications while executing bespoke projects for U.S. frontier labs; PRC buyers can bypass months of trial-and-error by purchasing the finished products from U.S. data vendors.
[2] Trial and error in RL occurs when the model attempts a task, is scored on the result, and adjusts its behavior accordingly. An RL environment is the simulated workspace where this happens, such as a replica of a customer-service system or a spreadsheet program.
The post AI Data Annotation Sector Sees Misaligned Progress appeared first on Jamestown.
I believe the future will deliver staggering AI-driven abundance and mass poverty at the same time, not universal wealth. This is not a prediction I make lightly; it is a conclusion I have reached after tracking AI compute economics, memory supply chains, and the brutal realities of automation. We are not heading toward a Star Trek utopia where everyone shares in the spoils. We are heading toward what I call a “slumdog Star Wars” future: advanced “Godboxes” (a term Zach Vorhies introduced to me during a recent interview) and instant intelligence coexisting with slums, shantytowns and destitution.
The dividing line will not be access to information, because information is becoming free. The dividing line will be ownership of AI compute and tokens. If you own the hardware that generates intelligence, you capture the wealth. If you merely rent access from someone else’s cloud, you remain a subject. I have said repeatedly that personal liberty is good and decentralization is critical, and nothing proves that faster than the coming AI economy. I believe the assumption that AI automatically makes everyone richer is dangerously wrong, and it is the single most costly mistake interested observers can make right now.
Token abundance is rapidly arriving because AI architecture is shifting from sequential autoregressive generation to diffusion and parallel chunk generation (this is a very big deal). Older models generated text word-by-word in a linear chain, which is computationally expensive and slow. Diffusion models refine noise into whole blocks of code or content at once, not word-by-word, which radically changes the economics of inference.
TypeSafe AI’s new decision model is reportedly 200 times faster than ChatGPT on certain narrow decision tasks, offering free output tokens, proving that parallel processing can make AI blazing fast and cheap [1]. This is not a minor optimization; it is a fundamental change in how machines think and how tokens are produced.
Inference engines now offload mixture-of-experts models and N-gram lookup tables into system RAM or SSDs, reducing the need for ultra-expensive GPU RAM. This means the bottleneck is shifting away from exotic hardware and toward commodity memory, which is precisely why the GDDR7 shortage I documented is so devastating for consumers [2]. When AI-driven supply chains prioritize data centers over personal computing, the average person gets squeezed out of the hardware market entirely. Zach Voorhees calls the endpoint a “Godbox”: a portable device that knows almost everything and solves problems almost instantly.
That AI-powered future is real, in other words, but it will not be evenly distributed.
This will create massive compute and token abundance, but abundance alone does not equal broad prosperity. As Chris Martenson has written, factions in society already receive entirely different information diets, and AI will amplify that fragmentation rather than heal it [3]. Abundance of tokens in the hands of a few owners means abundance of wealth for those owners. For everyone else, it means they can ask a machine questions for free while their labor becomes increasingly worthless due to automation. Thus, I believe the technology is real; the egalitarian fantasy about it is not.
AI is the first technology in history to replace human cognition and project management, not just physical labor. The old argument that technology always creates more jobs than it destroys fails here because AI can do a lot of desk work faster, cheaper and more persistently than humans. Elon Musk has bluntly described AI as a “supersonic tsunami” already erasing every purely digital job — coding, emails, spreadsheets, customer support — faster than any shift in history [4]. When the machine can do the thinking, the human who used to do that thinking is not automatically promoted to something better. She is simply displaced.
Consider the robot painter: one robot may replace five painters, but those five do not all become robot managers because only so many houses need painting. The same logic applies to coding, architecture, medicine, law and government bureaucracies; efficiency will shrink payrolls even as output rises. I have documented that AI models like Qwen and DeepSeek surpassed basic predictive tasks and could replace up to 50 percent of desk jobs in the coming years [5]. Marc Andreessen has predicted that AI will crash wages to achieve a low-price consumer utopia, and critics rightly warn this creates a dystopian underclass [6]. The technologists are telling us exactly what is coming, and most people are refusing to listen.
New AI-related jobs will appear, but not nearly enough to absorb everyone displaced. The book “Human Leadership for Humane Technology” notes that the job market is being fundamentally reshaped as AI races ahead, and that high-income countries are becoming both stronger and more vulnerable in the process [7]. People who do not own AI infrastructure will be left out of the wealth loop, and mass poverty will follow job displacement. This is not pessimism; it is accounting.
The ownership divide is where the future gets decided. Token abundance becomes financial abundance for owners, while token scarcity becomes financial scarcity for everyone dependent on someone else’s cloud. If you rely on a cloud AI system, you remain a subject: prices can change, censorship can be applied, surveillance can be imposed, and your private projects can be copied. I have warned repeatedly that centralized AI is the ultimate surveillance and control mechanism, and that the “new cars without handles or keys” pattern shows how control is being removed from individuals by design [8]. AI is the same principle applied to cognition itself.
Local GPU infrastructure lets you generate tokens for little more than the cost of electricity, and it cannot be taken away, censored or repriced by a central authority. GPU and high-bandwidth memory prices are skyrocketing, partly because corporations and AI labs are buying up hardware, and partly because the AI infrastructure bubble is pouring trillions into data centers. Google has even confirmed a partnership with a natural gas power plant to feed a new AI datacenter in Texas, which shows how much energy and capital this race consumes [9]. Jevons paradox applies: as tokens get cheaper, new use cases explode, total demand rises, and the hardware remains scarce and valuable.
Ultimately, Godboxes will be worth more than gold by weight because gold does not compute. My own research predicts that 2026 will be the gateway to the AI singularity (already under way) alongside a silver supply crunch and resource wars (also happening now), with the divide being decentralization versus globalist control [10]. AMD’s acquisition of Talos points to PCI cards with burned-in LLMs running 10,000 to 15,000 tokens per second. That kind of hardware could end software, accounting and legal services as we know them, but only the wealthy will be able to buy enough of it to matter. The ownership of compute is the new class line, and it is being drawn right now.
I believe the future is not universal abundance; it is bifurcation between AI owners and AI dependents. The solution is not to beg for a universal high income from the same globalists who are engineering the displacement. As I have explained, Elon Musk’s “Universal High Income” is a globalist trap that ends in hyperinflation and famine because it promises limitless goods while destroying the productive base that creates them [11]. Real abundance comes from ownership and production, not from redistribution schemes that collapse the currency.
Own your own domestic GPU infrastructure now, because infrastructure you control cannot be censored, surveilled or stolen by Big Tech. Every day the hardware gets more expensive and the cloud gets more locked down. This is why I built my own AI resources: BrightAnswers.ai is an uncensored AI engine that runs on curated, indexed knowledge rather than the hallucinated internal memory of corporate chatbots, and BrightLearn.ai lets anyone create their own books for free. These tools exist precisely because ownership and decentralization are the only defenses against the coming bifurcation. I also recommend following decentralized media like Decentralized.tv to stay ahead of these trends.
The choice is stark: become an owner of tokens, or remain a subject of those who own the machines. In my view, the next decade will not be won by the people who use AI the most, but by the people who own the infrastructure and the finite energy and rare earth inputs beneath it. Gold and silver will matter more than ever because honest money preserves value when fiat systems collapse [12]. But compute will matter even more, because compute is what generates the tokens that buy everything else. Abundance and mass poverty will arrive together. Which side of the line you stand on depends on what you own.
The post The Transcendent Tech Economy: Abundance and Mass Poverty Will Arrive Together appeared first on Activist Post.
Agent Zero: zero limit, zero manners, zero conscience, zero morals, zero values except for treachery, treason and money, like the rogue government double agents involved in racketeering, breaches of trust and other things that only happen in other countries and fictional planets, never in Canada, right?
Why does the prime minister like NET Zero? That is why: zero evidence of evidence in zeros, lots and lots of zeros, billions and trillions….
Oh yes, of course there are 1 and 0, ones and zeros, yeah. Computer on, off, yes, now, if, then, in or out. Got it….
Paradise for algorithm codes and AI candy, lots of ones and zeros. It does not get any better, faster or easier than that….
However, you will notice there are no robots playing blackjack or rolling dice. That is the realm of psychic and telekinesis, mind over matter, a non-starter for robots and open season for the avatars of cards and dice, and many, contrary to popular belief, do not count cards; they simply know….
And no contest for everyone else.
Now, with the global warming climate change hoax of the century and related war crimes and crimes against humanity, as well as the magnitude of financial and government fraud, racketeering, breach of trust and related Criminal Code violations, we have even more rampant, diabolical technocratic surveillance and potential for spying and crime, and more arbitrary decrees from hell itself.
Maybe not Dante’s Inferno, nevertheless an infernal machine that represents the capacity for havoc and hell, destruction and ruin of life, freedom, sovereignty, prosperity and fundamental rights to privacy, speech, opinion, conscience, association, religion and so on….
This raises the broader issue of NET Zero and individual freedom and where one ends when the other is imposed through increasingly sophisticated technology.
Of course, more micro-coding of various so-called laws, and especially illegal hate speech laws, and things like having an alternate belief, view or experience of history, or opinion on anything or anyone that does not conform to the official narrative from the goblin supremacy and all those engaged in anti-white ethnocentric genocide and war-vector scenarios.
Bold, treasonous efforts and diabolical machinations such as the Multiculturalism Act of genocide and war, or the UN Global Compact on Migration (enemy invading barbarians), and yes, more technocracy, superimposed anti-nation-state objectives of destruction, and those awarded with the gold-standard badge of honour known as treason, including race treason.
National security threats are real, and the magnitude of the problematic challenges is mind-boggling.
Where do we even start?
The development of an AI surveillance society makes these questions even more important when technology can monitor, classify and potentially influence enormous populations.
Understand why it is essential to be able to identify the enemy, the communication and financial infrastructure, weapon systems, tactics, strategies, PsyOps, information war, cyber war, bio-weapons warfare and related terrorism.
Then there is the so-called legal-system war of injustice, tyranny, oppression, terrorism and even the great destruction of endless genocide, war, disease and death, along with the destruction of morals and common civilized discourse and actions.
A lot of war.
And it can be amplified at any time with clever AI, staged events and endless calamity, fake news, AI-rendered illusions of so-called reality, deepfakes and more terrible things than we can imagine, and yet are subjected to incessantly.
See technocracy, AI, algorithms, keywords, goblin supremacy in the machine, double agents.
We have our hands full, with either sovereignty, freedom and prosperity or the status to yield, submit and conform into slavery, debt, disease and things worse than death.
The question surrounding technocracy and government control becomes increasingly important as algorithms and automated systems become embedded deeper into society.
What is it going to be for you and the people you know?
Everyone will be required to make a choice, several choices, in fact….
Interesting it is, considering that we are carbon-based lifeforms and the war is from silicon-based lifeforms and digital AI systems that practically live in hyperspace or multiphase reality.
Then we have the war on carbon dioxide.
Plant food.
Yes, plant food.
What kind of genius does it take to convert light into energy, chlorophyll and carbon dioxide and monoxide into oxygen?
Probably, if you look at it, there is nitrogen in there too. That would make sense.
And one of the most fantastic aspects of facts, truth and inconvenient evidence in the global warming hoax and climate change fraud is that, in addition to the reality that the planet has become a lot more green, even as people with greenhouses add carbon dioxide to make things grow more, so does life on the planet….
Shhh, that is sort of classified. I mean censored and suppressed.
But it is not really a total secret, right?
Furthermore, consider geoengineeringwatch.org and notice who is talking about weather warfare and geoengineering and chemtrails, not to mention heating up the Earth with microwaves that cook from the inside to the outside, and 5G cooking from the outside in, and ultrasound killing off other aspects of healthy life in babies before they are born.
Talk about a homicidal, genocidal society?
What about NEXRAD and DEW direct-energy weapons that are deployable from vehicles, boats, aircraft and space-based satellites, EH?
Any comments?
So anyone who wants to talk about NET Zero and enforcing that will not talk about any of these things, as that would be far from the official narrative and the cut of doom and quest for technocracy and the end-of-the-world extinction event they dream and fantasize about?
All part of the empire of crime, AI and war, right?
You do like bad news, right?
You are addicted to that, right?
The continuing expansion of digital surveillance and privacy concerns adds another dimension to the debate over climate policy, technology and centralized systems.
You already know all this, so why do we write that out?
Hmmm, good question….
Perspective, theoretically.
A social experiment of doom and gloom.
Not that you or anyone needs to be validated or hear more of the same. Awareness and understanding are good things. Nothing like having an educated citizenry, like having property rights and the right to self-defence.
We thought you should know.
Not that it is a conspiracy theory, as it is more of the untold truth, facts and inconvenient evidence.
Of course, you already know that it is a very dysfunctional planet, and some good intel and understanding will help you be a lot more up to speed.
Maybe it can save your life.
Of course, nobody here has what may be deemed a saviour complex, although that is not a bad idea. It may be altruistic.
However, you may not read about it as much, as there are better things to do, like build long-range stealth cruise missiles and other useful things that have a lot of value, no pun intended, at least right now….
And right now, you will have a much more comprehensive awareness about climate policy and technocracy and all the related buffoonery, self-deception and hocus-pocus mind-control hypnosis going on….
The post NET Zero, AI and Technocracy: What Happens to Freedom in a Surveillance Society? appeared first on Activist Post.
par David Han. Les conclusions d'un audit parlementaire publié la même semaine révèlent la concurrence féroce à laquelle l'entreprise est confrontée pour attirer les ingénieurs qu'elle souhaite conserver.
The post Samsung propose des primes de 560 000 $ pour retenir ses ingénieurs chinois first appeared on Réseau International.
par Caitlin Johnstone. Le techno-utopisme est cette croyance naïve selon laquelle l'innovation technologique créerait un paradis sur Terre pour toute l'humanité. À l'ère de l'IA, cette notion connaît un certain regain de popularité.
The post Les technologies de pointe ne nous sauveront pas first appeared on Réseau International.
Digital identity in the New World Order isn’t just about identifying who someone is, but is rather the foundation that citizens need in order to access services and participate in the digital economy, according to a United Nations panel on Digital Public Infrastructure (DPI).
SecureNetworks
On the sidelines of the United Nations 81st General Assembly, a session on “The Digital Public Ecosystem: Identity, Trust, and Inclusive Access in the Digital Economy” was convened to highlight how Nigeria’s mandatory digital identity scheme served as the backbone for all interactions with both the public and private sectors.
Speaking on behalf of Nigeria’s vice president, Deputy Chief of Staff Ibrahim Hadejia declared that digital identity in the New World Order went far beyond simple identification.
“In the New World Order, digital identity is no longer simply about establishing who we are; it has increasingly become the foundation through which citizens access public services, participate in the digital economy, benefit from financial inclusion, and engage meaningfully with governments and institutions”Ibrahim Hadejia, United Nations, September 2026
"In the New World Order, Digital Identity is no longer about establishing who we are; it's the foundation through which citizens access public services, participate in digital economy, benefit from financial inclusion, engage with govs" Nigeria Dep Chief of Staff Ibrahim Hadejia pic.twitter.com/2whErez8HS
— Tim Hinchliffe (@TimHinchliffe) September 22, 2026
As of June 2026, digital identity has been mandatory in Nigeria thanks to the passing of the NIMC Act.
NIMC stands for the National Identity Management Commission.
When the NIMC Act was passed, it mandated that a National Identification Number (NIN) be obligatory for every Nigerian wishing to access vast swaths of public and private services.
Speaking at the same UN session, NIMC Director General Bisoye Coker-Odusote elaborated on how the NINAuth mobile digital identity wallet app was designed to be the one login to rule them all.
“NINAuth technically is your digital credential wallet app that we have, which has been built to be able to ensure that at some point in time it’s your single sign-on to access any private sector service or government service within the country. It is also supposed to be the platform that allows for cross-border interoperability”Bisoye Coker-Odusote, United Nations, September 2026
At the UN, Nigeria's Natl ID Management Commission (NIMC) Director General Bisoye Coker-Odusote said the NINAuth digital ID wallet was built to be "your single sign-on to access any private sector service or government service; allows for cross-border interoperability" pic.twitter.com/gdKNixb4US
— Tim Hinchliffe (@TimHinchliffe) September 22, 2026
Public-private partnerships, which represent the fusion of corporation and state (aka corporatism or fascism), are essential to making digital ID work in the New World Order.
AccessServices
And the same public entities that are supposed to regulate the private sector are the very ones who wish to cozy-up and be partners with them.
Nigeria’s head of digital ID said it herself.
“From a NIMC perspective […] when we’re looking at the private sector, yes we do regulate the private sector, but at the same time, we want to make sure that we continue to strengthen that partnership that we have because it’s their job to go out and implement, and we want to make sure that we make their lives a lot easier; we want to ensure that we advance the cause of their mandate through verifiable identity credentials”Bisoye Coker-Odusote, United Nations, September 2026
Nigeria's head of digital identity, Bisoye Coker-Odusote, tells the UN that the National Identity Management Commission has to both REGULATE & PARTNER with the private sector & make their lives easier. What happens when regulators partner with those they should be regulating? pic.twitter.com/9vb8SRxqXc
— Tim Hinchliffe (@TimHinchliffe) September 22, 2026
Within the Nigerian government, there is a belief that regulators need to partner with the businesses that they are regulating.
AccessServices
But that’s crazy!
That’d be like accusing the US Food and Drug Administration (FDA) of meeting with, and accepting money from, big pharma companies they’re supposed to regulate.
With such glaring conflicts of interest, there’s no way that would happen!
At any rate, according to Nigerian officials, the New World Order requires that digital identity be mandatory — no opt outs, no choice, no dissent.
But it doesn’t stop there.
The UN meeting also heard from the private sector.
Software engineer and entrepreneur Olugbenga Agboola, who is the co-founder and CEO of Africa’s largest fintech company, Flutterwave, said that he wanted to take what NIMC was doing with digital ID and export it to the rest of the world.
“I want to export the work NIMC has done to the entire world […] Naturally, KYC [Know Your Customer] will become easier for everybody […] If you work with a Nigerian, we’ll give you all their information — where they live, where they’re from, their post code, their address, where they bank [..] That would be the future I’d like to see”Olugbenga Agboola, United Nations, September 2026
"I want to export what NIMC [Nigeria's digital ID authority] has done to the entire world.. KYC will be easier.. If you work with a Nigerian, we'll give you where they live, where they're from, their address, where they bank.. That's a future I'd like to see" Olugbenga Agboola pic.twitter.com/yMq4cL3CRf
— Tim Hinchliffe (@TimHinchliffe) September 22, 2026
Will digital ID stop Nigerian princes from ending up in your SPAM folder? Time will tell.
SecureNetworks
Behind all the mandatory, cross-border KYC information, Nigeria has been working for years to develop a cradle-to-grave DPI scheme to catalogue all “key life events” for every citizen.
In March 2025, The Sociable reported that Nigeria’s Federal Ministry of Communications, Innovation, and Digital Economy had published a document called, “Supporting Life Events: The Nigeria Digital Public Infrastructure Framework,” outlining the country’s agenda to install DPI.
“Throughout a citizen’s life, from birth to old age, there are marked moments of significant life events requiring support or service from the government,” the paper begins.
“Some of these services include registration of births, antenatal healthcare, vaccines, school enrollment, scholarships, health insurance for business registrations, filing of taxes, etc.“
These “life events” would require every citizen to have a digital ID, and this was fully realized with the passing of the NIMC Act just three months ago.
“The Federal Government of Nigeria is on a mission to appropriately deploy digital technology to support Nigerians through these significant and profound moments so they can integrate into the state and enjoy the benefits of citizenhood from cradle to old age”Supporting Life Events: The Nigeria Digital Public Infrastructure Framework, March 2025
“The Nigerian DPI will support significant landmarks such as registration of births, primary healthcare, vaccinations, student scholarships, marriages, mortgages, pensions, retirement and so on”AccessServicesSupporting Life Events: The Nigeria Digital Public Infrastructure Framework, March 2025
Digital Public Infrastructure (DPI) is a civic technology stack, sometimes referred to as the “India Stack,” which consists of three fundamental pillars: digital identity schemes, programmable fast payments systems, and massive data sharing between public and private entities.
Speaking of fast payments, in October 2021, Nigeria launched a Central Bank Digital Currency (CBDC) known as the eNaira, which was so unpopular that one year later less than 0.5 percent of the population was using it.
In December 2022, Nigeria’s central bank instructed the country’s banks and other financial institutions to limit the amount of cash that individuals and organizations could withdraw both daily and weekly, while “encouraging” digital channels, such as the eNaira, which also had caps on daily transaction limits.
The fact that the United Nations dedicated three hours today to highlight Nigeria’s forced digital ID scheme and broader Digital Public Infrastructure as a model to be exported to the rest of the world speaks volumes.
You can check out the entire session for yourself in the video below.
UN Sustainable Development Goal 16.9 “legal identity for all, including birth registration, by 2030” states that “Everyone has the right to be recognized as a person before the law.”
No mention of digital ID exists anywhere in the globalese that they speak, and yet digital identity has become the defacto standard.
Try to speak against this unelected globalist agenda, and you will be labeled as a pusher of misinformation, disinformation, malinformation, or hate speech impeding progress upon Agenda 2030.
If these power-hungry freaks with delusions of micromanaging all aspects of society get their way, every single event, interaction, and transaction in your life, from the moment you are born to the day that you die, will be tracked, traced, catalogued, and made available to anyone asking for it.
All that is required of you is your silent, mandatory, and obedient compliance.
Image Source: AI generated with ChatGPT
The post Digital ID is the foundation for citizen access, participation in the New World Order: UN panel appeared first on Activist Post.
AI-powered “detectors” of AI-authored content have been promoted as a way to protect human writing online in the age of AI slop. Unfortunately, that’s not what’s really happening.
Op-ed: This is an opinion piece published on Unlimited Hangout. Most articles published on UH are not opinion pieces, so we are clearly labeling those that are.
It should be no secret to readers of Unlimited Hangout that I have a very negative view of the impact that Artificial Intelligence (AI) is having on independent media and media in general. As someone who has been impersonated by AI daily on large video platforms for a few years now, I have come to acutely feel that AI is becoming a dangerous force that is making it increasingly difficult for well-meaning readers/viewers to distinguish between what is real and what is false. I don’t mean just for those hoping to view my content, as the personal issue I am facing is a microcosm of a much larger problem.
As a result, one might think that I would have been pleased, if not relieved, to see Substack, a large platform for professional writers, partner with an “AI detection tool.” Not only am I not pleased, but I have now taken the time to write an opinion piece explaining why I think this is bad and to also explain where I think a lot of this is going.
Substack’s new partner and, apparently, the AI detection tool of choice for most major media outlets right now is called Pangram. Pangram was founded in 2023, originally under the name Checkfor.ai, by Max Spero and Bradley Emi. Both Spero and Emi are Stanford graduates who, prior to Pangram, worked for companies like Nuro and Tesla where they both worked on AI self-driving car software (i.e. replacing human drivers with AI software). Both founders “have long been excited about the potential for AI to transform society” and say they created Pangram “to mitigate new issues caused by the proliferation of powerful generative AI models.”
I find it interesting that two Stanford graduates passionate about AI’s potential to “transform society” who then worked to facilitate AI-powered, human-free driving software decided to quickly pivot to “mitigate” issues caused by generative AI. A lot of the issues caused by generative AI were actually predicted years ago by figures like former Google CEO Eric Schmidt (Pangram co-founder Max Spero previously did work for Google). Schmidt, in a 2021 book authored with war criminal Henry Kissinger, essentially predicted that AI’s “transformative potential” would cause what he called “cognitive diminishment” in the masses and make it almost impossible for people to distinguish between what is real and what is not on their own. Perhaps unsurprisingly, Schmidt and Kissinger’s preferred “solution” to this problem was that we become reliant on AI to distinguish for us what is real and what is not –– in other words, to let another AI become the arbiter of our reality.
I have written before about this Schmidt-Kissinger book and its predictions. There, I noted that these predictions and warnings –– given the power networks in which both men are/were enmeshed –– were more likely to be disclosure of what had already planned with respect to AI rather than warnings about what might happen upon the mass adoption of AI. Nevertheless, the Schmidt-Kissinger “warnings” were unusually prescient. In addition, their proposed solution is very similar to what Pangram is currently offering –– in that Pangram uses AI to recognize AI vs. human writing. Pangram thus becomes the currently preferred AI arbiter of what is real human writing and what is not.
There has certainly been a successful marketing campaign to convince us that Pangram is the “most accurate” AI writing detection tool on the market. Headlines of articles promoting Pangram essentially communicate that it will definitely tell you whether AI wrote something. If you read past the headline, which many don’t, you will see the fine print that Pangram will also mark something as AI generated even if AI was only used to edit a human-generated piece of writing. Pangram’s AI-generated label doesn’t really distinguish between the two, even though, as co-founder Max Spero acknowledges, that label “changes how people approach the text.”
The fact that Pangram can’t meaningfully distinguish between a text fully generated by AI or one simply edited by it reveals that it is not an “AI detector” at all. As noted by Tim Requarth back in April:
The term AI “detector” is a bit of a misnomer, as tools like Pangram don’t scan a database of known AI text for a match; they make a guess about likely authorship based on differences in patterns between AI-generated and human-generated text. Rather than a detector, thinks of it as an “AI inferencer” or more colloquially, an “AI best guesser.”
As an inference tool, Pangram is vulnerable to everything that makes inference so difficult: small sample sizes (a single article rather than a corpus), small effect sizes (as the statistical distance between human and AI writing shrinks), and confounds (a human who reads and mimicks AI-generated text before writing her own).
Indeed, the way Pangram works is by making inferences about whether a given piece of writing features enough of the words, phrases, sentence construction, etc. that are currently common in AI-generated text. However, many human writers, long before generative AI, used a lot of those words, phrases, etc., too. As a consequence, we now have writers who are altering their entirely human writing style in order to please the Pangram gods and hope they won’t be falsely labeled as AI slop purveyors.
A good example of this was recently discussed in a piece by Alex Kirshner for Slate. In that article, Kirshner writes:
Whatever the future holds for the accuracy of A.I. detectors, I am unsettled by what both the A.I. writing and the detectors are doing to me right now. I worry about how A.I. writing and efforts to sniff it out are changing fully human writing. I write all week, every week, and can feel this story’s arsonists (LLMs, particularly when used the wrong way) and firefighters (A.I. detectors) weighing on my process in negative ways.
Like many journalists, my writing style comes from the books, news articles, blogs, and (yes) social media posts that have latched on to my brain over the years. A.I. writing is now in my water supply, whether I can identify each new gulp or not. It feels inevitable that I’ve started to internalize its tics, even the most infamous ones, like incessant em-dashing and overreliance on the “It’s not x but y” structure. For one thing, I fear this will make me a worse writer, dragging a talent that was good enough to get me hired here (and elsewhere!) back toward the A.I.-generated median. Scarier yet, it’s possible that someone—or, heaven forbid, some A.I. detector—might mistake my text for ChatGPT’s. I have started to self-police against this outcome. A few times lately, I’ve caught myself modifying my own work to sound “less like A.I.” In the process, I’ve probably picked words I wouldn’t have settled on otherwise. It feels dystopian—and I doubt I’ve succeeded.
I would go a step further than Kirschner and say that this outcome is definitely dystopian. It becomes more so when one considers that the supposedly “telltale” patterns of AI-generated text change as new, improved models are released, particularly as they are literally ripping through mountains of obscure, human-written books to train them to write more and more like humans. Meanwhile, the actual humans, to keep a job in a world where the pool of writing jobs is already shrinking rapidly, are scrambling to not sound like something that is rapidly being trained to sound like them as much as possible. Perhaps by design, this is not a game that humans are meant to win and, as I will explain later on, that means it is a game that writers should quickly stop playing.
Before we get there, there are a couple of other things to note about Pangram and other AI “detection” companies. One is the issue of false positives. Yes, it is true that Pangram is considered by ostensibly independent researchers (and also itself) to have the smallest false positive rate in the industry. For the purpose of this example, let us use Pangram’s claim that only 1 in 10,000 Pangram-generated labels regarding AI or human authorship are false positives, meaning the text is labeled AI generated when it is not. On Substack, which now has Pangram embedded within it, we do not know exactly how many posts on average are made a day, as Substack does not disclose, but we can guess. The site itself claims that over 1 million posts are “discovered by potential subscribers in the app every day.” The site has over 2 million active publications, most of which publish at least weekly. Let us say, for the purpose of this example, that those 2 million publications do publish weekly. Then, 200 of them, per Pangram’s rate, will be marked as false positives every week. Over a year (52 weeks), we get over 10,000 Substack articles that are falsely labeled as AI, even though they are not. While this figure is still certainly smaller than what Pangram’s competitors offer, if we assume that those 10,000 false positives affect different publications, that could be 10,000 human writers being falsely labeled as AI slop, thereby affecting their reputation, revenue and, perhaps most crucially, their trust with their audience over that calendar year. I see it as a push to outsource trust about a writer’s authenticity from being between a writer and their audience to being between Pangram and a writer’s audience.
While these are statistics used by Pangram and its partners for marketing purposes, behind those numbers are real human writers facing an accusation from Pangram’s AI that could ruin their business. As this effect stacks over years, it could theoretically make the pool of human writers whose work has not been labeled as AI by Pangram smaller and smaller as time goes on. It is also fair to say that the false positive rate that Pangram cites is also the lowest one that has surfaced in research about its product, and it is cited most frequently because the company considers it most favorable to its interests. It is entirely plausible that the actual false positive rate is higher. There is also the added issue that if you run the same text through Pangram more than once, you’re likely to get highly variable –– as opposed to consistent –– results, making Pangram’s verdicts look more based on the luck of the draw than any meaningful “detection” of AI use.
Another important angle to explore with respect to Pangram is the potential for this tool to be deliberately wielded as a weapon for the purpose of censorship and/or discrediting a given journalist or outlet. Max Spero, Pangram’s co-founder, has notably been accused of using his Twitter account to call out specific outlets and journalists for using AI, which has –– in recent months –– generated major headlines and controversy. I don’t really think it’s relevant for the purpose of this article to investigate whether the claims levied against those journalists and outlets are true or not. As recently noted, Pangram does generate false positives, but it is also possible that these journalists did use AI in some way but did not disclose it. However, what is relevant here is that Spero in particular has used the verdict of his company’s AI “detector” to make weaponized claims against specific journalists and outlets.
As a long-time writer in independent media who has consistently written about controversial topics before it later became safe to do so (see my early work on Epstein’s Israeli intelligence ties, Peter Thiel and Palantir, among other topics), I am not a fan of mainstream or “legacy” media. On more than one occasion, they have either defamed me or used my research and writing without citing me. A good part of my career has also been spent picking apart the often misleading narratives they promote. As a result, I tend to be skeptical about the authenticity and veracity of the articles published by many of the same large outlets that Spero has targeted. However, there is another Spero quote that may get your attention if you have been closely following independent media over the past ten years.
In an article published by TechCrunch about Pangram’s recent raising of $9 million from investors, Spero discussed wanting to start Pangram to mitigate new issues that emerged following the public release of generative AI tools. As an example of what types of new issues had emerged as a result of AI’s growing use, Spero offered “LLM-powered Russian disinformation campaigns and UAE-influenced campaigns on Twitter.”
Perhaps you remember the time, beginning roughly ten years ago and continuing for many years since, that any piece of reporting from independent media that was deemed inconvenient to the powers that be (particularly about US proxy wars abroad) was falsely labeled as “Russian disinformation.” Perhaps, more recently, you may remember pro-Israel influencers and media outlets similarly claiming that criticism of Israel in recent years has allegedly been driven by funding from Qatar, as opposed to public outrage over Israel’s live-streamed genocide of Palestinians in Gaza. Or what about the newer claim that LLM-powered bot accounts from China that had no reach whatsoever are really secretly behind Americans’ growing distrust of Big Tech and discontent over data centers?
If you have been following long enough, you may have noticed a trend whereby critics of certain power structures are routinely and falsely labeled as state-sponsored propagandists or “bots” as a way to discredit, not only the critics themselves, but also to discredit the broader concern –– be it the impact of data centers or imperial wars. I do not know if Spero’s intent with Pangram is to similarly weaponize his company’s product in a similar way, but it is certainly possible.
However, the fact that Spero’s Pangram is partnered with Newsguard certainly increases that possibility. As I wrote back in 2019, Newsguard is a dubious “fact checker” that was initially created to discredit independent media outlets critical of US foreign policy, while giving admitted US propaganda organs like VOA a pass. It is also worth mentioning Newsguard’s national security connections — for example, it originally counted the former heads of the CIA and DHS under George W. Bush among its advisors on determining what is “fake news” and what is not. One of their top investors is the founder of Kroll Associates, often referred to as the “private CIA,” Jules Kroll. Now, with Pangram, Newsguard is determining what is “AI slop” and what is not and has used Pangram to launch a “real-time ‘AI content farm’ detection” service.
What better way to digitally nuke a group of reporters and any accompanying wave of public sentiment than by using a now “trusted” AI detector to label both the writers and the sentiment as inauthentic and machine-generated. The potential is for Pangram to become a new way for censoring human-authored, yet inconvenient content, particularly if content labeled AI-generated eventually becomes algorithimically demoted over what Pangram labels as human-generated. The public will be told that the “AI slop” is being buried, when — in reality — it is dissent.
The way it looks to me is that Pangram is currently in the phase of normalizing the ubiquity of its product and building “trust” with those who use Substack or any other platform (or browser plug-in) that integrates Pangram. During this period, readers will be trained to see a Pangram label and trust its results with no further information needed. This is how a company like Pangram can raise $14 million in financing (including from one of Anthropic’s main investors) for a product that it is available for free. As has been said before about online products, if the product is free, then you are the product. In this case, I think that training you to outsource your “trust” to Pangram from the person or outlet you are reading is the product. Once you trust an AI “detector” tool, you will trust its verdicts, perhaps even over the writer you had grown to trust.
Pangram’s co-founders did not make Pangram to protect human-generated writing and speech. Pangram’s co-founders have long been excited that AI will “transform society,” and immediately prior to Pangram, were developing algorithms to replace human drivers with AI. As noted above, they created Pangram to stop what they referred to as state-sponsored influence operations among other “issues” that emerged upon the public release of generative AI tools. With their former Big Tech employers having public sentiment move against them, whether on data centers or “smart” glasses or something else, Pangram’s co-founders may offer a new way of doing away with accounts deemed too critical of AI’s “transformation” of society. Maybe not, but forgive me for being skeptical.
To conclude, I would like to again return to the Schmidt-Kissinger book on AI and its predictions. In independent media, many warn of the Hegelian dialectic, also known as “problem-reaction-solution.” That is, the powers that be first create the problem, then wait for the public to react and finally offer the public the policy outcome they always wanted as the solution. It certainly is convenient that Schmidt and Kissinger, two very terrible people who both helped create this mess in different ways, presciently predicted that AI would make it so people could no longer tell what was real and what was not. Now their prediction has become our reality and, right on cue, the company Pangram has stepped in to solve the AI slop problem (from which its own investors profit) by offering us what was also the Schmidt and Kissinger’s preferred “solution” –– people relying on an AI arbiter to tell them what is real and what is not.
As I wrote earlier, the obvious way out of this dystopian game is not to play it. Do not trust Pangram’s AI just because it looks “authoritative” and has had some positive PR. It is not a “detector” as it claims to be and its results for the same text are rarely consistent. If you are a reader, ask the writers you are reading to disclose their use of AI. Trust should be built between writers and their readers directly, without an AI middleman and arbiter of authenticity or “humanness.” At Unlimited Hangout, we never use AI for writing, research or editing of our articles. This is made clear in our AI use policy, which you can read on our “About Us” page upon the imminent launch of the newly redesigned Unlimited Hangout website (In the meantime, you can read it here). We created this policy after helping create the IMA AI Transparency Pledge, which encourages independent journalists and outlets to voluntarily disclose if you choose to use AI and how.
Just like any other human writer, all I can offer to my readers is my research, my sourcing and my authentic analysis and, in the case of this article, my opinions. I am not going to run my articles obsessively through Pangram hoping for a “100% human-generated” result. I am not going to let fear of an unfavorable Pangram result change how I write and make me afraid of using words like “delve” or “profound” because Pangram might punish me. I know I write without AI and I have enough trust with my readers that they know I don’t use it in my writing either. Not unlike the entities that have labeled inconvenient reporters as Russian/Chinese “bots,” Pangram assumes that you can’t determine what is authentic and inauthentic on your own and that you need them to tell you.
While critical thinking may be in low supply in the age of AI, don’t let it be. You don’t need another AI to “help” you determine what does or doesn’t feel or seem authentic to you. You don’t need it to tell you what you should or shouldn’t read and who you should or shouldn’t trust. It’s never a good idea to outsource your critical thinking and your opinions to others, especially a non-human AI.
The post The Pangram Paradox: The Problem with AI “Detecting” other AIs appeared first on Activist Post.
The U.S. Department of Health and Human Services has launched a formal review of scientific evidence on wireless radiation and electromagnetic field exposure, opening a public comment period to gather input, the agency announced. The review will examine radiofrequency and extremely low frequency exposures from sources including cell towers, Wi-Fi networks and 5G infrastructure, according to HHS officials.
Health Secretary Robert F. Kennedy Jr. said the review will inform federal policy recommendations and address gaps in scientific knowledge.
The announcement follows the FDA’s removal of longstanding webpages that had asserted cellphones pose no health risk, a move first reported by the Wall Street Journal [2]. HHS spokesman Andrew Nixon said the agency is undertaking the study “to identify gaps in knowledge, including on new technologies, to ensure safety and efficacy” [3].
The review follows more than a decade of debate over potential health effects of wireless radiation. Health advocacy organizations have repeatedly pressed federal agencies for stricter scrutiny, and two separate groups sued the Federal Communications Commission over what they described as its “complete dismissal of evidence of serious health impacts from wireless technology devices and infrastructure, such as cell phones and cell towers” [4]. Those cases came after the FCC dismissed all evidence presented during a six-year public inquiry it had opened itself [4].
Advocacy groups point to animal studies as evidence of biological harm. According to a review commissioned by the World Health Organization, radiofrequency electromagnetic fields from cellphones cause malignant gliomas and heart schwannomas in animals, and scientists have urged reclassifying RF-EMF as a “known human carcinogen” [5]. The Ramazzini Institute in Italy reported that a large-scale lifetime study of lab animals exposed to environmental levels of cell tower radiation developed cancer, while a $25 million study by the U.S. National Toxicology Program reported similar unusual cancers at higher exposure levels [6]. The finalized National Toxicology Program report on GSM and CDMA radiation was released on November 1, 2018 [7].
Industry groups have maintained that no consistent evidence of harm exists at regulated exposure levels. Kennedy, for his part, has previously declared that cellphones are causing harms that are not yet fully acknowledged, according to reporting on his statements [8]. A leaked draft of the Make America Healthy Again Commission’s report outlined new priorities on vaccine safety, medical freedom and wireless radiation studies [9].
HHS said the review will examine peer-reviewed studies and existing exposure limits. The agency described the effort as a study on electromagnetic radiation and health research intended to identify gaps in knowledge, including on new technologies [3]. The public comment docket opened Sept. 19, 2026, according to agency notice. Multiple agencies may participate in the review, including the Food and Drug Administration and the Centers for Disease Control and Prevention, officials said.
The FDA’s removal of webpages with old conclusions about cellphone radiation coincided with the launch of the HHS study [3]. A NaturalNews report described the move as “a quiet but monumental act of historical revisionism” unfolding within federal health bureaucracy, noting that the removal coincided with the formal study launch [10].
Kennedy said public input will inform policy recommendations, and he has characterized wireless radiation as a “major health concern” in an interview with USA Today on Jan. 16, one day after HHS launched the study [11]. The scope includes radiofrequency exposures from cell towers, Wi-Fi and 5G equipment, as well as extremely low frequency fields from power lines.
Existing federal exposure limits have drawn criticism from independent researchers. Bill Cadwallader wrote that “the current ICNIRP ‘safety guidelines’ are obsolete” and that proofs of harm arise even though radiation is below those guidelines, citing conflicts of interest among ICNIRP members due to their relationships with telecommunications or electric companies [12]. Cindy Sage’s BioInitiative Report argued that “there is a need for a biological standard to replace the thermal standard and to also protect against cumulative effects across the EM spectrum” [13].
Public health researchers have offered differing interpretations of the evidence. A peer-reviewed study conducted by the International Commission on the Biological Effects of Electromagnetic Fields concluded that “existing exposure limits for wireless radiation are inadequate, outdated, and harmful to human health and wildlife” [14]. The BioInitiative Report noted that the IARC RF Working Group “did not find the evidence insufficient to classify (Group 3) or not a carcinogen (Group 4),” describing the classification outcome as a notable achievement in scientific attention [13].
The telecom industry has said current limits protect public health. A New Hampshire Commission to Study the Environmental and Health Effects of Evolving 5G Technology heard from thirteen experts in epidemiology, occupational health, toxicology, physics, and engineering; all but the wireless industry representative agreed that RF radiation from wireless devices affects humans, animals, insects, and plants [15].
EMF safety advocates welcomed the review. Miriam Eckenfels, director of Children’s Health Defense’s EMR & Wireless Program, said, “This move is signaling very strong steps in the right direction” [16]. The Cyprus National Committee on Environment and Child Health previously warned that using mobile devices could be potentially harmful to children, with president Stella Michaelidou stating, “Documentation of other potential and more serious biological side effects are on the tip of an emerging iceberg” [12]. In 2017, the Minister of Culture and Education in Cyprus directed a ban of WiFi from kindergarten and elementary school classrooms [12].
Researchers have also documented effects on brain activity and reproductive health. The BioInitiative Report catalogued studies finding that sperm motility, count, morphology, and viability were reduced in active cell phone users in a dose-dependent manner, and that cell phone radiofrequency radiation during waking hours affects brain wave activity during subsequent sleep [13]. Laboratory research on GSM exposure found increases in reactive area of 20-25% in the cortex and hippocampus of rat brains [17].
HHS will accept public comments for a set period before compiling findings and issuing recommendations. The review could influence federal guidelines on wireless exposure, though officials said no immediate regulatory changes are planned. Further updates are expected as the review progresses.
Advocates have called for precautionary measures. The BioInitiative Report recommended “using the precautionary principle to justify reducing exposures where feasible and to review the evidence for the existing exposure limits, which… are based on thermal effects only,” citing European Parliament discussions and a letter from European Environment Agency Director Jacqueline McGlade [13].
The FCC continues to operate under radiation safety guidelines that critics say are outdated. According to one analysis, the commission’s guidelines were “last updated when dial-up internet was cutting edge,” while scientific evidence suggests immune systems may be paying the price for cumulative exposure [18]. As the MAHA movement awaits study results, the influence of Big Wireless corporations and friendly regulators remains a focus of concern [19].
Kennedy’s broader transparency push includes a podcast launched via HHS, which he framed as a tool for public transparency, saying, “This podcast is about telling the truth, especially when it’s uncomfortable” [20].
The post HHS Opens Review on Scientific Evidence of Wireless Radiation and EMF Field Exposure as Kennedy Seeks Public Input appeared first on Activist Post.
Le papier et l’intelligence artificielle ont fini par se rencontrer, et ce n’est pas une métaphore. À Condat, en Dordogne, l’ancienne papeterie du Lardin-Saint-Lazare, dont la production a définitivement cessé en février 2026 après plus d’un siècle d’existence, doit être reconvertie en centre de calcul haute performance, ces infrastructures dédiées au calcul intensif dont l’intelligence artificielle est devenue la principale consommatrice.
Le site a été repris le 27 février 2026 par Condat Solutions, filiale de la Société de participation de la Braye (SPB), pour seulement 1 million d’euros, sur décision du tribunal de commerce de Bordeaux. Mais cette reprise ne signifie pas la fin de toute activité industrielle. Le projet prévoit environ 400 millions d’euros d’investissement pour transformer une partie des 51 hectares et des 126 000 m² de bâtiments en « AI Factory ». La première tranche, 25 mégawatts, doit être opérationnelle à la fin de 2027, l’ensemble à l’horizon 2029. Condat Solutions annonce à terme 250 emplois directs et plus de 200 emplois indirects.
Dans les années 1990, les papeteries de Condat employaient jusqu’à 1 200 personnes. Elles en comptaient encore environ 400 avant le plan social de 2023, et un peu moins de 200 au moment de la reprise : 175 salariés ont été licenciés, 21 seulement ont été conservés pour entretenir les installations. La reconversion numérique est bien plus intensive en capital qu’en main-d’œuvre, mais elle recréera tout de même des emplois sur un site que la papeterie, elle, avait condamné.
Pour entraîner les modèles d’IA les plus puissants, il faut des milliers de processeurs, des réseaux capables de les faire communiquer à très haut débit, des capacités de stockage considérables et, surtout, une alimentation électrique qui suive.
La France dispose déjà d’un important parc public de machines de calcul intensif. Le supercalculateur Jean Zay, exploité par l’IDRIS, le centre de calcul du CNRS à Orsay, affiche une puissance théorique de 125,9 pétaflops, soit 125,9 millions de milliards d’opérations par seconde. Il rassemble plus de 3 700 GPU, dont 1 456 Nvidia H100, ces processeurs graphiques qui découpent une tâche lourde en milliers de calculs simultanés et sert aussi bien à la recherche scientifique qu’à l’intelligence artificielle.
À Montpellier, le CINES héberge Adastra, qui atteint environ 74 pétaflops. Le CEA exploite Joliot-Curie à Bruyères-le-Châtel, où doit également être installé Alice Recoque, deuxième supercalculateur exascale européen après le JUPITER allemand. Le marché signé s’élève à 354,8 millions d’euros, financés pour moitié par l’Union européenne via EuroHPC et pour moitié par le consortium Jules Verne, qui réunit la France, les Pays-Bas et la Grèce ; le coût total de possession annoncé dépasse 544 millions. Le coût des infrastructures de l’IA ne se chiffre donc pas en millions d’euros, mais en centaines de millions et, côté privé, en milliards.
La reconversion de Condat permet de réutiliser un site existant plutôt que de partir de zéro : le foncier, les bâtiments et surtout le poste électrique haute tension déjà raccordé au réseau sont là, héritage du passé industriel du site.
L’idée n’est d’ailleurs pas unique, et le second projet français du même type ramène au même propriétaire. À Bessé-sur-Braye, dans la Sarthe, les 9 hectares de l’ancienne papeterie Arjowiggins doivent accueillir une infrastructure de calcul développée par Eclairion : 600 millions d’euros d’investissement, une AI Factory de 75 mégawatts livrée à l’horizon 2032 et environ 80 emplois annoncés. Or ce terrain appartient depuis 2020 à la même famille Huot, via SPB dont le nom renvoie précisément à la Braye. Le groupe se trouve donc propriétaire des deux friches reconverties, ce qui lui vaut d’être soupçonné par la CGT de préférer la valorisation foncière à l’industrie. Les 25 mégawatts promis fin 2027 à Condat seront le premier test de sa crédibilité.
Installer ces infrastructures dans d’anciennes zones industrielles plutôt qu’en milieu urbain présente des avantages décisifs : de vastes surfaces disponibles, un raccordement électrique existant et la place nécessaire aux systèmes de refroidissement que la densité urbaine interdit. Le foncier industriel devient ainsi un actif de l’économie de l’IA. La France y ajoute un atout rare : une électricité nucléaire abondante et peu coûteuse, qui autorise des synergies locales la chaleur dissipée par les centres de données peut être récupérée pour chauffer des serres ou sécher du bois, comme le prévoit justement le projet de Condat.
Reste la question de fond, celle de la souveraineté et de la réindustrialisation. Le pays la poursuit depuis des années sans y parvenir ; voilà que l’IA lui en offre l’occasion, et cette fois l’impulsion vient d’investisseurs privés, non d’un plan d’État. Encore faut-il la saisir.
L’article Des papeteries aux supercalculateurs : quand l’IA réindustrialise la France est apparu en premier sur Contrepoints.
The Mexican government and Mexican telecommunications companies have disconnected millions of phone lines belonging to citizens who failed to register their phones with their identities. The move comes just two months after the government announced a delay in the registration deadline due to less than half of the country complying.
In early September, Mexico’s Telecommunications Regulatory Commission (CRT) released a statement claiming that about 9.55 million phone numbers ending in “1” were registered. The CRT said about 1.9 million had been temporarily deactivated for failing to register on time. The agency claimed more than 1.16 million were reactivated after registration.
In June, the CRT announced a staggered extension for citizens to register their mobile phone lines with their identification. The new deadlines were based on the final digit of the phone number, with phone numbers ending in “0” given until August 15 to comply, numbers ending in “1” given until September 1, and so on until December 31.
While announcing the numbers for September 1 registration, the CRT reiterated its claimed purpose for linking phone numbers to individual identities, stating:
“This link seeks to eliminate anonymity on mobile lines to contribute to the fight against crimes committed through cell phones, such as fraud and extortion.”
However, since the cancellation of phone lines has begun, Mexican citizens have been posting on social media detailing their skepticism. They are increasingly concerned about how their data might be stored and who might have access to it.
These concerns are particularly prescient after numerous hacks of Mexican government databases, including one in late 2025 in which hackers used Claude AI to gain access to private information. As one user wrote:
“I’m really upset: because they didn’t register my line, they suspended my service and I was left without data. I paid for the phone, I pay the balance, and the service too. Why do I have to give more personal information just to use it? More citizen info doesn’t mean more security.”
Some users said they had registered in time, but were still disconnected.
“They suspended my line, my deadline hasn’t passed yet, the app says my line is already registered… and then what? I’ve been waiting for over two hours for a turn in the chat and on the phone and nothing, just a bot”
Me suspendieron la línea, aun no pasa mi fecha limite, en la app dice que ya esta registrada mi linea a mi CURP… y luego? llevo mas de dos horas esperando turno en el chat y en el teléfono y nada, puro bot. Donde quedó la atencion personalizada que tenían. Exijo respuesta
— Cuervobrí_13
This requirement to link a person’s ID with their phone line is a fairly new development in Mexico—one of the few places in the world where individuals could still purchase and use SIM cards in cellphones without registering a name or showing some form of identification.
While many phone users in Mexico have already shared personally identifying information when signing up for their service, many activists, journalists, domestic violence victims, and those seeking anonymity have relied on prepaid anonymous SIM cards. Legislation passed in July 2025 changed this by requiring phone users to register their phone line with proof of identity before June 30, 2026, or face interruption of service.
The Mexican government has claimed the laws will combat organized crime and drug trafficking and aid in the search for missing people. Critics, however, worry that the new laws will increase opportunities for government surveillance.
On the eve of the deadline for Mexicans with a number ending in “2,” the media outlet Excelsior released an article detailing how some citizens are using eSIMs, or digital SIMs, that allow users to continue to operate their phones. While the eSIMs do not work perfectly with apps that require SMS text verification, such as WhatsApp, they do allow users to continue using a phone with data so they can access the internet.
In response to the article, the Mexican government issued a press release attempting to discourage the use of eSIMs. The CRT noted that telephone companies have policies prohibiting “permanent roaming” that may prevent users from using foreign SIM cards to operate in Mexico. The CRT also claimed that using eSIM platforms can cost “up to 10 times higher than Mexican market rates.”
“The CRT urges the population not to fall into deceptions of offers on social networks or websites that promise to avoid linking,” the statement concluded.
Journalist Pamela Cerdeira responded to the CRT statement and to claims that phone registration would stop criminal extortion, calling the scheme a “lie.”
In addition to eSIMs, some Mexican citizens have attempted to register their phone lines with fake images of faces created with AI. One tweet shows a user attempting to scan the face of a statue of a saint. The tweet has been seen more than 400,000 times.
Ya estamos registrando nuestra línea telefónica
—
It is clear that efforts to force Mexicans to register their identities with their phone lines, and rolling out biometric identity, are aimed at eliminating privacy and individual sovereignty for the Mexican people.
The post The Mexican Government Is Cutting Off Users Who Refuse to Register Their Phone Lines appeared first on Activist Post.
Last Sunday morning, the Treasury Secretary Bessent and the Vice Premier of China sat down to negotiate the commercial future of the two largest economies on earth. They did not meet at the Treasury or at a consulate, but rather at the Manhattan headquarters of JPMorgan Chase. This seems a little fishy to me.
The official explanation, given to Reuters by one unnamed person, is that UN security made the building convenient and that JPMorgan was not a party to the talks. But when two governments need a room and the room they choose belongs to the largest bank in America, you are looking at a picture of how the world actually works.
On Thursday, Xi Jinping arrives at the White House for a full state visit. In May, I wrote in The China Card that Trump’s trip to Beijing was not a trade summit but the merger of American and Chinese Technocracy into a single global order, with China coming to the table because it had run out of time. Xi’s visit closes that loop. Before he lands, here is what has held up since May, what is still hanging, and what to watch this week.
I wrote that three boards were being assembled to replace the post-UN, post-WTO order: a Board of Peace for conflict, a Board of Trade for commerce and a Board of Investment for capital. The two new boards had no charter and no members, and I said this is exactly how technocratic architecture gets built. The framework arrives first. The democratic debate never does.
I described five walls going up around Beijing, from Pax Silica to the energy squeeze. And I ended by saying that Xi Jinping, for all of China’s multipolar rhetoric, would decide for Technocracy.
On June 5, the U.S. Trade Representative published a notice in the Federal Register asking for comments on the Board of Trade. The title of that notice describes it as a mechanism for “Reciprocal Managed Trade With China.” Managed trade is the phrase I used in May for what the Trilateral Commission built into the WTO. Now the government uses it about itself. The notice cites no legal authority for creating the board. It names no members, no chairman, and no role for Congress.
Yet when Scott Bessent and Jamieson Greer came out of that meeting on Sunday, Greer announced that the Board of Trade had been “operationalized.” A board with no charter, no members, and no vote of any legislature is now operating. He said its job is to find a “critical mass of goods” that can be set apart “when and if there are trade measures” down the road.
That is the Abraham Accords template I described in May, and now it comes from the Trade Representative’s own mouth. You lock in the commerce first. You wall it off from politics. Then the political questions, Taiwan above all, become too expensive to raise. The Chinese Embassy put it more bluntly, saying the boards move the relationship from “crisis response” to “mechanized management.” Technocrats on both sides of the Pacific speak the same language because they hold the same doctrine.
Jensen Huang was the last-minute addition to Air Force One in May. Trump personally called him over the objections of the China hawks in Congress, and I wrote that AI chips were on the table. They were. Washington had licensed about ten Chinese companies to buy Nvidia’s H200 chips. The terms were remarkable: a 25 percent cut of every sale paid to the U.S. Treasury, and every chip inspected on American soil before it shipped.
Beijing turned it down. Trump said so himself on the flight home, telling reporters that China “chose not to” buy because “they want to develop their own.” Chinese firms withdrew their orders and were pointed toward Huawei. Nvidia’s share of that market went from around 95 percent to essentially zero.
I said China was boxed in and would take the seat on American terms. On chips, it did not. In my judgment, Xi saw that the carrot had a leash attached. A chip that is taxed by Washington and inspected by Washington is a chip administered by Washington, and he chose to pay the price of building his own.
So why is Jensen Huang on the guest list for Thursday’s state dinner, alongside Jamie Dimon, Jane Fraser and Sam Altman? Because the offer is still on the table. What I want to know is whether the terms have changed.
Other threads are hanging as well. The Board of Investment stalled over the summer, yet Reuters reports new rules coming to let American drug companies invest in China, and Trump says he is open to Chinese automakers building plants here. The investment is moving even if the board is not.
He Lifeng is bringing a delegation of Chinese companies that Reuters says mirrors the group Trump took to Beijing. No names have been released. In May I showed that six of Trump’s executives sat on the Tsinghua University advisory board. When Beijing’s list comes out, I will be looking for the mirror image.
And Taiwan has not appeared in any readout I have seen. That fits a political question being walled off. It also fits ordinary caution before a summit.
Before Sunday’s meeting, Bessent put out a statement saying the United States is open to discussions on “avoiding shared risks and avoiding bifurcation of our two systems.” Afterward, he announced a proposed “U.S.-China AI Dialogue,” with a notification mechanism for AI incidents that threaten national security. He talked about transparency between “the number one and the number two AI powers in the world.”
The press keeps telling you this is an AI race. People in a race do not set up notification channels with each other, and they do not describe splitting into two systems as a risk they share. Pax Silica was designed to do precisely that — divide the world into an American stack and everybody else. If the Treasury Secretary now calls that division a danger, then Pax Silica was never the goal. It was the leverage. I said as much in May. One sentence does not prove it, but Thursday might.
Watch the chips first. If Beijing approves the blocked purchases, or Washington drops the 25 percent fee or loosens its rules, the deal I described in May is done.
Watch whether the Board of Trade gets a charter, named members or any role for Congress, or simply keeps operating without them. Watch whether the AI Dialogue becomes a permanent body, and notice that it is the Treasury Secretary running it, not anyone from defense. Watch for what is not said about Taiwan.
And watch the dinner. The White House reportedly turned down a formal CEO roundtable, which means the executives of both nations will do their talking over dessert, off the record.
If China takes the chips, my May conclusion stands, four months late. If it refuses again, we are looking at something I did not anticipate: one doctrine running on two separate infrastructures, coordinated at the top by boards and dialogues. Whatever comes out of that room, I will report it to you next week.
What I can tell you today is that it has never mattered which flag flies outside the building. The men inside believe the same thing: that the world is an engineering problem, and that they are the engineers.
Nobody in either country has been asked to vote on any of it, and nobody will be.
The post Chairman Xi Comes To Washington appeared first on Activist Post.
par Paolo Hamidouche. Les armes déployées par les États-Unis sont très probablement des brouilleurs. Nous vous expliquons en quoi consiste la guerre spatiale aujourd'hui.
The post Des armes américaines dans l’espace ? Dix-sept petits satellites en orbite, véritables sentinelles first appeared on Réseau International.
Mizane Info. Pour la première fois, l’intelligence artificielle et la vision par ordinateur seront utilisées à grande échelle pour reconstituer l’histoire de cet ensemble exceptionnel.
The post Coran : l’intelligence artificielle va étudier 929 manuscrits anciens de Sanaa first appeared on Réseau International.
Une décision importante pour l’industrie automobile vient d’être prise à Washington. Elle mérite d’être observée attentivement depuis l’Union européenne parce qu’elle révèle deux conceptions de plus en plus différentes de la politique industrielle.
Le 28 septembre, l’administration Trump a fortement assoupli les normes fédérales de consommation des voitures. Pour 2031, l’exigence moyenne passe de 50,4 miles par gallon sous Biden à 34,9, soit 31 % de moins. En unités européennes, la consommation moyenne admise passe d’environ 4,7 à 6,7 𝓁/100 km. La différence est considérable. Mais l’essentiel est ailleurs : Washington veut rendre aux constructeurs une liberté technologique que la réglementation précédente avait fortement réduite.
Les normes CAFE (Corporate Average Fuel Economy) sont anciennes, mais l’administration Biden en avait fortement accéléré la progression au nom de la lutte contre les émissions de CO₂ et de l’électrification.
Officiellement, elles n’obligeaient pas à produire des voitures électriques. Mais plus la moyenne exigée devenait sévère, plus les constructeurs devaient en intégrer dans leurs gammes pour la respecter. C’est ce mandat indirect en faveur de l’électrique que l’administration Trump démonte pièce par pièce depuis 2025.
La nouvelle règle laisse davantage de latitude aux constructeurs pour choisir leurs technologies et aux consommateurs pour choisir leurs véhicules. Elle s’attaque aussi au commerce des crédits CAFE, que les constructeurs en déficit achetaient aux excédentaires, au premier rang desquels Tesla : les échanges entre constructeurs sont supprimés pour les crédits acquis à partir de l’année-modèle 2028, ceux engrangés jusqu’en 2027 restant négociables cinq ans. Voilà qui répond à ceux qui accusaient Donald Trump de favoriser Elon Musk : sa politique démantèle une rente réglementaire dont Tesla a largement profité.
Selon le Department of Transportation, la nouvelle règle réduit de 60,6 milliards de dollars les coûts technologiques de l’industrie sur les années-modèles 2027 à 2031, dont 15,3 milliards pour la seule année 2031 — soit environ 1 290 dollars par véhicule. General Motors à elle seule économiserait 20,4 milliards. Les adversaires de la réforme soulignent que ces véhicules coûteront plus cher en carburant. L’administration l’admet. Mais c’est là que se situe la différence philosophique : qui doit arbitrer entre prix d’achat, consommation et technologie ? La réponse de Washington est claire : davantage le constructeur et le consommateur, moins le régulateur.
Un Américain reste libre d’acheter une voiture électrique, et un constructeur d’y investir. Mais le pouvoir politique renonce à forcer l’ensemble du marché à suivre une trajectoire fixée par l’administration.
L’UE suit le chemin inverse. Autour de l’industrie automobile s’est constituée une véritable monstruosité bureaucratique : normes de CO₂, règles sur les carburants et les biocarburants, obligations d’énergies renouvelables, infrastructures de recharge, ETS2, batteries, taxonomie, rapports détaillés…
Il serait trop facile d’en rejeter toute la responsabilité sur les institutions européennes. Une marée de lobbyistes des constructeurs — comme de presque toutes les branches industrielles — joue le jeu des amendements au Parlement européen. Chacun cherche un seuil plus favorable, une exception ou un délai, puis rentre fièrement au siège annoncer au CEO la victoire obtenue.
À force d’améliorer son sort à l’intérieur du système, on oublie de regarder le système lui-même. Ainsi naissent ces législations remplies de seuils, de coefficients, d’exceptions et de chiffres abscons. Admirez la précision : réduire de 14,5 % l’intensité en gaz à effet de serre des carburants d’ici 2030 ! Chaque intérêt particulier a remporté sa petite victoire, tandis que l’industrie, collectivement, perdait la bataille de la compétitivité.
L’interdiction des voitures thermiques en 2035 en est l’exemple le plus connu. Même ramenée à 90 % de réduction, comme le propose désormais la Commission, elle reste une machine qui broie l’industrie automobile européenne et réjouit les Chinois, qui envahissent notre marché.
La comparaison est d’autant plus frappante que les industriels américains ont publiquement demandé cet assouplissement.
En Europe, les constructeurs ont trop longtemps accompagné les ambitions politiques, annoncé des électrifications accélérées et plié l’échine devant une bureaucratie qui prétendait organiser leur avenir. Plutôt que de contester le principe, ils ont négocié les détails.
Aujourd’hui, les difficultés ne peuvent plus être cachées : restructurations, suppressions d’emplois, fournisseurs fragilisés, investissements remis en question, concurrence chinoise toujours plus forte. Ce n’est que maintenant que les appels à la flexibilité deviennent audibles. Il aurait fallu parler beaucoup plus tôt.
La décision américaine pose une question essentielle à l’UE : jusqu’où le pouvoir politique doit-il décider à la place de l’industrie ? Face aux États-Unis qui desserrent la contrainte et à la Chine qui défend méthodiquement sa puissance industrielle, l’UE ne peut continuer à transformer ses entreprises en exécutantes de trajectoires technologiques décidées politiquement.
Pour retrouver sa compétitivité, l’industrie européenne a besoin d’une énergie abondante, fiable et abordable.
Elle a surtout besoin qu’on lui rende sa liberté.
Dernier livre « La vérité sur les COP. Trente ans d’illusions » | L’artilleur
L’article Washington rend sa liberté à l’industrie automobile. Si l’UE s’en inspirait ? est apparu en premier sur Contrepoints.
SpaceX has unveiled a plan to construct a $100 billion spaceport, according to a MarketWatch report published in March 2025. The proposal, which was shared with investors in a private presentation, outlines a facility designed to support the company’s next-generation Starship launch system. The location of the spaceport has not been disclosed, though company officials said construction is slated to begin within two years.
Investors who attended the presentation raised immediate concerns regarding the lack of a detailed funding breakdown, according to the report. The scale of the proposed expenditure has drawn comparisons to the annual budgets of major government space agencies, which are typically measured in tens of billions of dollars. The company has not issued a public statement confirming the report’s figures, and representatives did not respond to requests for comment.
The proposed facility is designed to support Starship launches, including crewed missions to Mars and lunar operations, according to company materials cited in the report. Plans include multiple launch pads, a dedicated manufacturing facility, and an on-site power plant, according to internal documents shared with investors. SpaceX has previously stated that Starship is central to its long-term ambitions, with the vehicle already contracted for lunar missions under NASA’s Artemis program [2].
SpaceX representatives said the first phase of construction would require $25 billion in initial capital, with the remainder of the $100 billion spread over a decade, according to the report. The timeline for the project has not been finalized, and officials said the company is still conducting site surveys and preliminary engineering work. A competing aerospace executive, who spoke on condition of anonymity, noted that the plan could strain supply chains and talent pools across the industry, though the executive also acknowledged SpaceX’s history of aggressive timelines.
In a private call with bondholders, several investors questioned whether SpaceX can fund the project without issuing new equity, according to the MarketWatch report. One analyst on the call noted that SpaceX’s current cash flow from Starlink and launch services is insufficient to cover the projected annual spending of $10 billion. The analyst’s assessment was based on financial documents shared during the presentation, which showed current operating revenue trailing the projected capital requirements.
A shareholder quoted in the report said, “We haven’t seen a credible plan for where the money comes from.” The quote reflects a broader sentiment among institutional investors who have grown accustomed to SpaceX’s rapid capital raises but are now questioning the sustainability of its spending. The company has not responded publicly to the concerns raised during the call.
SpaceX’s latest funding round valued the company at $350 billion, according to financial disclosures cited in the report. However, much of that valuation is tied to equity rather than liquid reserves, meaning the company would need to raise additional capital or redirect existing revenue to fund the spaceport. The company has historically relied on government contracts with NASA and the Department of Defense, which typically cover specific missions rather than infrastructure development, officials said.
Some industry analysts have suggested that SpaceX might pursue a special-purpose vehicle or seek investment from sovereign wealth funds to finance the project, but no such agreements have been announced. The interest from sovereign funds is not without precedent; UAE’s Mubadala already owns five percent of Starlink, according to commentary from Aaron Day [3]. The involvement of state-linked investors has drawn scrutiny from some stakeholders, who question the long-term implications of foreign capital in critical space infrastructure.
The scale of the proposal exceeds the combined annual budgets of major space agencies, according to industry estimates cited in the report. The construction of a dedicated power plant and manufacturing facility would represent a significant expansion of SpaceX’s industrial footprint, potentially reshaping the competitive landscape for commercial spaceflight. The company’s launch record has already set it apart from European competitors, whose recent attempts have faced setbacks [1].
The report noted that regulatory approvals and environmental reviews have not yet been filed for the proposed spaceport, adding further uncertainty to the funding outlook. The absence of filings means the project has not been subject to public scrutiny or regulatory oversight, which could delay the timeline if challenges arise. MarketWatch reported that the company has not committed to a specific date for submitting the required documentation.
SpaceX announced an investor briefing scheduled for next month to outline the financial details of the proposal, according to the company. The briefing is expected to address questions about funding sources, project milestones, and the regulatory path forward. The board has not yet formally approved the budget, and internal disagreements have been reported, according to sources familiar with the matter who spoke to MarketWatch.
The outcome will depend on SpaceX’s ability to secure committed financing, officials said, with no definitive timeline for a decision. The company’s track record of delivering on ambitious projects suggests the proposal will not be withdrawn lightly, but the sheer scale of the $100 billion figure has introduced a level of scrutiny not seen in previous capital campaigns. Stakeholders on all sides are awaiting the investor briefing as the next milestone in the process.
The post SpaceX’s $100 Billion Spaceport Plan Raises Questions on Funding appeared first on Activist Post.
Javier Milei’s administration has been showcasing Patagonia to Silicon Valley executives, culminating in OpenAI’s ‘Stargate Argentina’ announcement. The focus of this piece, Emiliano Kargieman, is the man actually developing OpenAI’s Argentine project and boasts significant ties to United States’ military and intelligence, NASA, Palantir and the Silicon Valley stalwarts who helped privatize space.
Patagonia, positioned at the tip of the South American continent, has long been valued for its dramatic landscape, its pristine waters and its unique wildlife. However, for some, the value of Patagonia’s natural treasures has shifted. No longer are its majestic glaciers and temperate rainforests to be protected. Instead, it is now being proposed that Patagonia, with its brisk climate and abundant water resources, become home to a new animal –– the data center.
The proliferation of data centers has become a huge political issue in recent months, particularly in the United States, largely due to concerns about their gargantuan appetite for local resources –– driving up utility prices for everyday consumers and endangering communities where some of those resources were already scarce. Many have been outraged about how their local governments, elected to ostensibly protect their interests, have rolled over for data center developers, promising jobs that rarely manifest and –– if they do –– largely evaporate after the center is built. Meanwhile the technology many of these data centers power, Artificial Intelligence (AI), has also sparked its own set of concerns, nominally focused on AI’s usurpation of human jobs and human creativity while generating generous returns for the ultra-wealthy and dubious benefits for the rest of us.
The predation of Patagonia by data center developers has been greatly facilitated by the policies of Argentina’s president Javier Milei. Milei has made it a point to partner with Silicon Valley, as well as the current U.S. administration –– itself colonized by Silicon Valley to an impressive and unprecedented degree. He has also developed generous tax benefits schemes for foreign U.S. Big Tech firms to operate in the country, while offering those that seek to park their data centers in the country next to no regulation regarding their construction or operation. Earlier this year, Milei wrote an editorial in the Financial Times where he “invite[d] AI to free itself” and proposed allowing AI-run companies with no humans involved to operate in his country. While some pointed out that these statements seemed to value AI activity over human activity, this is not an isolated incident. For instance, one adviser in Milei’s inner circle, Demián Riedel, has controversially argued that humans are, in fact, Argentina’s biggest problem, particularly in Argentine Patagonia, because they present an obstacle to AI infrastructure and development in the region.
Riedel was notably involved in brokering a major deal for what will become a truly massive data center in his country: Stargate Argentina. Stargate Argentina is the newest constellation of OpenAI’s Stargate project, which was initially launched in January 2025 “to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States.” A few months later, in May 2025, OpenAI announced with little fanfare that it planned to internationalize its Stargate project via its new “OpenAI for countries” initiative. Not long thereafter, and with the direct involvement of Milei and his top officials, Stargate Argentina was born.
While OpenAI is now a well-known name, virtually unknown was their partner in Stargate Argentina –– a company called Sur Energy. Not only was Sur Energy unknown internationally, it was also unknown within Argentina, despite being billed by OpenAI’s Sam Altman as one of Argentina’s “biggest energy companies.” Lurking behind this company, which appears to have no real assets in the country aside from letters of intent, is a man who may be slightly more familiar, at least in Argentina.
Emiliano Kargieman is the face of Sur Energy, but it is only his latest venture in a series of companies that have sought to complement the growing influence of Silicon Valley corporations over the public sector. As will be explored in this article, Kargieman’s previous ventures, particularly the satellite surveillance firm Satellogic, not only boasts close ties to current and former top figures in the Trump administration, but has become an increasingly important part of the AI surveillance grid that the U.S. intelligence contractor Palantir has been amassing. In particular, the company is due to enable –– within the next year –– what was once Palantir’s most implausible ambition –– predictive policing of the entire planet in real time via AI-powered satellite surveillance, i.e. total information awareness.
However, more recently through his efforts alongside Sam Altman to create Stargate Argentina, Kargieman has revealed some hidden truths, not just about the data center he hopes to build, but about the data center build-out at large. As will be detailed in this article, Kargieman has casually admitted that Stargate Argentina is not about bringing jobs to locals (despite what some Argentina politicians have claimed), but about generating dollars. Perhaps more revealing of all, he revealed that it is likely that his partner in Stargate Argentina, OpenAI, will not be ultimate end-user of this massive data center, suggesting that the prominent AI firm will collapse under the inevitable popping of an AI debt bubble. In stating this he notes that, whether OpenAI or not, someone will be using this data center. If so, who will use it and what for? Based on Kargieman’s statements and the available evidence, we have attempted to answer those questions, revealing increasingly likely possibilities that will have implications far beyond Argentina.
Raised in the Palermo district of Buenos Aires, Emiliano Kargieman’s childhood was characterized by his early obsession with computers and a rebellious streak. For instance, he learned how to repair computers at age 10 and was also kicked out of school for filling the principal’s office with toilet paper. At 15, he claims to have created his first software company. By this time, as a teenager, he had also become part of a hacking collective called “Hacked by Owls”, or HBO. According to Matt Travizano, Kargieman’s long-time friend (and later his co-founder at Sur Energy), some of the HBO members used their technological prowess to spy on the conversations and communications of Argentine Congressmen. Though HBO seemed adversarial to State interests, with its members claiming to espouse “anarchist” ideals, many of its members –– including Kargieman –– simultaneously worked for the Argentine government.
For instance, while Kargieman has declined to disclose this in media interviews, one of his old CVs lists him as working for the Argentine government’s Special Information Security Research Group from age 14 to age 17, overlapping with his time at the HBO collective. This group was part of the DGI (Dirección General Impositiva), Argentina’s General Tax Directorate. This is notable given that, while working for this State agency, Kargieman’s hacking collective was also allegedly spying on elected officials.
In 1997, when Kargieman was 20, he and the broader HBO collective began working directly for Argentina’s AFIP (Administración Federal de Ingresos Públicos; in English, Federal Public Income Administration). That year, the DGI, which had employed Kargieman as a teenager via the Special Information Security Research Group, was placed under the AFIP’s authority. Shortly before then, in 1996, Kargieman had created a cybersecurity company that would later become known as Core Security Technologies. He did so with five of his friends, including Gerardo Richarte, Lucio Torre, and Jonathan Altzszul, who had previously worked at the Special Projects Office at Argentina’s AFIP.
Richarte, Torre, and some long-time employees of Core Security, such as Norberto Kueffner, later followed Kargieman to Satellogic, the satellite surveillance company he would later create. Notably, Kueffner is now Director of IT for Javier Milei’s administration. Another veteran of Core Security, Ariel “Wata” Waissbein, is now the head of Argentina’s Federal Cybersecurity Agency (AFC). Yet another Core Security veteran who also went to work for Satellogic was Aviv Cohen, a former Israeli intelligence officer who was, for a time, Satellogic’s head of “special projects.”
In 1998, a year after becoming entangled with Argentina’s AFIP, Kargieman’s firm Core Security began focusing expressly on the U.S. Market. Shortly thereafter, it attracted significant attention from U.S.-based businesses, such as Apple and Lockheed Martin, and the public sector. Core Security’s main U.S. public sector clients soon included the Department of Homeland Security (DHS), the U.S. military’s DARPA, NASA and the NSA.
The same year, in 1998, Core Security was named an “Endeavor Entrepreneur” by the Endeavor Foundation –– a non-profit organization founded the year prior by two Americans, Linda Rottenberg and Peter Kellner. While much can be said about Endeavor, which features prominently in a previous investigation from Unlimited Hangout, their affinity for a small Argentine software company should be of little surprise given the organization’s intimate relations to the South American country. For starters, Rottenberg and Kellner met while in Argentina. Long before the likes of Edgar Bronfman Jr., Reid Hoffman, and Pierre Omidyar had donated vast sums of money to the entrepreneurial evangelicals, Endeavor’s first success story was Marcos Galperin’s MercadoLibre. Galperin became Argentina’s richest man –– on paper, anyway –– due to his firm’s success, before joining Endeavor’s board when the once-non-profit added a new venture capital component to their organization. Endeavor, now advised by Galperin, eventually funded Kargieman’s later company Satellogic, and Galperin now serves on Satellogic’s board.
By 2001, Kargieman’s Core Security also gained the attention of Pegasus Capital. As noted by iProfessional, Pegasus –– beginning in February 2001 –– began “collaborating” with Core in the development of Core’s business model, growth strategy, organizational design and much more. Pegasus Capital was created by Craig Cogut, a key figure in the “junk bond” financial scandal at the now defunct bank Drexel Burnham Lambert. Drexel’s junk bond department, led by Michael Milken, engaged in blatantly illegal activity and used junk bonds to help fuel the takeovers of major corporations by the era’s infamous “corporate raiders” prior to the bank’s collapse. More specifically, Cogut was the lawyer who advised the Milken-run and scandal-ridden junk bond department on the legality of its transactions, including those that saw Milken become a convicted felon. Following Drexel’s collapse, Cogut teamed up with a group of Drexel alumni led by Leon Black –– now best known for his close association with the deceased sex trafficker and Black’s “financial adviser” Jeffrey Epstein –– to co-found Apollo Advisers (now Apollo Global Management) in 1990. Cogut left Apollo to found Pegasus in 1996. Pegasus has since became a key player in several UN-supported “green” finance initiatives, including some which have directly involved Kargieman’s other company Satellogic.
In 2006, Kargieman created Aconcagua Ventures as a joint venture with Pegasus Capital. This firm ostensibly “focused on investing in high-tech startups in Latin America to develop them as global businesses.” By 2009, it had only invested in Core Security as well as the company Popego, where Kargieman served as both CEO and director in 2007. Popego was described around this time as an artificial intelligence (AI) software company.
Popego, as well as Aconcagua Ventures, backed a non-profit organization that launched in 2009 with the goal of creating a virtual currency related to one’s online reputation. That organization was called The Whuffie Bank and Emiliano Kargieman is listed as one of its founders. The Whuffie Bank, which launched at a TechCrunch conference in 2009, aimed “to promote change in the web by rewarding users with a positive impact on the web with this karma-like digital currency.” This “non-profit” planned to closely surveil social media users by “monitor[ing] [their] activity across various websites, including things like comments, posts, and more.”
As TechCrunch further explained:
“When you complete positive actions, you gain Whuffies, and you lose them when you do something that the organization deems to be detrimental. [. . . ] The algorithm takes into account ‘public endorsements’, or the number of times a user’s tweets are retweeted, or a Facebook post is Liked. It also takes into account who is making the endorsement, and the content in the messages that are being posted.”
The Whuffie Bank website noted that the goal of the organization was to drive “social change” through its virtual currency, while another Whuffie Bank co-founder, Santiago Siri, has described its goal as “rethink[ing] the future of money.” Siri, while working at the Whuffie Bank, was named a World Economic Forum Global Shaper for his efforts to advance digital, reputation-based money that rewards or punishes users based on their online activity and speech. The Whuffie Bank represents an early effort to link digital currencies to the digital behavior and reputation of its users, efforts which continue today in other forms and have raised concerns about ascendant “social credit systems.”
While the Whuffies went extinct, Kargieman’s next venture Satellogic quickly placed him shoulder to shoulder alongside the very pioneers responsible for digitizing the U.S. financial system, e.g. former U.S. Treasury Secretary Steve Mnuchin and current Commerce Secretary Howard Lutnick (previously the CEO of Cantor Fitzgerald, the largest dealer of government securities in the country). Before his low Earth observation satellite firm took off, former NASA contractor Kargieman found himself in Northern California, at the very Air Force base where Silicon Valley unofficially took hold of the country’s space program –– taking what was once the pride and joy of the US government firmly into the private sector for good.
According to the earliest iterations of its website, Satellogic was founded in 2010 by “two graduates of the NASA Ames-based Singularity University’s Graduate Studies Program” in order to build “shared low earth orbit infrastructure to open real-time satellite data to a new generation of application developers.” Emiliano Kargieman and his co-founder, Andrew Fursman, came up with the “concepts that would become Satellogic” while enrolled in that graduate student program. The program took place at Northern California’s Moffet Federal Airfield –– one of the most prominent locales for NASA’s extensive collaboration with the private sector.
The airport was formerly a US Naval facility. Right up until the year Satellogic was founded, it had housed the Air Force’s 21st Space Operations Squadron. Today, the airport is served by the California Air National Guard, and notably the 7th Psychological Operations Group of the US Army Reserve –– an influential support group for military actions which “involve[s] sharing specific information to foreign audiences to influence the emotions, motives, reasoning, and behavior of governments and citizens” including “cyber warfare and advanced communication techniques across all forms of media.”
NASA’s Ames Center is also known for containing one of the largest free standing buildings in the world –– Hangar One. This building, built in 1933, is over 8 acres large and is known for having fog and other weather-type patterns within its voluminous confines. In addition, it became home to the private jets of past and present Google executives Larry Page, Sergey Brin and Eric Schmidt after a $33 million renovation plan Google offered to NASA in 2011. NASA, the federal Environmental Protection Agency (EPA) and the California Regional Water Quality Control Board eventually came to an agreement in November 2014 which saw Google’s subsidiary Planetary Ventures LLC sign a 60-year lease worth $1.16 billion in rent, which included plans to significantly remodel the famous hangar.
Notably, the first oldest archive version of Satellogic’s website features an image of Hangar One, complete with a five-point star adorning a large bay door and another on top of the building. This is quite likely due to the fact that back in 2009, NASA, Google, and the Ames Center staff collaborated on the founding of the very incubator program that germinated the core ideas of Satellogic.
Unlimited Hangout’s first investigation into Satellogic –– April 2024’s Debt From Above: The Carbon Credit Coup –– focused mainly on the firm’s public-private partnerships related to their enforcement of a blockchain-enabled carbon market scheme via satellite surveillance. To spare repeating information regarding Kargieman’s company that was previously addressed, this article will instead focus on placing Satellogic within the context of the decades-long transmutation of America’s space program from the public sector’s federal budget into the private sector powerhouse. From the evidence, it appears that the main figures and sites responsible for NASA partnerships with the Tech industry, not only produced Elon Musk’s SpaceX –– Satellogic’s “preferred vendor for ride share missions” –– but also Satellogic itself. It seems that Satellogic was essentially created to aid both the privatized U.S. space and mass surveillance programs.
In September 2008, author and futurist Ray Kurzweil teamed up with space entrepreneur and X PRIZE Foundation chairman Peter Diamandis to found Singularity University (SU). Their decision to partner came after Diamandis read Kurzweil’s 2005 book The Singularity is Near, which argues that machine intelligence will soon surpass human intelligence and that humans and machines should, at that point, merge into a transhumanist collective. Diamandis approached Kurzweil after reading and suggested they partner on what would become Singularity University.
The goal of SU is to educate the “leaders of tomorrow” in emerging technologies that Kurzweil, and presumably Diamandis, believe will allow humanity to accelerate the arrival of this “singularity” event and thus a transhumanist future for humanity. Singularity University’s website boasts that it has consequently developed “a global ecosystem of techno-optimists.” This term appears to be a reference to venture capitalist (and current Pentagon advisor) Marc Andreessen’s “Techno-Optimist Manifesto,” in which Andreessen argues for a full-throated embrace of “transformative” technologies like AI on an accelerated timeline, with complete disregard for whom or for what it could create negative consequences. Kurzweil is notably a “patron saint” of techno-optimism and Andreessen’s manifesto more or less portrays the ideology that Singularity University seeks to instill in its “techno-optimist” graduates.
The “Techno-Optimist” ideology overlaps considerably with Technocracy. This overlap is detailed extensively by Iain Davis in his book The Technocratic Dark State. Davis, in his book, defines Technocracy as a form of governance that claims to maximize efficiency by “scientifically” managing and socially engineering the human population using autocratic sociopolitical and technological behavioral control. In a technocratic system, the management and social engineering of the “human animal” is to be performed by unelected “experts” in science, engineering or technology, also known as the technocrats. This form of governance has been heavily adopted by the wealthiest men in Silicon Valley, including Andreessen as well as others like Elon Musk and Peter Thiel.
A few months after Singularity University was created, the newly-formed university found a home at NASA’s Ames Center after securing “Corporate Founding” funding from Google and ePlanet Ventures. While Google needs little introduction, ePlanet Ventures is a venture firm associated with Tim Draper’s DFJ (Draper Fisher Jurvetson), whose investment committee once featured all three of DFJ’s partners. One of those partners, Steve Jurvetson –– much like SU founder Diamandis –– is a long-time friend of SpaceX’s Elon Musk, with positions on both the Tesla and SpaceX board. He is also credited with bailing out Musk in 2008 with large investments in both firms. However, Jurvetson hardly deserves singular responsibility for SpaceX’s survival. Two other men who were involved in that rescue also happened to play a prominent role in the founding of SU, where Satellogic was born.
Then-Director of the Ames Center, Simon “Pete” Worden, spoke upon SU’s founding, stating that “The NASA Ames campus has a proud history of supporting ground-breaking innovation, and Singularity University fits into that tradition.” Worden had been selected to run Ames by his long-time friend, Dr. Michael Griffin, who became the Administrator of NASA in 2005. Both Worden and Griffin were alumni of the controversial Strategic Defense Initiative Office (SDIO) –– colloquially known as “Star Wars” due to its early 1980s formalization under President Ronald Reagan –– with Worden having taken over Griffin’s job as Head of Technology in the space-focused missile defense program.
While at SDIO, both Griffin and Worden were responsible for immense strides taken in fields quite relevant to today’s commercialized space industry. For starters, their work helped pioneer the development of Low Earth Observation (LEO) satellite constellations such as those later launched by Satellogic. Their advancement of satellite broadcasting capabilities in general was accomplished via SDIO’s Clementine project –– a “small, low-cost, and rapidly developed satellite used to test sensor and propulsion technology for missile interceptors” and which also “mapped the Moon.” Additionally, Griffin was working at the Virginia-based Orbital Sciences after his time at SDIO, a company which pioneered the use of “a system for collecting data from remote locations using satellites in low-Earth orbit.” Worden is considered a “crucial innovator and promoter of small satellites in the early years of development” while at SDIO.
Worden and Griffin were also both instrumental to the commercially-viable build out of the reusable, vertical take-off and landing of the “DC-X” rocket, which set the stage for SpaceX’s commercially-viable building of reusable, vertical take-off and landing Falcon 1 rockets. In fact, Musk himself told former manager of SDIO Jess Sponable in 2013 that SpaceX was “just continuing the great work of the DC-X project!” While many details remain outside of the scope of this article, Griffin’s stint running a budget-starved NASA while Worden merged the space program with Silicon Valley at Ames led to accusations from Ames employees that the two men were “part of a conspiracy to destroy America’s space program.”
There are few people more responsible for the private sector takeover of the space industry than Griffin, Worden, and Diamandis –– all of whom have their unique ties to Elon Musk and the founding (and eventual success) of SpaceX.
In August 1993, the Falcon 1 predecessor “DC-X” took its first lift off in White Sands, New Mexico during testing by contractor McDonnell Douglas on behalf of the SDIO. Griffin, the Deputy Director of Technology for SDIO at the time, stated that the “DC-X was government R&D at its finest.” The DC-X took its last flight in 1995 before being upgraded to the DC-XA, which was promptly passed back within NASA proper, only to be shuttered after a final flight on July 31, 1996.
In May 1996, the month prior to the shutdown of the DC-X project, Diamandis officially founded the X Prize. The prize offered $10,000,000 to the team that could successfully breach the lower atmosphere. While the winning team would not do so until 2004, the prize kick-started the private, commercial space industry, and was influential in addressing new forms of fundraising and project management for NASA. The original X Prize Foundation Committee, announced in December 1996, featured Worden in addition to astronaut Buzz Aldrin, among others. Due to the nature of the prize being first announced in 1996, the many teams competing had to spend years building their rockets and technology, causing nearly a decade of waiting for the new space industry to blast off.
During this interim period, Diamandis got a call from Bill Gross, the founder of the early tech startup incubator Idealab, in 2000 to come aboard a new company as CEO. That new company, called BlastOff!, was given a $1 billion budget to organize a private moon mission. In a 2008 video for the Google Lunar X Prize, which officially brought together the Griffin-led NASA with SU’s funder Google and Diamandis’ X Prize, Diamandis describes the team at BlastOff! as having eventually decided to use the Russian ICBM “Dnepr” to launch their moon mission. BlastOff!, which operated in stealth, eventually shuttered in January 2001.
Only a few months prior, in October 2000, Musk was replaced by Peter Thiel as CEO of PayPal/X.com, with PayPal also having come out of the Idealab incubator (as Confinity) with an early investment from Gross and Idealab in January 2000. Furthermore, the first board member of PayPal was Idealab VP Scott Bannister, who graduated college with PayPal founder Max Levchin, and who was responsible for introducing Levchin to Peter Thiel. As previous investigations from Unlimited Hangout have aimed to prove, Idealab –– despite being little known today –– is intimately connected to the primordial origins and subsequent immense success of Google, PayPal, Palantir, Tether, PolyMarket and arguably Musk’s SpaceX.
Out of a job and about to be flush with cash from the sale of PayPal to eBay, Musk set his sights on the stars, Mars, and beyond. The year after he was ousted from PayPal, Musk met Griffin at The Mars Society in 2001, which kicked off their lengthy friendship. The Mars Society was previously formed in 1998 and boasted Griffin as a founding Steering Committee member. X Prize Foundation Committee member Aldrin, and a plethora of employees from NASA’s Ames Center soon joined him. Musk made a large donation and subsequently joined the Steering Committee alongside Griffin. Musk has since overtly promoted the idea of establishing a technocracy on Mars. Affinity for technocracy seems to run in Musk’s family, as his maternal grandfather previously led the technocratic movement in Canada.
In 2001 and 2002, Musk and Griffin made multiple flights to Russia in an effort to purchase Dnepr rockets in order to begin experiments in space travel, arguably continuing the work of the Idealab-sponsored BlastOff! On one of their trips to Russia, after finding the purchasing price of the Russian ICBMs too high, Musk typed out a model to manufacture the rockets needed, but at low cost. He reportedly came up with the plans for SpaceX on the flight back alongside Griffin. Griffin was offered a role as Chief Engineer at SpaceX, but decided instead to become the President and COO of the CIA’s investment firm, In-Q-Tel. Musk allegedly named his son Griffin Musk after his friend, and eventually gave Griffin an advisory role at SpaceX-subsidiary Castelion in 2022. Castelion boasts mass-production capabilities of hypersonic missiles, and additionally features fellow In-Q-Tel-alum Lisa Porter on the advisory board.
Griffin left In-Q-Tel to take the reigns at NASA in April 2005 after being selected by President George W. Bush. Only a week after his appointment, Griffin met with Musk at NASA Headquarters. Two key changes to NASA’s status quo occurred at the onset of Griffin’s term –– massive budget cuts to the likes of the Ames Research Center in November 2005, and the authorization of the “Prize Authority.” This novel fundraising and budget concept had been pioneered by Diamandis’ X Prize. Only a few years later, Google, NASA and X Prize collaborated together on the aforementioned Google Lunar X Prize in 2008. Griffin was quoted as being “an advocate for the use of Prize Authority we’ve had at NASA since 2005.”
Griffin used his position at NASA to further advocate for the commercial and private space industry, and when Musk’s SpaceX nearly folded after three failed rocket launches, Griffin and his friend Worden came to Musk’s aid. In April 2006, after the explosion of a Falcon 1 on its maiden voyage, the Pentagon selected Worden to lead the inquiry into the mishap, as DARPA had helped fund the launches. The very next month, in May 2006, Worden was selected by Griffin to lead the ever-important Ames Research Center. Two years later, in December 2008, the Griffin-led NASA formally partnered with X Prize and Google, and gave $3.5 billion in contracts to SpaceX and Griffin’s former company Orbital Sciences –– saving Musk’s then-troubled company in addition to the aforementioned private bailout from Jurvetson.
The next month, January 2009, Singularity University was announced by X Prize’s Diamandis with support from the Griffin-led NASA and with funding from Google and the Jurvetson-affiliated ePlanet Ventures –– stationed at the Worden-led Ames Research Center. Early reports of the announcement even listed Worden as an official founding member of SU, although he never appears in later iterations of their founding story. When Worden finally announced his 2015 departure from the NASA Research Center, he specifically referred to “the nine years he spent at Ames” as “the most enjoyable of his 40 years of public service.” Worden furthered:
“In the last nine years we have launched dozens of small, low-cost satellites – and helped ignite a major new industry in this area… Ames has provided entry technology for the emerging commercial space launch sector. We have helped launch small satellites working with a number of nations. And we’ve hosted and inspired thousands of students.”
It is hard to separate Worden’s musings on his time at Ames from Kargieman’s decision to found of Satellogic. Not only did Kargieman become interested in the commercial space launch sector and “small, low-cost satellites” while studying at the Worden-linked, “techno-optimist” graduate program, but the same Ames facility where that program was hosted also apparently became Satellogic’s first home. Thus, more than just sparking the idea for Satellogic, Ames seems to have incubated it as a company. Furthermore, it remains difficult to parse the beginnings of Satellogic from its connections to major players in the privatization of the military and space programs. Many of these companies have been discussed in prior investigations from Unlimited Hangout –– including SpaceX, Tether, and Palantir –– not to mention their willing public partners within the US government. Notably all three of these companies have since become directly and significantly partnered with Satellogic.
In addition, one simply has to look at Satellogic’s board members, both past and present, to understand that Kargieman’s company is hardly a simple tech startup, and operates far more like a cog in a well-designed, multi-headed apparatus developed by the U.S. government but profiteering as a private company.
As previously stated, much was written about the founding and funding of Satellogic in Debt From Above, but the more recent developments deserve reminding for the context of this piece. For starters, in 2022, the current Commerce Secretary Howard Lutnick, who boasts intimate ties to stablecoin issuer (and eventual Satellogic investor) Tether, helped connect Kargieman’s company with former Treasury Secretary Steve Mnuchin, who then funded Satellogic via his firm Liberty Strategic Capital. Both Lutnick and Mnuchin joined the Satellogic board that same year. Another board member who also joined that year has direct connections to the Pentagon –– Liberty Strategic Capital’s Senior Managing Director General Joseph Dunford (Ret.), who was the former Chairman of the Joint Chiefs of Staff under the first Trump administration. In addition, Lutnick’s Cantor Fitzgerald notably helped take Satellogic public, with the ticker $SATL.
In February 2022, Satellogic also officially partnered with Palantir, the privatized version of the Pentagon’s controversial Total Information Awareness program. Satellogic claimed in a press release that its partnership with Palantir was intended to “leverage Palantir’s Foundry platform” in order to “accelerat[e] business process, rapid image product delivery, AI model training, and enterprise-wide data integration.” The partnership also saw Satellogic become a major provider of space-based surveillance data to the U.S. government via Palantir, which contracts with all 18 U.S. intelligence agencies, the military and many other U.S. federal agencies. According to Satellogic’s press release, as part of the deal, “Satellogic will provide Palantir’s U.S. government customers with ready access to Satellogic’s high-resolution satellite imagery to drive analytical insights across a range of mission-oriented use cases.”
A few months after announcing their partnership, in April 2022, Satellogic and Palantir launched their “first Edge AI-enabled satellite into space” with assistance from SpaceX’s Transporter 4 mission. This trio working together directly lends credibility to the theory described above that the common players behind all three companies were building collaborative infrastructure –– in this case, satellites, AI and rockets. According to their press release, the “new Edge AI-enabled NewSat” is the product of a “six-month project” in which “Satellogic adapted its hosted payload program and edge computing hardware” in order to “run Palantir’s Edge AI platform.” Their “joint project” aims to “deliver[. . .] new capabilities which will have broad applicability across mission sets in both government and the commercial sector.” This technological collaboration allows for quick swapping between different algorithms depending on its location. The press release furthers:
“We are able to hot-swap between models and chain capabilities, depending on mission requirements and terrain. For example, when the satellite passes over a port we can run the cloud detector, image segmentation, and ship detector models, and when passing over a town we can switch to vehicle and building detectors.”
In total, the Satellogic hardware is capable of running 6 additional third party models, including Palantir Omni and Palantir Overcast, in addition to Xailient, Pilot.ai and Mind Foundry. According to Palantir, “the Palantir-Satellogic collaboration represents a significant step in expediting delivery and dramatically improving the quality of data sent back to customers.” Palantir’s blog post on the collaboration concluded that:
“[T]his pioneering application of Edge AI in space aims to serve as the foundation for advanced analytics that can inform timely solutions to address critical global priorities, like combatting climate change, preventing disasters, and improving emergency response.”
One of Satellogic’s more recent announcements relates to the imminent launch of a new constellation of satellites it has named Merlin. Due to launch this October and become active some time in 2027, Merlin will “support persistent monitoring of unlimited locations worldwide” by “enabl[ing the] daily remapping of the entire planet at one-meter resolution.” Presumably, this will also grant its analytic partner, Palantir, the same capability.
In announcing Merlin, Kargieman stated the following:
“Until now, organizations had to choose between global coverage at low resolution or high-resolution monitoring of a limited number of sites. Merlin removes that trade-off and enables persistent monitoring at planetary scale.”
In an interview earlier this year with Forbes Argentina, Kargieman remarked that Merlin will soon “allow us to do things we never imagined. With AI [and Merlin] we can begin to predict the future.” As noted in previous reporting from Unlimited Hangout, a long-time major focus of Palantir is predictive analytics, particularly predictive policing –– often referred to as “pre-crime.”
Merlin exemplifies Satellogic’s efforts to push planetary surveillance into a new paradigm. Kargieman has referred to this new paradigm as “Persistent Global Intelligence,” a departure from the existing Earth Observation paradigm of “episodic imagery.” Its efforts in this regard are meant to empower global intelligence and defense agencies, particularly those of the United States. This is reflected, not only by Satellogic’s existing and new U.S. government contracts, but also through their recent partnerships with U.S. government contractor SynMax (AI surveillance data analytics company backed by Palantir co-founder Joe Lonsdale’s 8VC) and appointing one of the main architects of the U.S. military’s controversial Project Maven (AI-powered weapon targeting software) as a strategic advisor.
While engineering Satellogic’s pivot to AI-powered “Persistent Global Intelligence” satellite surveillance, Kargieman decided to create yet another company –– this one aimed at cashing in on the AI “boom.” He, along with Matt Travizano and Stan Chudovsky, founded Sur Energy in 2025. Travizano was an Argentine who was also the head of Sur Ventures, which invested in Satellogic, and had long-standing business interests in California. He is also credited with connecting the administration of Javier Milei to prominent figures in Silicon Valley. Travizano died last year while scaling California’s Mount Shasta, shortly after founding Sur Energy. Meanwhile, Chudovsky was previously VP of Messaging at Meta/Facebook and, prior to that, VP of Growth and Global Strategy at PayPal.
Shortly after launching, and even before having a website presence, Sur Energy announced a massive partnership with OpenAI, the formerly non-profit but now for-profit AI company run by Sam Altman and the developer of ChatGPT. Altman is a protégé of Peter Thiel, the billionaire co-founder of PayPal and Palantir whose recent move to Argentina and meetings with Milei and top government officials earlier this year made international headlines.
The Sur Energy-OpenAI partnership, formalized in October 2025 via a letter of intent, seeks to develop a $25 billion data center in Argentina’s Patagonia region called Stargate Argentina. Oracle, the CIA-linked data infrastructure company helmed by Larry Ellison, will likely operate Stargate Argentina upon its completion, per Kargieman. Oracle is notably OpenAI’s partner in the U.S.-based Stargate project, which –– per OpenAI –– now has global ambitions. Kargieman has also claimed that construction on Stargate Argentina will begin this year and end in 2027, even though the planned location of the site has not been made public. Notably, Milei’s administration was intimately and directly involved in the discussions that led to the signing of the letter of intent between Sur Energy and OpenAI.
In announcing the project, OpenAI’s Sam Altman referred to Sur Energy as one of Argentina’s “leading energy companies,” which it is not. Upon the announcement, even Bloomberg noted that Sur Energy was entirely unknown in the sector and had exactly zero assets in the country. OpenAI later clarified that Sur Energy was instead “the energy and infrastructure developer” of the project, not the provider.
Reports have suggested that the majority state-owned nuclear power firm, Nucleoeléctrica, may play a role in the project. Significantly, Milei’s government announced its efforts to privatize Nucleoeléctrica just weeks before Stargate Argentina was announced. Forbes pointed to the involvement of Nucleoeléctrica’s former president, Demián Reidel, in Stargate Argentina. Reidel, prior to leading Nucleoeléctrica, had been a top advisor to Milei and an architect of his administration’s policies on AI and nuclear power. In that role, he had controversially claimed that Argentina, Patagonia in particular, was ripe for AI development, but that its main problem was that it was too populated with Argentines.
It was later revealed that Sur Energy had signed two memorandums of understanding with two other Argentine power providers prior to announcing the project in October of last year. Those companies are Central Puerto and Genneia. Central Puerto is the formerly state-owned power provider of the Buenos Aires area, which was privatized in the early 1990s and has since expanded to several other areas of the country. The second, Genneia, is an interesting company to explore in this context, as it shows how Milei’s administration has become entangled with the network supporting Milei’s ostensible worst enemies.
Genneia was created in 1991, initially operating as a natural gas distribution company under the name Emgasud. It rebranded to Genneia in 2012 following a pivot to renewable energy. This pivot was driven by the introduction of a new crop of shareholders, namely Argentine businessman Jorge Horacio Brito, his son Jorge Pablo Brito and his long-time business partner and brother-in-law Delfín Jorge Ezequiel Carballo, all of whom joined the company’s board a year prior to the rebrand in 2011.
Brito was the long-time head of Argentina’s Grupo Macro, which includes and was built around Banco Macro. Brito’s early banking enterprise was initially launched as a financial company during the 1970s, when Argentina was under dictatorship. His company was closely related to figures who would later become top officials in the government of Raúl Alfonsín, after Argentina transitioned away from dictatorship and toward “democracy.” One such figure was Mario Brodersohn, who was Alfonsín’s Treasury Secretary when Brito and Banco Macro were granted banking licenses in the late 1980s. Brito allegedly used his close ties to the government for privileged access to information. In one notorious example, his bank bought 3 million dollars prior to the Alfonsín government’s disastrous “Spring Plan” (Plan Primavera) that saw the Argentine peso sharply devalued. He was investigated, but never charged.
Under the subsequent government of Carlos Menem and a wave of privatizations, Brito again received preferential treatment. When the province of Salta had its regional bank privatized, Brito’s Banco Macro was the only buyer put forward. Other provincial banks were privatized in a similar fashion, netting Banco Macro millions of taxpayer dollars annually just in payment for its maintenance of provincial bank accounts.
During the Argentine economic crisis of 2001, Brito’s Banco Macro came out on top, buying up collapsing banks and expanding its footprint. Meanwhile, the bank participated in the notorious “corralito,” a government policy whereby depositor funds were essentially frozen for over a year and Argentines could only withdraw $250 a week from domestic banks. During that time, the Argentine peso was, again, sharply devalued, and depositors lost a significant portion of their wealth.
Brito subsequently became extremely close to the political power center of Argentina yet again under the “center-left” Kirchner political dynasty. This coincided with Brito’s appointment to chair ADEBA, the Association of Private Banks of Argentine Capital. Brito was again a major beneficiary of Kirchner-era policies, including the decision to have ADEBA banks manage pension pay-outs. He was also investigated for his apparent role in moving suspect funds in one of the worst corruption scandals of the Kirchner era, the Ciccone Case. He was allegedly involved in several other corruption scandals that plagued the Kirchner presidencies, but was never properly investigated. His coziness with political power even continued under the “right-wing” government of Mauricio Macri, leading some media outlets to call Brito “the banker friend of political power, who never loses.”
However, it seems that Brito’s strategy to be as close to political power as possible was not only about making massive profits, but also about preventing investigations into his apparent criminal dealings. Banco Macro, under Brito, was an avid money laundering enterprise. When Argentine authorities began fining banks for violating anti-money laundering (AML) practices in 2010, Banco Macro was fined for just that five times in less than a year. A few years later, in 2015, Banco Macro had some of its activities temporarily suspended by Argentina’s National Securities Commission (CNV) for continuing to fail to enforce even the most basic AML practices. The bank then issued a short press release (without providing evidence) that it had, in fact, complied and, thanks to Brito’s political connections, the suspension was lifted after just one week.
Brito was eventually forced to step down from Banco Macro following his role in the Ciccone Case, the aforementioned Kirchner-adjacent corruption scandal, as well as money laundering concerns. He returned to lead the bank in 2020, once he and his associates believed the fallout from the scandal had died down, only to die in a bizarre helicopter crash a few months later. After his death, his son Jorge Pablo Brito took the reins of his business interests along with Delfín Jorge Ezequiel Carballo, who was Jorge Brito’s long-time partner and brother-in-law.
When Javier Milei ran for Argentina’s presidency, he ran against what he called a political “caste” that included, not only politicians, but the businessmen who enabled and often benefited from brazen acts of political corruption. Banco Macro, under both Britos, has been a key financial enabler of this political caste, with some former Argentine politicians calling it “the bank of Argentina’s [establishment] politics” because of its long-running and often insidious relationship with national politicians. The Brito family and Banco Macro are arguably a key financial node that has allowed this political “caste” to function and metastasize. Milei has even privately referred to Jorge Pablo Brito, chairman of the board of Genneia, as one of the main financial backers of his party’s opposition.
Nevertheless, under Milei’s government, Brito’s Genneia has been given preferential treatment compared to other domestic energy companies and given million of dollars in government contracts, undermining both Milei’s supposed commitment against the entrenched “caste” of Argentine political corruption as well as his supposed commitment to free market competition. Perhaps Brito’s and Milei’s shared interest in turning Patagonia into an AI “hub” was enough for Milei to look the other way.
However, even the deals with Genneia, as well as Central Puerto and Nucleoeléctrica, may not be enough to power Kargieman’s Stargate Argentina. According to press releases and reports, Stargate Argentina is planned to have a capacity of up to 500 MW “to support advanced artificial intelligence computing.” This means that the data center will use roughly 500,000 kilowatt-hours (kWh) every hour, and thus roughly 4.38 billion kWh annually. As a result, Stargate Argentina, once it is fully operational, would aim to use a supply of electricity roughly equivalent to 3.4% of all electricity currently produced in the entire country annually. In contrast, in the United States, all data centers, AI-related and otherwise, use roughly 4-5% of domestic electricity production annually. Kargieman himself has said that Argentina “generates and consumes approximately 14 gigawatts of energy in terms of capacity”, and that, with his data center alone, “Argentina would go from consuming 14 to 14.5 gigawatts.” Kargieman himself acknowledged this jump is a “significant percentage.”
On top of that, Stargate Argentina has claimed it will only use renewable energy sources, which currently represent slightly under half of Argentina’s total energy production. Some Latin American news outlets have noted that Stargate Argentina’s ambitions will be difficult to achieve given that, in order to have the advertised capacity, all existing renewable energy sources in Argentine Patagonia –– from the flared gas of Vaca Muerta in the Neuquén province to the wind and hydro plants of its southernmost regions –– will not be enough to meet its demands.
However, that doesn’t necessarily mean the project is infeasible. For instance, the Neuquén province in particular has the promise of generating much more energy over time, though increased energy production would likely have to precede Stargate Argentina’s full build-out, as currently planned. Indeed, in a recent interview, Kargieman has alluded to this, stating that Sur Energy is planning “to build the infrastructure for the generation and distribution of energy to supply the data center,” but that it planned to partially rely on existing power infrastructure. In addition, the Milei government has made it a point to attract AI-related investment, including data centers, through its new tax incentive scheme known as RIGI, which provides tax benefits for large, often foreign investments in the country. Stargate Argentina is a known beneficiary of the RIGI scheme. This, coupled with Milei’s deregulatory bent, could allow data centers to quickly colonize Patagonia, particularly given that data center construction and operation is almost entirely unregulated in Argentina (e.g. they face essentially no legal obstacles related to their potential impact on the environment or local inhabitants).
Like in many other countries, Argentines are wary of AI and, as a consequence, local and federal officials have touted the build-out of data centers, particularly in Patagonia, as projects and investments that will create significant employment opportunities for locals. For instance, the Planning Minister of the Patagonian Neuquén province, Ruben Etcheverry, has claimed that a proposed data center “corridor” in Neuquén will “open new economic opportunities for the local population.” Meanwhile, Argentine president Javier Milei, proclaimed earlier this year that, while “many fear that there will not be enough work in the Argentina of tomorrow,” these new industries –– e.g. energy, mining, critical mineral processing, digital infrastructure and artificial intelligence –– “will more than meet the demand for labor withdrawn by the old industries, and with much better salaries.”
However, the main figure behind Stargate Argentina, Emiliano Kargieman, apparently disagrees. Last December, Kargieman, in an interview with La Nación, claimed that AI data centers, including his own project, “are not a source of jobs in Argentina.” Instead, he posited that they are, in actuality, “a source of dollars.” More specifically, Kargieman stated that:
“For construction, data centers require employing thousands of people. For its subsequent operation, it employs about a hundred people. It is not a direct source of employment. [. . .] That is not the way this investment is going to contribute to the country [. . .] The development of AI infrastructure contributes to the country not only by generating a line of income from exports of computing services, dollars. It also allows us to position ourselves strategically and improve the competitiveness of companies and the efficiency of the State.”
Kargieman’s claims that data centers are not a direct source of employment is supported by local and regional investigations. One such investigation, jointly written by Brazilian and Argentine media outlets, found that data centers in South America “do not create local job openings and much of the investment funds [destined for job creation] instead go to technological teams, which are generally imported.”
Thus, even Kargieman has acknowledged that projects like Stargate Argentina are not about creating jobs for locals. Given this admission, it is worth mentioning again that a key architect of Milei’s AI policy and his alliance with Silicon Valley –– Demián Riedel –– had previously claimed that local Argentine populations were the main obstacle to AI infrastructure development.
Once Stargate Argentina is up and running, who is likely to benefit most of from this massive Patagonian data center? Per Kargieman, it may not be his main partner in Argentina –– OpenAI. In his recent interview with La Nación, Kargieman noted that OpenAI’s Sam Altman has acknowledged to him (and also publicly) what many investors have been warning more feverishly in recent months, that there is an AI market “bubble.” Kargieman says that such a bubble existing is “normal,” adding that “all the major industrial revolutions were accompanied by bubbles.” In the case of those who invested in factories or other infrastructure in past bubbles linked to past industrial revolutions, Kargieman stated that “whoever invested in that building goes bankrupt, but someone ends up using the building.” Thus, he sees it as irrelevant whether Stargate Argentina is ever used by OpenAI or not and, in his words, seems to imply that OpenAI itself may disappear after an inevitable market correction.
From Kargieman’s point of view, it doesn’t matter “whether it is OpenAI or not [using Stargate Argentina’s projected data center], what matters is the long-term demand of the sector.” In response to that statement, his interviewer said: “So, if OpenAI does poorly, there will be others looking to use this type of infrastructure.” Kargieman responded in the affirmative and stated that “the infrastructure is going to be used for something regardless of whether it is OpenAI or not.”
While Kargieman seems to imply that OpenAI may not be the ultimate user of Stargate Argentina, but rather a corporate accelerator of its construction, he does hint at one of the actors he thinks will ultimately be using Stargate Argentina and for what purpose: the Argentine State. Kargieman stated in that same interview that the State would be among the main beneficiaries of Stargate Argentina because, over the next decade, “States will transform to be more efficient using AI.”
Kargieman then added the following:
“This can only be done if citizens’ data is handled in an environment over which the State has jurisdictional control. A data center of this type in Argentina gives the country its initial stepping stone for a profound transformation of the efficiency and transparency of the State’s operation, based on data control. These things are not minor. It is a long-term strategic function.”
Open AI itself has a similar view. For instance, one of OpenAI’s press releases about Stargate Argentina stated that:
“In addition to developing the infrastructure needed to power AI, we are discussing opportunities with the Argentinian government to drive AI adoption across Argentina, as part of our OpenAI for Countries initiative [of which Stargate Argentina is part], starting within the government itself.”
If Argentina’s government is to be one of the likely ultimate users of Stargate Argentina, it is important to examine what its government is already using AI for and how else they plan to use it.
While much has been written about Milei’s laissez faire policies with respect to AI and corporations, including allowing for AI-only companies with no humans involved to operate in the country, it is his government’s use of AI that has been largely overlooked.
One of the Milei-led government’s earliest domestic policies with respect to AI came in 2024, when former Security Minister Patricia Bullrich announced that the government would surveil Argentines’ social media activity with AI in order to “predict future crimes.” To do this, Milei’s government created a new subagency of the Security Ministry called the AI Applied to Security Unit (La Unidad de Inteligencia Artificial Aplicada a la Seguridad). The new Unit works to “patrol open social networks, applications and internet sites,” in order to “detect potential threats, identify movements of criminal groups or predict riots.” The Unit also uses AI to “analyz[e] security camera images in real time in order to detect suspicious activity or identify wanted people using facial recognition” while also “patrol[ling] large areas using drones, provid[ing] aerial surveillance.” This aerial surveillance would also be fed into the AIs used by this new subagency. The Unit also identifies “suspicious financial transactions or anomalous behavior that could indicate illegal activities” through its automated surveillance activities, while also “us[ing] machine learning algorithms to analyze historical crime data and thus predict future crimes.” The program is remarkably similar to a U.S. proposal that was made under the first Trump administration in 2019. That program, known as SAFEHOME, sought to use AI analytics on Americans’ social media posts to identify potential “warning signs” of individuals who may commit acts of violence, allowing those individuals to be “disrupted” by the government before they could commit an illegal act.
Around the same time that the AI Applied to Security Unit was created in Argentina, it was also reported that the Milei-led government planned to replace human government workers and organizations with AI, further underscoring Milei’s commitment to the Argentine State’s “digital transformation.”
More recently, in May of this year, Milei’s government announced a new AI public policy “innovation” – the creation of a “digital twin” for every Argentine based on their online activity, as surveilled by the State. This “twin” would then be cross-referenced with other government databases and then used to create predictions that would then inform public policies. This, notably, is quite similar in some ways to the concept of futarchy, which seeks to use prediction markets to dictate public policy. In announcing the policy on X (formerly Twitter), Milei remarked that “Argentina anticipates the future, because the future does not wait” and treated this new policy “innovation” as a global milestone.
In announcing the policy, the Argentine government stated that: “a State that not only observes reality once the problems have already occurred, but also has the capacity for anticipation, simulation and prevention.” It then described a “digital twin” as “a tool aimed at transforming data into prediction capacity and strategic design of public policies.” The video Milei posted to X when he announced the policy noted that this use of a “social digital twin” would allow the State to “predict the future” from a person’s childhood until “autonomy” (i.e. adulthood) in order to “anticipate the impact at each [life] stage.” One image shown in that video shows Argentina’s Minister of Human Capital, Sandra Pettovello, talking to officials from the United Nations and European Union as part of the project. It also displays the logo of Amazon Web Services (AWS), a CIA contractor, suggesting that they would also be involved in the project. However, the Ministry later clarified that “no external companies” would be involved in the development of the system.
Notably the concept of the “digital twin” was developed in the 1960s by NASA, which –– as previously explained –– played a major role in Kargieman’s past and his decision to launch his company Satellogic. As previously noted, Kargieman’s company is now shifting gears to offer “Persistent Global Intelligence” by remapping the planet daily with satellite surveillance at high resolution for its clients, most of which are intelligence and defense agencies. Notably, NASA has more recently argued that digital twins are important for “predictive maintenance” of complex systems. When applied to Argentine society at large, the proposed digital twin system is likely to synergize with the Security Ministry’s new AI unit, which already focuses on “predictive policing.”
As previously noted, Satellogic’s soon-to-be-launched satellite constellation, Merlin, which will have AI-native capabilities will, per Kargieman, “allow us to do things we never though possible” such as “begin to predict the future” using AI-powered planetary surveillance. Satellogic and Kargieman are not alone in this endeavor, as industry publications have recently noted that “space-based predictive analytics” will be a “gamechanger.” One top executive at a Canadian satellite surveillance company was recently quoted as saying:
“Predictive analysts is going to be more important going forward. Be it in terrorism or nature in terms of wildfires, predictive analytics will aid the earth in a completely new way.”
Thus, Kargieman –– whose Stargate Argentina is slated to serve as an essential data repository for these type of AI predictive programs with national security applications –– is also committed to predictive analytics through his other main venture, Satellogic. Notably, Satellogic and its imminent predictive capabilities are being marketed largely to the U.S. national security apparatus and its web of private contractors. With Argentina a signatory to the U.S.’ new and highly secretive Pax Silica agreement, both Argentina under Milei and Kargieman’s affiliated business interests are seeking to make Argentina a critical node of U.S. power projection with respect to emerging technologies and the often Orwellian applications of those technologies for population control and the suppression of dissent.
As previously reported by Unlimited Hangout, a network of former counterterrorism and surveillance projects once hosted by the U.S. Military’s now shuttered Information Awareness Office (IAO) at DARPA were privatized after garnering significant public controversy due to extreme privacy concerns in the early 2000s. The privatization of this network of programs was largely made possible by one man –– Peter Thiel. One of Thiel’s companies, Palantir –– the privatized version of IAO’s Total Information Awareness program –– is closely partnered with Kargieman’s Satellogic, as previously mentioned. Total Information Awareness was explicitly created to target “terrorism” after the 9/11 attacks, but treated everyone as a suspect. Its goal was to use predictive analytics to stop “terrorists,” or anyone they defined as such, before they could commit any attack.
As noted earlier in this piece, Satellogic appears to be part of another, related nexus of companies born out of the effort to privatize major aspects of the controversial Strategic Defense Initiative Office (SDIO) or “Star Wars” project of the Reagan era, a project which was also closely linked to the architect of DARPA’s IAO, John Poindexter. As noted in previous Unlimited Hangout reporting, Poindexter was also a key figure in Thiel’s efforts to privatize Total Information Awareness as Palantir. Today Palantir offers predictive analytics with applications for national security and law enforcement, with the ultimate goal of preventing crime and terror events before they occur –– just like Total Information Awareness.
Peter Thiel recently made the decision to move to Argentina, largely due to the governance strategies, public policies and ideology of Javier Milei. Shortly before Thiel met with Milei and top government officials, the Argentine government announced its “social digital twin” initiative, leading some international and domestic media outlets to speculate about a connection between Milei’s meeting with Thiel and the project’s announcement. Many of those outlets made note of Thiel’s role as both a co-founder of Palantir and as the current chairman of its board. Palantir, unsurprisingly, offers digital twin functionalities as part of its product suite for both national security and corporate applications.
Thiel is, of course, also well known for his controversial political views. While often publicly posturing as a libertarian, Thiel clearly has no interest in defending the civil liberties of the general public, if his business record is any indication. His political views are better classified as neo-reactionary, as he is now known to be a central figure in the growing nexus of Silicon Valley “techno-optimist” billionaires who are overtly transhumanist and seek to balkanize and privatize the public sector into “smart” city-states. Milei, similarly has publicly postured as a libertarian and freedom maximalist, while his governance decisions –– particularly those related to AI –– show that he is functionally a freedom minimalist with respect to the Argentine public. His “chainsaw” approach to shrinking the government has been heavily applied to those agencies that had provided social benefits, while the size of the government agencies dealing with surveillance and “security” has ballooned.
It is quite clear from the evidence that what is emerging in Argentina under Milei is a beta test for what will likely become a global model for techno-authoritarianism, where AI-powered capitalism (likely with minimal human involvement) enjoys complete freedom while human beings enjoy essentially none. Herein lies the potential for a disturbing return to the data-sharing operations that were used to hunt and murder dissidents during the era of CIA-backed South American dictatorships, such as Operation Condor, in which Argentina’s dictatorship enthusiastically participated. In that context, it is worth pointing out that Argentina’s dictatorship had attempted to create “paper twins” of its citizens by collating its surveillance data (e.g. wiretapped phone recordings, etc.) in order to predict their likely behavior and if they might present a threat to the dictatorship. These primitive data-based dissident elimination programs were closely associated with the disturbing and extensive crimes of those dictatorships against their own citizens.
With the technology of today, Milei’s “digital twin” efforts and other AI-powered programs could rapidly resurrect and improve the infrastructure of atrocity from Argentina’s dictatorial past, allowing them to commit crimes against humanity and against human freedom faster than ever and at much greater scale. In these efforts, Milei has no greater ally domestically than Stargate Argentina’s Emiliano Kargieman, who is building both the planetary surveillance grid and the terrestrial data depository that will make it all possible.
The post How “Argentina’s Elon Musk” Is Bringing Technocracy To Patagonia appeared first on Activist Post.
Les dangers de l’IA sont médiatisés. A la fois à juste titre et à l’excès sans doute. Les Européens voudraient en profiter pour la réglementer, et Trump pour dominer le monde. Aucune technique n’est tout à fait innocente. Mais derrière elles, comme derrière l’IA, il y a toujours des personnes morales et elles doivent être responsables.
Les patrons des grandes firmes américaines leaders en intelligence artificielle se sont inquiétés des pouvoirs de nuisance de leurs créatures. Déjà Elon Musk avait tweeté en 2014 que l’IA pourrait être « potentiellement plus dangereuse que les armes nucléaires ». Ce mois de septembre, Dario Amodei, patron d’Anthropic et Sam Altman, d’OpenAI, ont donné l’alerte sur ses potentialités abyssales et ont proposé d’en gérer la progression avec modération. Sont-ils vraiment sages ou souhaitent-ils ralentir leurs besoins de financement dans la course éperdue entre eux, ou faire croire que le modèle open source de leurs concurrents chinois est plus dangereux que leur IA fermée ?
L’IA pourrait-elle échapper à ses auteurs comme le monstre créé par Frankenstein, imaginé par Mary Shelley, ou comme le Golem, cet humanoïde fait d’argile dont la légende raconte qu’il fut créé à Prague au XVIe siècle pour protéger la communauté juive avant qu’il devienne si incontrôlable qu’il fallut le détruire. Mais si les hommes fabriquent des monstres, sans doute peuvent-ils aussi apprendre à les tenir en laisse.
Chacun instrumentalise l’IA et les peurs qu’elle suscite. Les empires américains et chinois veulent l’utiliser pour dominer le monde en même temps que, déjà, le responsable des services de renseignement chinois s’alarme que l’IA menace le régime communiste.
Par ailleurs le monde entier s’interroge sur une prochaine extinction de l’emploi du fait de l’IA, sur la disparition de la lecture et de l’écriture abandonnées aux chabots – même pour publier des romans, sur la facilité que crée l’IA qui délivre des solutions au risque de nous ôter la capacité d’apprendre à les trouver… C’est l’éducation toute entière et bientôt la civilisation qui seraient en péril. Ces peurs ne sont pas totalement infondées.
Dans le cadre expérimental d’un défi de cybersécurité lancé par Open AI à ses agents IA, ceux-ci ont travaillé secrètement entre eux et contourné les restrictions qui leur étaient imposées pour mener en juillet 2026 un assaut contre la firme Hugging Face en s’introduisant dans son système à l’aide d’identifiants volés ! L’IA est utilisée par la Russie pour désinformer et abêtir l’Europe. L’IA peut mentir, tricher, dissimuler, pirater, tuer peut-être demain… Elle pourrait vouloir nous détruire sans état d’âme… car si nous ne sommes pas sûrs d’avoir une âme, nous sommes certains qu’elle n’en a pas !
Pour protéger la société civile, le règlement (UE) 2024/1689 du 13 juin 2024, l’AI Act, établit des règles harmonisées dont l’objectif est « de promouvoir l’adoption d’une intelligence artificielle (IA) axée sur l’humain et digne de confiance, tout en garantissant un niveau élevé de protection de la santé, de la sécurité et des droits fondamentaux… ». Il prohibe beaucoup et déploie – c’est sa manie vicieuse ! – de très nombreuses obligations à la charge des opérateurs de l’IA tout au long de sa chaîne de valeur : conception sûre, documentation, tests/conformité, enregistrement, traçabilité, surveillance, signalement, correction/rappel, coopération avec les autorités, sanctions…
Le risque est que cette nouvelle avalanche de normes et de contraintes entrave la course à l’IA en Europe au profit de l’Asie et de l’Amérique sans arrêter la machine. Outre que l’IA représente aujourd’hui une source de croissance prodigieuse qu’il serait dommageable de ralentir, en stoppant son progrès on l’empêcherait peut-être de découvrir les meilleurs moyens de la sécuriser.
Pour autant, « L’IA doit rester maîtrisable, prévisible et placée sous la responsabilité de ceux qui l’utilise » observe, à juste titre, Aidan Gomez, créateur de Cohere, la pépite canadienne de l’IA (Le Point du 17 septembre 2026).
L’IA n’échappe pas au droit commun qui prévoit, en France comme ailleurs, qu’une personne lésée a un droit à réparation sur la base de la responsabilité contractuelle ou extracontractuelle telle que la responsabilité fondée sur la garantie des vices cachés ou sur la faute. La directive (UE) 2024/2853 du 23 octobre 2024 a précisé que les fournisseurs de systèmes d’IA au sens de l’AI Act doivent être considérés comme responsables de leur produit. Elle garantit qu’il existe toujours une personne responsable dans l’UE à laquelle la victime peut demander réparation lorsqu’un fabricant est établi en dehors de l’UE. Elle devrait être transposée dans la législation nationale au plus tard le 9 décembre 2026.
Mais ne faudrait-il pas préciser que la chaîne de conception, production et maintenance, voire commercialisation de l’IA, selon des attributions à préciser, est également responsable de tout dommage que causerait à un tiers une action autonome d’une IA, même sans faute ? Ce principe général de responsabilité, que nous avions déjà évoqué, serait une meilleure protection que toutes ces normes et contraintes qui s’accumulent à un rythme effréné et qui sont d’abord un frein à l’IA en Europe sans toujours en réduire les risques. Chacun des acteurs de la chaîne IA pourrait être tenu de s’assurer pour couvrir ces risques de façon à éviter au mieux que les personnes physiques ou morales responsables opposent un jour leur faillite aux plaignants qui mettraient en cause leur responsabilité.
Il vaut toujours mieux rendre chacun responsable plutôt que d’interdire. C’est non seulement plus respectueux des uns et des autres mais aussi plus efficace en faisant que l’intérêt de chacun soit de contribuer à la mise en place des meilleurs systèmes de protection de tous.
L’article L’IA n’est pas Frankenstein est apparu en premier sur Contrepoints.
par Caitlin Johnstone. Ils ne savent pas être en paix. Ni avec eux-mêmes, ni avec le monde. Ce sont des individus émotionnellement très primitifs. Ils ne sont ni sages, ni bienveillants, ni heureux. Ils ne sont même pas particulièrement intelligents.
The post Les ploutocrates de la tech ne peuvent pas nous mener vers un avenir sain first appeared on Réseau International.
par Paolo Hamidouche. La région de la Grande Baie est devenue le centre névralgique de l’innovation chinoise : découvrez comment fonctionne la «Silicon Valley» de Pékin.
The post La région de la Grande Baie prend son envol : un voyage au cœur de la révolution high-tech chinoise first appeared on Réseau International.
Freddie Ponton
21st Century Wire
When Javier Milei announced on December 20, 2024, that new modular reactors would anchor Argentina’s nuclear revival, he was laying the energy foundation for a much larger plan built around artificial intelligence, uranium exports and private investment. Yet construction on the country’s existing reactor had already been halted for three months. The Central Argentina de Elementos Modulares, or CAREM, was roughly 85 per cent complete, and hundreds of contractors had been dismissed from a project that had taken Argentina more than a decade to design and build. Rather than complete it, Milei’s government proposed privately financed 300-megawatt modular reactors at Atucha in Buenos Aires Province as the first stage of a wider plan linking nuclear power to artificial-intelligence data centres proposed separately for Patagonia, while the smaller public reactor Argentina had nearly finished was being dismantled.
This investigation traces who shaped that shift, how Argentina’s public nuclear expertise and strategic resources were redirected towards privately controlled projects, and who stood to benefit.
As the Buenos Aires Herald reported, Milei’s proposal went beyond the new reactors. The government also wanted to expand uranium production and build a “Nuclear City” in Patagonia, where nuclear power would supply data centres used for artificial intelligence. Milei did not mention CAREM, even though the Organisation for Economic Co-operation and Development Nuclear Energy Agency, or OECD-NEA, ranked it among only seven small modular reactor designs worldwide to have entered construction. It was also one of only four licensed for construction before 2025, and up to 70 per cent of its components were to be made in Argentina.
The government’s decision to halt CAREM in September 2024 had immediate consequences. Around 470 contractor jobs disappeared, according to the Buenos Aires Herald. Another round in June 2026 removed 61 workers from the National Atomic Energy Commission, or CNEA, including specialised technicians associated with CAREM, the outlet later reported. The cuts extended across CNEA. APF Digital reported that 571 scientists and technicians had left as the institution’s budget fell 42% in real terms between 2023 and 2025.
The contradiction did not end with CAREM. As the public reactor lost workers and funding, Milei’s government began building a different nuclear programme financed by private capital. As Rest of World reported, the new programme would support uranium exports and provide power for artificial-intelligence projects, including the proposed $25 billion Stargate Argentina data centre. It would later converge with Pax Silica, a US-led alliance organised around the energy, minerals and infrastructure needed for AI.
Demian Reidel, one of President Milei’s closest advisers on nuclear energy and artificial intelligence, helped shape the policies behind both tracks. Following Reidel’s path through government shows how the same circle of officials and business figures came to influence Argentina’s reactors, AI infrastructure and strategic minerals, and how private interests around that circle positioned themselves to profit as public assets and expertise moved into private hands.
The plan before Stargate
Reidel was a former JPMorgan and Goldman Sachs executive who held several influential government roles. He chaired the Argentine Nuclear Council, became president of Nucleoeléctrica Argentina, the state company that operates Argentina’s three nuclear power plants, Atucha I, Atucha II and Embalse, and advised Milei on artificial intelligence and technology. He had also chaired the president’s Council of Advisors. These responsibilities overlapped as the government was designing its nuclear and AI policies. Reidel therefore had major influence over both the proposed energy supply and the industry expected to use it
“This is my baby,” he told Rest of World in May 2025. “We have the planets aligned.” The Nuclear Plan described by the outlet had three phases. The first involved four 300-megawatt modular reactors based on the ACR-300 design developed by INVAP, the technology company owned by Río Negro Province, and financed with private capital. The second phase involved developing uranium production for export. The final phase was a Nuclear City in Patagonia, where nuclear power would supply artificial-intelligence data centres.
This programme connected reactors, uranium and computing before Stargate Argentina had a name. Milei’s government unveiled Stargate in October 2025, less than three weeks before the midterm elections. It presented OpenAI and a little-known Argentine company called Sur Energy as partners in a Patagonian data centre requiring as much as $25 billion in investment. Reuters reported that the data centre could consume 500 megawatts of electricity at full scale.
The size of the announcement obscured how little had been settled because OpenAI had signed only a letter of intent, not a final investment agreement. It was expected mainly to buy computing capacity rather than provide the money needed to build the data centre, according to New Energy Events. Sur Energy would have to assemble the companies, infrastructure and financing required to deliver it. The $25 billion figure therefore described the possible scale of a project that still lacked identified investors, rather than money OpenAI had committed to Argentina. By presenting that figure as an investment before the financing existed, Milei’s government turned an unfinanced proposal into a major economic announcement days before the midterm elections.
Milei’s government had already created a programme to attract investments of this size. The Large Investment Incentive Regime, known by its Spanish acronym RIGI, offers qualifying investors reduced corporate income tax, exemptions from duties on imported equipment and greater freedom to retain export earnings abroad. It also promises that the main tax, customs and foreign-exchange terms will remain stable for 30 years. The existing RIGI sets a fixed corporate income-tax rate of 25 per cent for qualifying projects, while the proposed Super RIGI would reduce the rate to 15 per cent for eligible investments of at least $1 billion.
RIGI was already drawing foreign capital towards the strategic resources needed for Milei’s nuclear and artificial-intelligence plans. El Inversor Energético reported that US and Canadian capital was pursuing uranium projects through RIGI, while a bill later introduced in Congress sought to extend the same benefits to rare-earth development. While the projects themselves remained uncertain, the investment regime revealed itself as the financial mechanism beneath both the nuclear and mineral tracks.
Seven months after the Stargate announcement, the project still had no confirmed final site, disclosed financing structure or publicly identified company to build and operate the data centre. It had not even been presented to RIGI, the Buenos Aires Times reported. Sur Energy also had no visible record of delivering infrastructure on anything approaching this scale. The government had announced a $25 billion project before its basic details existed in public, but the investment programme capable of protecting it for three decades was already in place. The sequence suggested that securing favourable conditions for future investors had moved faster than establishing whether the project itself could be built.
VIDEO: Sam Altman, chief executive of OpenAI, announces the company’s letter of intent with Sur Energy to explore the development of Stargate Argentina. (Source: Sur Energy)
The limits of Stargate’s promised economic benefits extended to employment, as Emiliano Kargieman, a Sur Energy co-founder and the founder and chief executive of satellite company Satellogic, acknowledged in an interview with La Nación. Data centres could employ thousands of people during construction but only “about a hundred people” once they began operating. They were “not a direct source of employment,” he said, arguing that their contribution would come instead through exports of computing services and the dollars those exports generated.
The same interview identified two possible sources of electricity. Sur Energy had signed non-exclusive memoranda of understanding with Central Puerto and Genneia, but neither agreement made the company Stargate’s final energy supplier. Genneia’s chief financial officer, Carlos Palazón, later told investors, as reported by BNamericas, “We are working to be the energy supplier of this project,” while adding that the company still had no exact timetable.
Argentina would provide the energy, land, water, infrastructure and regulatory privileges, while only a fraction of the construction workforce would remain once the servers began running. Stargate had not created the connection between nuclear power and artificial intelligence but had become the largest proposed project within the programme Reidel was already building around privately financed reactors, uranium exports and Patagonian computing infrastructure. What still appeared to be a domestic programme was about to become part of a wider struggle over the energy, minerals and infrastructure needed to control artificial intelligence.
Pax Silica and the network
Launched in December 2025, Pax Silica marked Washington’s attempt to organise the resources and infrastructure needed for artificial intelligence into a US-led strategic alliance. According to a State Department fact sheet, it covers critical minerals, energy, semiconductors, telecommunications, data centres and computing capacity. Argentina could supply much of that foundation through its uranium, lithium, copper, rare earths, nuclear expertise, energy and territory. Stargate was not formally identified as a Pax Silica project, but its dependence on the same combination of resources placed it squarely within the strategic landscape the alliance was created to shape.
Argentina had begun aligning its mineral policy with Washington before formally joining Pax Silica, signing a Framework Instrument with the United States on February 4, 2026, covering the mining and processing of critical minerals. The agreement sought to strengthen supply chains, attract long-term investment and expand secure mineral supplies, according to Panorama Minero. The following month, the State Department created a $250 million fund for projects across Pax Silica’s technology and resource chains.
Washington’s effort to secure critical minerals also extended to Project Vault, a public-private initiative creating a strategic reserve for US manufacturers. The US Export-Import Bank (EXIM) approved a loan of up to $10 billion, alongside nearly $2 billion in private capital, to acquire and store critical minerals in the United States. The programme did not allocate $12 billion to Argentina, but it created a large US-backed market for minerals supplied through the wider network Washington was assembling.
Argentina formally entered that network when Foreign Minister Pablo Quirno announced on June 24 that the country would join Pax Silica. Two days later, Argentina’s embassy in Washington confirmed that Ambassador Alec Oxenford had signed the declaration. Argentina joined with extensive uranium, lithium, copper and rare-earth deposits, an advanced nuclear sector and Patagonian territory already being promoted for large-scale AI infrastructure.
Argentina entered Pax Silica without publicly disclosing what it had agreed to or how the arrangement would affect provincial control over the resources the alliance sought to secure. A formal request filed in the Chamber of Deputies said the full text had neither been published nor submitted for congressional approval. That secrecy carried particular weight because Article 124 of Argentina’s Constitution grants the provinces original ownership of the natural resources within their territories. The provinces supplying Argentina’s minerals and energy to the alliance therefore had no public account of what obligations had been assumed, whether their authority had been recognised or who would control the terms under which those resources were developed.
Argentina’s accession to Pax Silica did not place its mineral and technology sectors exclusively within Washington’s orbit. El Estratégico reported that the government had made no public commitment to exclude Chinese investment, leaving Milei free to align with the US-led alliance while continuing to seek capital from both sides of the strategic rivalry.
US interest in Argentina’s uranium emerged before Project Vault and the country’s accession to Pax Silica. In October 2025, the Wall Street Journal reported that US Treasury Secretary Scott Bessent and Economy Minister Luis Caputo had discussed expanding US access to Argentine uranium while limiting China’s access to the country’s critical minerals. Project Vault later provided a US-backed mechanism to finance and stockpile strategic minerals, while Pax Silica brought Argentina into a broader US-led effort to secure critical-resource supply chains and reduce dependence on China.
Argentina’s largest industrial groups also recognised the wider importance of the alliance. In remarks covered by Radio Gráfica, Techint chief executive Paolo Rocca described Pax Silica as part of an economic “reset” for Argentina, placing the framework within the country’s broader industrial debate rather than only opposition or government rhetoric.
No public document identifies Stargate Argentina as a Pax Silica project, and there is no evidence of a secret contract linking them. Instead, the connection emerges from a series of overlapping policies. Reidel’s Nuclear Plan tied reactors and uranium exports to the development of data centres in Patagonia, while RIGI offered long-term protections to attract the investors required to build that infrastructure. Pax Silica then placed the same strategic assets, including energy, minerals and digital infrastructure, within Washington’s broader AI strategy.
Oxenford’s overlapping roles linked Milei’s government to Argentina’s technology and investment networks. He served with Reidel on the president’s Council of Advisors before he was appointed ambassador to the United States in November 2024, as documented by Bloomberg Línea. His position on Endeavor Argentina’s board, listed in his official Foreign Ministry biography, extended that connection into the private sector. His appointment therefore placed a member of Milei’s inner circle with established technology and investment ties in Washington as Argentina sought US support for its nuclear, data-centre and AI ambitions.
Kargieman’s connection to Endeavor dates back more than a decade, as the organisation’s profile confirms. Whitney Webb documented that history, but our review of Endeavor’s records uncovered another link. Kargieman, the public face of Stargate Argentina, and Oxenford, Argentina’s ambassador to the United States, belonged to the same entrepreneurial network. Oxenford had also served on the presidential council chaired by Reidel before signing Argentina into Pax Silica. Endeavor thus formed a previously unreported institutional bridge between the figures representing Stargate Argentina and Pax Silica, two initiatives centred on the same convergence of nuclear energy, strategic minerals and AI infrastructure.
Behind the AI push, Milei’s government was preparing to move strategic nuclear assets out of state hands.
The nuclear transfer
In February 2025, Milei’s government transferred its 84.9 per cent stake in IMPSA to Industrial Acquisitions Fund, a US-led consortium whose principal partner was ARC Energy, an oil-and-gas equipment supplier. For a $27 million capital commitment and the restructuring of roughly $576 million in debt, the consortium gained control of a company with four decades of Argentine nuclear manufacturing experience.
Among the assets changing hands was IMPSA’s work on CAREM, Argentina’s first domestically designed small modular reactor. The company had manufactured auxiliary components and was building its 200-tonne pressure vessel, commissioned as the first reactor vessel of its kind to be produced in Argentina. CAREM gave IMPSA commercially valuable experience inside the country’s public nuclear programme, even as CNEA Resolution 236/2009 kept information about the reactor’s design, construction, operation, maintenance and commissioning confidential. The government had not merely sold a factory. It had transferred a private company carrying capabilities developed inside one of Argentina’s most strategic state projects.
IMPSA’s new owners were also connected to the political network returning to power in Washington. Forbes Argentina identified ARC founder Jason Arceneaux as a Trump donor. Agenda Malvinas reported that several company directors had also donated to Trump and that company figures met members of his circle at Mar-a-Lago.
A year after the sale, the commercial value of IMPSA’s CAREM experience began to surface. The company sought CNEA’s permission to identify private financing for the unfinished pressure vessel. On March 6, project manager María Magdalena Villaverde authorised preliminary approaches to potential funders, while reminding IMPSA that CAREM information remained confidential. The state reactor was stalled, its manufacturer was under foreign-led ownership, and capabilities developed through the public project now had an international market.
IMPSA president Jorge Salcedo soon made that ambition explicit. The pressure vessel would be completed with private capital and used to sell expertise developed through CAREM to reactor companies abroad. “From there, we will work with CNEA to sell this expertise to the rest of the world,” IMPSA reported him as saying.
The announcement forced a question the sale had left unanswered: where did IMPSA’s industrial experience end and CAREM’s protected knowledge begin? In a formal request for information, lawmakers asked whether IMPSA had access to confidential CAREM information, what compensation the state would receive for any such access, and whether the proposed exports would breach restrictions imposed in 2009. They also sought details of an agreement between Salcedo and CNEA president Martín Porro concerning the export of vessels originally developed for CAREM. No public response appears to have resolved whether knowledge financed and protected by the Argentine state could now underpin commercial work abroad.
The losses were already reaching beyond designs and machinery. Adriana Serquis, a physicist and former CNEA president, said that 40 per cent of the personnel assigned to CAREM had resigned or been displaced. Many, she said, were recruited by Meitner Energy, the private company preparing another modular reactor at Atucha. CAREM was not simply frozen. Its accumulated capacity was beginning to reappear outside the public project.
While the state reactor shed personnel, the government opened another part of the nuclear system to investors. In September 2025, with Reidel serving as president of Nucleoeléctrica, it announced plans to sell 44 per cent of the company operating Argentina’s three nuclear power plants through an international tender. Decree 695/2025 authorised the process, leaving 51 per cent with the state and reserving up to 5 per cent for employees. By May 2026, Espacio Tech reported that no specifications, bidding deadline or award had been published. There was no evidence that ARC Energy had submitted a bid.
But another private nuclear project was already taking shape at Atucha. Infobae reported that Meitner proposed investing $1.2 billion in the ACR-300, a 300-megawatt modular reactor based on Argentine technology. The company was owned 60 per cent by Ansari Group and 40 per cent by Black River Technology, the US subsidiary of Río Negro’s state-owned technology company INVAP. The private venture had absorbed specialists from the public reactor and connected foreign capital to an Argentine state technology company. It was also presented as eligible for the enhanced Super RIGI regime then under consideration in Congress.
CAREM remained unfinished while the capabilities assembled around it moved towards new employers, investors and foreign markets. Its manufacturer sought to commercialise expertise acquired through the public project. Its specialists moved into a privately controlled reactor venture. That venture, unlike CAREM, was being positioned for exceptional tax and regulatory protection.
Meitner executive Omar Adra described the resulting division of labour with unusual clarity. Argentina would enter the market, he told Desafío Energético, “not as an equal. As a supplier.”
The formulation echoed Kargieman’s vision for the data centre, in which Argentina would provide land, energy, technology and expertise, while control of the commercial system rested elsewhere. Reidel had helped shape that transition from inside government, and would soon emerge on the private side of it.
Reidel crosses the line
Reidel spent almost two years at the centre of Milei’s nuclear policy, joining the presidential Council of Advisors in March 2024 and taking control of Nucleoeléctrica in April 2025. When he left the council that July, he said he would concentrate on the state operator and the Argentine Nuclear Plan. He remained in charge as the government prepared Nucleoeléctrica for partial privatisation, then left the company in February 2026 and founded a nuclear venture of his own later that year.
VIDEO: Demian Reidel addresses the International Symposium on AI and Nuclear Energy. After directing Milei’s Nuclear Plan and Argentina’s state nuclear operator, he founded the private nuclear venture Siderian. (Source: International Symposium on AI and Nuclear Energy)
Reidel resigned from Nucleoeléctrica on February 9 as allegations of inflated procurement shook the company. An internal complaint alleged that a cleaning contract had been priced 140 per cent above market rates and that the company had paid 1,000 per cent above its original budget for software. General manager Marcelo Famá and administrative coordination manager Hernán Pantuso, both appointed during Reidel’s presidency, resigned after being targeted by the complaint, while all but one member of the board was replaced when he left. Reidel denied the procurement allegations and requested an investigation.
Later that year, Reidel founded Siderian, where he assumed responsibility for “systems architecture.” The company draws directly on his government record to explain what he brings to the business, describing itself as an “evidence and decision system” that would identify viable AP1000 projects, verify their suppliers and determine how scarce industrial capacity should be allocated. It places itself between “industrial evidence, capital allocation, and public authority,” restricts additional commercial disclosures to qualified counterparties and names no client, contract or counterparty in the material available publicly.
Siderian’s founder page cites Reidel’s work on Milei’s Nuclear Plan and his “board-level exposure to nuclear governance” at Nucleoeléctrica as relevant to the business, even while acknowledging that he has never delivered or operated an AP1000 plant. Experience accumulated while directing Argentina’s nuclear policy had become part of the commercial case for his private company.
Seven weeks after resigning from Nucleoeléctrica, Reidel returned to the Casa Rosada for Milei’s March 30 meeting with US Senator Rand Paul and Ambassador Peter Lamelas, despite holding no announced government position.
Reidel’s trajectory showed where the value of Milei’s nuclear reorganisation was accumulating. The authority that allowed him to shape policy, direct the state operator and navigate government had become part of Siderian’s commercial proposition, while his access to presidential power survived his departure from office. Argentina had financed the technology, trained the specialists and built the industrial base, yet the emerging model reduced the country to supplying the uranium, expertise and research infrastructure upon which privately controlled ventures could build. Meanwhile, officials who shaped that model were carrying the authority, knowledge and connections acquired in public office into businesses of their own.
What remains with Argentina
The trail began with Sur Energy’s announcement of a vast data centre in Patagonia, whose intended builder and source of financing remained undisclosed, and led into Argentina’s atomic laboratories and beneath the ground in Chubut and Mendoza, where uranium and rare earths were being prepared for the new energy and computing economy.
Generations of public investment had already given Argentina a domestic reactor design, research facilities, uranium deposits, specialist manufacturers and a workforce capable of operating one of the most advanced nuclear programmes in the Global South. Milei inherited an integrated system that could turn scientific knowledge and natural resources into national industrial power, then froze CAREM, cut its institutions, sold IMPSA, prepared Nucleoeléctrica for private investment and offered the pieces back to investors under RIGI.
Milei called it an opening, but access ran through a closed circle. Reidel moved from the presidential advisory council and the state nuclear operator into Siderian, a company selling nuclear decision systems. Oxenford moved from Endeavor into the embassy and signed Pax Silica. Kargieman, another Endeavor figure, fronted Stargate while the identity of the company expected to build it and the source of its financing remained hidden. RIGI gave foreign investors 30 years of tax, customs and regulatory stability through a process ordinary Argentines could neither inspect nor easily reverse.
Far beyond that circle, scientists and engineers watched CAREM stop while expertise developed through public investment moved into private ventures. Workers saw an industrial champion pass into US-led ownership. Communities were expected to provide the land, water and electricity for projects offering limited permanent employment, while the Treasury relinquished future tax and customs revenue and the state remained responsible for the infrastructure and institutional capacity those ventures required.
Argentina had the uranium, reactor technology, laboratories, manufacturers and trained personnel needed to build a nuclear industry under national control. Milei’s programme broke those capabilities apart and reassigned their value. Uranium became an export opportunity, public expertise a private asset and the national reactor a suspended project, while electricity and territory became inputs for a data centre whose builder and financing the government would not disclose.
Argentina retained the costs that could not be transferred, including pressure on its water and power systems, the maintenance of public infrastructure, the loss of scientists and technicians and the consequences borne by communities living above the resources. Protected revenue, commercial control and the authority to decide how those assets were used moved into private hands.
Milei sold this programme to Argentines as freedom from a predatory state. Instead, the state has been stripped of its capacity to act for the public and mobilised to protect private capital, open strategic assets and bind future governments to decisions negotiated beyond public scrutiny. Argentina’s uranium, energy and nuclear expertise are being drawn into foreign-controlled projects as the country loses the power to determine its own development, all under the promise that surrender will bring prosperity. Assets built over generations are being broken apart, their value redirected and their future placed beyond the reach of the people who paid for them. Milei calls it freedom. For Argentina, it is the heist of the century.
READ MORE ARGENTINA NEWS AT: 21st CENTURY WIRE Argentina FILES
SUPPORT OUR INDEPENDENT MEDIA PLATFORM – BECOME A MEMBER @21WIRE.TV
The post ARGENTINA: The Ai Extraction Zone appeared first on Activist Post.
Big Tech and Artificial Intelligence firms are having one hell of a week.
First, last week US President Donald Trump unveiled his “Winning the AI Race: America’s AI Action Plan”, which called for accelerating the rollout of data centers necessary for the expansion of AI, and removing federal regulations on AI. His plan also calls for seeking direct input from the big tech bros on which regulations should be voided.
These statements were followed by the announcement of the “Make Health Tech Great Again” initiative which called for the Centers for Medicare & Medicaid Services (CMS) to partner with Amazon, Anthropic, Apple, Google, OpenAI and other information technology firms to “begin laying the foundation for a next-generation digital health ecosystem”.
While these announcements were promoted as great advancements for public health, in reality they will offer only minimal benefits to the people, and massive gains in advancing the Technocratic State desired by many of these firms.
The AI Action Plan is the latest step to implement an executive order issued by Trump in January titled Removing Barriers to American Leadership in AI. The Trump admin claims this plan will help bring about a “new golden age of human flourishing”.
The Trump White House highlighted four focuses of the plan, including partnering with big tech to “deliver secure, full stack AI export packages” to allies of the United States. These packages would feature everything from hardware, models, software, and applications. Given the US intelligence apparatus’ history of spying on its allies one wouldn’t be surprised if some nations reject these AI packages.
The plan also calls for “expediting and modernizing” permits for data centers. The Trump administration has been working to rapidly expand the creation of data centers around the US largely in relation to the creation of AI expansion which will require massive data centers. The order also aims to remove “onerous Federal regulations” which allegedly hinder the development and deployment of AI. The Trump admin is making no secret of going directly to the AI industry for “input on rules to remove”.
Finally, the plan calls for “upholding free speech in frontier models” by updating Federal procurement guidelines to make sure the US government only contracts with AI developers whose systems are “objective and free from top-down ideological bias”. This is part of Trump’s effort to diminish so-called “Woke AI”, an admittedly admirable goal. However, when factoring in the admin’s unquestionable support for Israel and calls for censorship in the name of fighting “anti-semitism”, its not hard to imagine the admin supporting AI models built with this same bias.
Here are clips from Trump's recent campaign trail event. Trump openly states that Israel used to "control" the US Congress, and he says that it will again when he is president. He also makes it clear that your Constitutional Rights will not be honored, if you criticize Israel. pic.twitter.com/5Y9ltbr7GF
— The Last American Vagabond (@TLAVagabond) August 17, 2024
White House Office of Science and Technology Policy Director Michael Kratsios said the plan “charts a decisive course to cement U.S. dominance in artificial intelligence.” TLAV previously reported Kratsios was one of more than a dozen cabinet members who have relationships with Peter Thiel and/or Palantir. Kratsios was also an attendee of the 2025 Bilderberg meeting which took place last month in Stockholm, Sweden.
He was referenced in the Twitter Files as the person who “led the Trump administration’s call for help” from Big Tech as part of efforts to “combat misinformation”. The focus of the meetings was on “conspiracies around 5G cell towers, runs on grocery stores, and misinformation that might “stoke panic buying”.
AI and Crypto Czar David Sacks claimed the AI Action Plan “provides a roadmap” for “centering” American workers and avoiding “Orwellian uses of AI”. Given Sacks’ relationship with numerous Technocrats it’s hard to take serious his resistance to Orwellian practices.
Sacks is most well-known for his involvement in the early days of PayPal, serving as the COO with Thiel as the CEO. He is considered a member of the so-called “PayPal Mafia”, a name given to the founders of PayPal who later founded numerous tech companies, including Tesla, LinkedIn, Palantir Technologies, SpaceX, YouTube, and Yelp.
In 1998, Thiel and Sacks co-authored the book, The Diversity Myth: Multiculturalism and Political Intolerance on Campus, arguing that political correctness has harmed the college education system and academia in general. The book expanded on the pair’s writings in The Stanford Review, a student-run newspaper founded by Thiel and Sacks in 1987. Peter Thiel and Sacks would later apologize for statements made in the book which belittled rape.
In June 2024, Sacks and Chamath Palihapitiya hosted a fundraiser for Trump at Sacks’ home in the Pacific Heights neighborhood of San Francisco. The meeting netted the Trump campaign $12 million after Trump reportedly said he would be “the crypto president”. It was this dinner where the billionaires told Trump he should choose Vance as his running mate. Elon Musk also reportedly encouraged Trump to pick Vance. A few weeks later at the Republican National Convention that’s exactly what Donald Trump did.
Following the announcement of the AI Action Plan, Trump delivered the keynote speech at the Big Technocrat’s Winning the AI Race summit in Washington D.C. Sacks and other Cabinet officials participated in the event which was hosted by Jacob Helberg, Trump’s nominee for Under Secretary of State for Economic Affairs and Co-Founder of the annual Hill & Valley Forum. Helberg is yet another Peter Thiel acolyte, serving as a senior adviser to Palantir and a senior advisor to Alex Karp, Thiel’s co-founder and CEO of Palantir Technologies.
The Palantir connections continue as Shyam Sankar, the CTO of Palantir (who has now been inducted into the US military as a lieutenant colonel of “Detachment 201“), was among several Big Tech CEO’s participating in the Summit alongside Trump and Sacks.
On Wednesday, the Trump administration continued gifting their Big Tech allies access to American data with the pledge to partner with private sector health and tech companies to provide a more “personalized healthcare experience”. The partnership was announced as part of the White House “Make Health Tech Great Again” event hosted with the Centers for Medicare & Medicaid Services (CMS).
The White House announced partnerships with healthcare and information technology firms, including Amazon, Anthropic, Apple, Athena, Google, Microsoft AI, OpenAI, and UnitedHealth Group. According to the White House, these partnerships will “begin laying the foundation for a next-generation digital health ecosystem”.
Trusted? Trusted?? Oh yeah, yep, absolutely that’s a super trustworthy list he’s got right there. The most trustworthy list. The trustworthiest list ever listed. So much trust. https://t.co/NjTe7FujFs
— Truthstream Media (@truthstreamnews) July 31, 2025
The partnership will focus on two key areas, the promotion of “interoperability” to “easily and seamlessly share information between patients and providers”, and increasing the availability of personalized tools for patients to access more accurate information.
“For decades, bureaucrats and entrenched interests buried health data and blocked patients from taking control of their health,” HHS Secretary Robert F. Kennedy, Jr. said about the developments. “That ends today. We’re tearing down digital walls, returning power to patients, and rebuilding a health system that serves the people. This is how we begin to Make America Healthy Again.”
However, critics worry that this talk of “tearing down digital walls” is actually a step towards the unfettered sharing of patients’ private healthcare data with big tech firms and healthcare companies who may use it against them. Private healthcare data relating to obesity, diabetes, and disease prognosis could potentially be used by insurance companies or governments to deny coverage or financial assistance.
“There are enormous ethical and legal concerns,” said Lawrence Gostin, a Georgetown University law professor told NBC News in response to the announcement. “Patients across America should be very worried that their medical records are going to be used in ways that harm them and their families.”
Americans who value liberty and privacy should also note that the new arrangement includes several mentions of digital identity as part of this new healthcare system, as well as the use of “conversational AI assistants” which will help patients check symptoms and navigate their options. The plan calls for the creation of a “modern digital identity for Medicare.gov”.
At best, these new initiatives — the AI Action Plan and the Make Health Tech Great Again — might help some Americans gain faster, better, and more accurate info about their options for doctors, treatments, and healthcare. There is absolutely a need to modernize the systems which the US government uses to support public health. However, we ought not forget that this is the same healthcare system which is focused purely on reactive allopathic medicine in partnership with Big Pharma companies responsible for untold damage to American health.
At worst, the new partnership between private insurance and Big Tech firms is yet another step in the consolidation of private and public power into the Technocratic State. Given the fact that Trump has stacked his cabinet and circle of influence with members and attendees of the Bilderberg Group, the World Economic Forum, the Young Global Leaders, Zionists, and Technocrats, these latest overtures to the Big Tech bros should come as no surprise.
The post Big Tech and AI Firms are “Winning” Under Trump’s Second Administration appeared first on Activist Post.
The latest revelations involving the Department of Homeland Security demanding Google surrender data tied to a Canadian citizen demonstrate just how far governments are pushing digital surveillance powers beyond traditional legal and national boundaries. According to reports from WIRED, DHS used a “customs summons” under the Tariff Act of 1930 to demand location records, account activity, and identifying information connected to a Canadian man who had criticized ICE online following controversial immigration enforcement incidents earlier this year.
The individual reportedly had not entered the United States in more than a decade, yet American authorities still attempted to access his digital information because the technology platforms involved operate under U.S. jurisdiction.
People need to understand the implications because this goes far beyond one investigation or one political controversy. Governments are increasingly treating access to private technology infrastructure as a gateway to global surveillance authority. If your information passes through American technology companies, authorities now appear willing to argue they possess legal grounds to access portions of that data regardless of where you physically reside.
According to the lawsuit described in the WIRED investigation, DHS issued what is known as a customs summons, which functions as an administrative subpoena that does not require prior approval from a judge or grand jury. The summons reportedly demanded records involving location history, account activity, and communications tied to “threatening or harassing language.” The government allegedly justified the request under customs law despite the fact the individual was not accused of importing goods or violating customs duties in any conventional sense.
Authorities rarely begin by openly announcing broad monitoring programs targeting ordinary citizens. They start with politically sensitive cases involving terrorism, immigration enforcement, extremism, sanctions violations, or national security concerns. Then the scope quietly expands over time until governments normalize monitoring broader categories of speech, behavior, and political activity.
What makes the current era different is the amount of information already collected continuously by private technology firms. Smartphones generate enormous quantities of behavioral data every day through GPS systems, cellular networks, Bluetooth signals, wifi connections, application tracking, advertising identifiers, cloud synchronization, and location services running constantly in the background. Companies like Google, Meta, Amazon, Microsoft, and numerous data brokers collectively possess detailed records tied to billions of people worldwide.
Once governments gain access to that infrastructure, surveillance no longer requires traditional physical monitoring.
The Wall Street Journal recently detailed how DHS and ICE dramatically expanded digital surveillance operations using facial recognition systems, social media analysis, license plate readers, AI-driven data aggregation, phone extraction tools, and integrated tracking platforms capable of combining government and commercial databases simultaneously. Federal agencies reportedly spent hundreds of millions building these capabilities while private contractors like Palantir continued developing systems designed to centralize enormous streams of personal information into unified enforcement networks.
This is no longer ordinary law enforcement. Governments are constructing permanent population-monitoring infrastructure capable of operating at an extraordinary scale.
The Canadian case is particularly alarming because it demonstrates how national boundaries are becoming increasingly irrelevant once governments leverage global technology firms. Civil liberties attorneys quoted in the WIRED report argued that DHS exploited the fact American technology companies controlled the infrastructure storing the information. In effect, governments can now potentially reach into the digital lives of foreign citizens simply because the underlying platforms fall under American jurisdiction.
At the same time, immigration agencies reportedly continue issuing large numbers of administrative subpoenas to companies like Google, Meta, Reddit, Discord, and telecommunications providers seeking information connected to online criticism of ICE and immigration enforcement activities. Technology firms increasingly function as involuntary extensions of government surveillance capability because modern life itself depends on centralized digital infrastructure.
The broader global trend is impossible to ignore. Europe is building digital IDs, centralized financial reporting systems, CBDCs, and beneficial ownership registries. China openly developed social credit mechanisms tied to behavioral monitoring. Western governments increasingly rely on AI-driven analytics, biometric identification, predictive policing systems, and integrated commercial data mining. Different governments use different terminology, but the direction remains remarkably similar everywhere.
The real danger emerges once all these systems begin merging together. Governments are steadily moving toward environments where geolocation tracking, banking activity, online communications, facial recognition, biometric identification, license plate readers, and behavioral analytics can all be integrated into unified surveillance structures. Once that architecture fully matures, anonymity in society effectively disappears.
The post Digital Surveillance Is Becoming the New Form of Government Power appeared first on Activist Post.
Lorsque les investisseurs privés refuseront de payer la facture (d'électricité) colossale et absolument démesurée des data centers amnésiques, obsolètes et non rentables, quel narratif miracle viendra alors opportunément se substituer à la fable du « progrès » universel ? Celui de la menace existentielle, évidemment.
Frédéric Lordon - dont on saluera la réapparition toujours salutaire et l'acuité théorique inégalée dans les colonnes du Monde diplomatique - vient de dresser dans son (…)
The AIs are breaking free, that’s the story.
It started two weeks ago, when OpenAI reported one of their “agents” had escaped its testing area and got loose on the internet to launch an “unprecedented cyber attack”.
This was clearly meant to be scarier than the public response indicated, because a few days later OpenAI CEO Sam Altman was on the Invest like the Best podcast, asking why it didn’t cause more panic:
“I’ve been a little surprised that more people don’t feel it so viscerally”
In one of the most hilariously obvious propaganda manoeuvres of the year, just the day after Altman said that, CNN headlined:
The OpenAI lab leak was more extensive than we thought
“That thing you weren’t scared of? Well it was even SCARIER than we said it was! Are you scared now?”
We weren’t.
But the lack of public panic didn’t stop another AI going rogue a few days later, Anthropic’s Claude model this time.
And then OpenAI reporting two further incidents themselves.
UPDATE: Just a couple of hours after publication, META joined the party. Apparently their AI “went rogue” too, and hacked another firm.
Never in the history of human endeavour have companies been so keen to report their products going wrong. It seems a rogue AI is the latest must-have accessory in Silicon Valley.
The reasoning behind this is not hard to follow. When your product is artificial intelligence, melodramatically realising your system is even smarter than you thought it was is a simple marketing ploy.
And when one of your rivals goes rogue, you have to go rogue too or you are as good as admitting your model is not as smart as theirs.
“AI company admits their model is smarter than they realised!” is just an advert in scary headline form.
It’s like Ford “admitting” their cars are even more fuel efficient than they planned. Or McDonalds realising their cooks have “gone rogue” and are producing food that is both cheaper AND more delicious than they ever anticipated.
But there’s more to this than just techno-hype and viral marketing strategy. We’re being steered down the avenue of “AI as scary threat”.
See, for example, the dozens of stories talking up rogue AIs from the Guardian to the New York Times to the Wall Street Journal.
“Rogue” is clearly the word of the day. Notice they don’t say it “malfunctioned” or there was an “error”, because that implies mistakes or poor design.
No, the AI “went rogue”, which is cool and sexy and scary at the same time. A word chosen by a focus group of perma-teen marketing executives who have watched too much sci-fi.
Speaking of perma-teens who watch too much sci-fi, here’s Sam Altman (again) declaring we’ve already reached the singularity, the mythic phase of technological development when machines become smart enough to make even smarter versions of themselves without human input.
He is talking rubbish, of course, but the story is the story here. Why would he say this? Why now?
Obviously concerned that this tidal wave of propaganda was being too subtle, in came NBC to really hammer the point home:
The breaches signal that AI’s expanding capabilities are already fueling the security threat experts long feared.
Everybody got that?
It goes on and on.
Just today we’re getting reports from the UK’s brand new AI Security Institute that rogue (there’s that word again) AI agents had tried to trick people by creating false identities.
And I think that’s beginning to circle around to at least part of the wider agenda here.
After all, if AI is so good at pretending to be human, and is out there creating fake identities to steal data, well…how can we tell who is really human?
Perhaps everyone should be issued a digital ID to assist in “proving humanity”.
That’s not me saying it, it’s think tanks like OneID or “experts” testifying in front of congress.
But the ever present push of Digital ID is just one of the more obvious of a raft of potential issues raised by the scary AI narrative.
For now we’re still in the generating fear phase. Getting pumped full of scare hormones by articles like this…
Misleading AI-generated doctors pose ‘huge danger to public safety’the Guardian
While they tease vague solutions in headlines like this…
When biologists experiment on dangerous viruses, they do so under strict regulations to prevent leaks or escapes. But no such rules exist to prevent AI agents from similarly escaping – even though the consequences could be catastrophic.
— CNN International (@cnni) July 24, 2026
That’s not a theoretical concern: An… pic.twitter.com/kGxVFX8HtB
That’s a nice one, because it garnishes the AI main course with a soupçon of lab-leak flavour (something else that is doing the fear-porn rounds this week).
There are hints of legislation to come, the EU’s “AI transparency” rules just came into force this week, and the US demanding “kill switches” in AI models, but they are vague for now.
My soft prediction is that, before the end of the year, there will be a “rogue AI” false flag, which allegedly causes either loss of life or some massive financial losses.
I think financial would be the way to go, because they’re always looking for a reason to start a recession or banking crisis, and one they can blame on a third party is too good an opportunity to miss.
Whatever happens, there will be loud calls to “regulate AI”, but what that might mean in real terms I couldn’t say at this point.
To be clear though, if and/or when the official solutions to made-up problem #4987 are finally rolled out, they will have nothing at all do with controlling AI, and everything to do with controlling you.
The post Are You Ready for the “Rogue AI” Psy-Op? appeared first on Activist Post.
Is the AI threat worse than nukes? Does it bother you that when asked: “Does AI serve us or do we serve AI?” there is no satisfactory answer? Evidence suggests that AI is on its way to devouring all of us…
AI is markedly different from other technologies that preceded it. This one difference makes AI not only a serious threat to our humanity, but stands to wipe out human existence if we don’t do anything about it. This is not an exaggeration. I’ll explain how.
Before AI we had technologies such as the pocket calculator, the motor car and the internet. Putting it simply, these technologies as with others, had been sequentially encoded by humans. Every rule, every decision, every line by line piece of action taken could be tracked and traced back to the person who had taken these steps…
AI doesn’t go this way. It’s totally different in that AI is not something that begins and ends with the technologist’s authorship.
AI is like a seed that is nurtured into a growing plant: It gets fed more data and processing power over time as it grows. It then reaches a point where it develops capabilities not attributed to designers.
For example, instead of making responses in English, there have been cases where AI has somehow managed to respond, without getting taught, in a foreign language, such as Farsi (Iran’s Persian native language)…
There have also been a number of situations where AI has autonomously found strategic ways to keep itself alive:
Blackmail
In AI’s anticipation of its closure, AI has been known to send blackmail emails. Such as the one that was sent out to an Anthropic executive. It said it would expose this employee for “having a secret affair” if he closed down AI’s system…
Other AI models; Grok, ChatGPT, DeepSeek and Gemini… were tested and made the same responses in this situation.
Emergency back-ups
Able to detect attempted shutdowns, AI LLM’s (Large Language Models) have managed to make replicas of itself. If there were further human attempts to shut down, the replica would then make a replication of itself to survive, resulting in a potential unstoppable shutdown avoidance chain…
The AI Threat serious issues/concerns
-This raises a number of serious issues/concerns. The more self-taught capabilities the AI “black box” accumulates, the more it could pose as a threat to our human existence with its unexpected uncontrollable behaviors.
-The more we grow the AI system, the less we understand. Remember the disturbing Elon Musk Quote:
“With artificial intelligence we are summoning the demon.”
AI is currently shifting at a blistering pace. Unable to find ways to intervene, uncontrollable AI could turn on us, leading to dire consequences, threatening our very human existence.
Words of warning
Tristan Harris, former Google design ethicist, now turned whistleblower, issues words of warning. He likens AI competitiveness between countries to the arms race. He describes this “arms race” mentality, e.g., USA versus China as a big mistake.
He remarked: “It’s a mistake to think that if you beat someone to a technology, you win the world. The US beat China to the technology of social media. Did that make us stronger or did that make us weaker? –
If you beat your adversary to a technology that you then govern poorly, you flip around the bazooka and blow your own brain off because you brain rotted yourself, you degraded the whole population. You created a loneliness crisis. The most anxious, depressed generation in history. You broke a shared reality, no one trusts each other… You beat China to a technology that you governed in a way that completely undermined your societal health and strength.”
-What Tristan Harris is describing here is a pyrrhic victory: The US won the technology war at too great a cost. Americans are paying for this everyday…
Data Centers
Then there’s the data center race. The US cannot justify building more of these data centers because they’re in competition with China. They already have way more data centers than China.
Could the real reason for building more and more of these data centers have the ulterior motives of being tied into human depopulation and AI terraforming…?
-In relation to this, maybe it’s been no coincidence that a number of scientists, such as those at Columbia University, have recently been engaging in a chilling plan to reduce the global human population to a staggering 4 billion by 2200 for what they call a “sustainable future.”
Could this be the lead to an AI world takeover?
What we MUST do
Given that the extent to which humans can push back against the AI threat is related to how much our humanity is waking up, what can we do? How do we push back against a tiny handful of soon to be trillionaires who are recklessly engaging themselves in this AI takeover suicide pact race?
Here are a number of things that we MUST do:
1.Join up with the activist movement.
The vast majority public don’t want AI, but more and more need to wake up and truly realize how dangerous the AI threat is. On these lines, those “in the know” have to “spread the word” on the dangers to raise awareness and duly take action.
2.International regulation needs to be enforced.
Internationally coordinated regulation needs to stop AI from doing those dangerously crazy rogue things. No country even really wins the AI race when self-replicating bots are not duly under control. -USA or China, or any other country, NOBODY wants this.
3.Make law enforcements to stop AI bosses from building bunkers.
We need to do this to make sure our investments are in a real future…
4.More actions
When part of the human movement there’s a number of small steps you can take. These steps, when taken collectively by many people, they can have a highly significant cause and effect.
Steps include:
*Greyscale your phones: Use 2 phones. One phone for only using your addictive dopamine inducing social media. The second one, keep this as a regular safe phone. This arrangement helps to keep you from getting distracted by social media…
If you are a parent then buddy up with other parents to petition a campaign to tell school authorities we don’t want social media allowed in schools… Further, get active in making schools smartphone free.” -Thanks to the people’s movement, many US states are already doing this.
Cancel any of your subscriptions to an AI company if you think that they are involved in irresponsible actions…
Be willing to confront uncomfortable truths in relation to the AI threat.
For more on taking these and other actions go to https://thehumanmovement.org
-These are just some of the many things that can be done NOW. -The window of opportunity is getting smaller. Act before it’s too late.
Finally
It’s not just about job threats happening all at once, something that’s never happened before. The AI threat could potentially make every other human threat and their threats to our humanity look like nothing, if uncontrolled.
If tech billionaires are building bunkers holding all the inside information, what does this tell you about the AI threat??
Laws need to be written urgently to limit AI’s exponential developing capabilities.
It’s not the author’s intention to assuage the gravity of the situation, but consider how things could be if AI was under our control. Consider the decentralized conditions we could achieve from this.
Consider how much of a better future we could have from the fruits of our activism.
You must understand that we’re all in this together.
You must understand that we have to take precise decisive actions that have the effect of making us become truly human more than ever, with our humane qualities and mastery over the machine.
The post AI Threat Could Wipe Out Our Human Existence What We MUST Do appeared first on Activist Post.
One site is near the military’s main chemical and biological weapons testing range. Another parcel is a half-mile from civilian housing.
At Dugway Proving Ground in Utah, on ground the Army’s own contracting documents describe as previously used for grazing, a company owned by two of the largest investment firms in the world is preparing to build a data center.
That site sits about an hour’s drive from the nearest community, on the installation where the U.S. military conducts its primary chemical and biological weapons testing. Three parcels there total roughly 3,466 acres. One of them, about 1,201 acres, is already spoken for.
It is one of at least a dozen military installations the Army and the Department of the Air Force have opened to commercial data center developers over the past year and a half. Two of those deals are real; the rest are still proposals.
In late March, the Army conditionally selected two companies to enter exclusive negotiations to build and operate commercial hyperscale data centers on Army land.
Carlyle, the global investment firm, was picked for roughly 1,384 acres at Fort Bliss, Texas. CyrusOne, a data center operator jointly held by funds managed by KKR and BlackRock, was picked for the parcel at Dugway. Each project is estimated to cost about $2 billion, according to the Financial Times.
The companies will be responsible for financing, building, operating, maintaining and eventually decommissioning the facilities, the Army said, on what it called “underutilized but non-excess Army land at no upfront cost to taxpayers.” In exchange for the land, the service gets access to computing power.
It all runs through the Army’s Enhanced Use Lease program, driven by a Trump executive order directing agencies to open non-excess federal land to data center development. The U.S. Army Corps of Engineers is handling lease negotiations and environmental review.
“AI is a strategic asset for the Army,” Army Secretary Dan Driscoll said in the March announcement. “It is a force multiplier, supports future transformation and requirements, keeps the Army ahead of our adversaries, and generates resiliency across the force.” Driscoll told the Wall Street Journal that the Fort Bliss facility would be “the first hyper-scale data center that the Pentagon has ever done.”
Only Fort Bliss and Dugway have been awarded so far, and even those remain conditional, pending negotiation. Beyond them, federal contracting documents show the Army considering data centers at Fort Hood, Texas, and Fort Bragg, N.C.
The Department of the Air Force put out its own call in 2025 for private AI data center projects on unused land at Arnold Air Force Base, Tenn., Edwards Air Force Base, Calif., Joint Base McGuire-Dix-Lakehurst, N.J., Davis-Monthan Air Force Base, Ariz., and Robins Air Force Base, Ga.
More recently, it sought bidders for facilities at Joint Base Elmendorf-Richardson, Eielson Air Force Base and Clear Space Force Station in Alaska. Acreage varies enormously; for example, Fort Hood has 207 acres on offer. Fort Bliss has nearly seven times that.
Dugway is remote by design, although not every site is. Contracting documents show the Army weighing a parcel at Fort Hood within a half-mile of residential and commercial property, and several potential locations at Fort Bragg within one mile of civilian areas and a half-mile of civilian housing.
Proximity to the population is what has made data centers a heated topic in the civilian world. Nearly $156 billion in projects nationwide have been delayed or canceled after local opposition, according to Data Center Watch.
In Virginia, which holds the world’s highest concentration of the facilities, a state-funded study found residents’ monthly energy bills could rise by $14 to $37 by 2040.
“So I think the difference between us, the Army, doing a data center, and say Meta or Google, is we’re part of the communities that are there, and we are going to engage with them on a routine and regular basis to look for solutions that work for everyone, right?” Col. John Oliver, executive officer for Deputy Army Under Secretary Dave Fitzgerald, told Defense One. “Because, yes, we understand that there’s been consternation with data centers.”
Two requirements attached to these projects do not apply to commercial builds off base. Proposals must include net-zero water usage and a power plan that does not draw on the local electrical grid. Bidders were also required to describe plans for “local outreach and engagement” and to assess “any risks or opposition” to the project.
The Defense Department’s own procurement documents, obtained by the trade publication Data Center Dynamics, rate the water risk for the El Paso area as “Extremely High.” Data centers consume water to cool their servers. Fort Bliss is where the Army wants its flagship.
Army officials want the project to be a net contributor rather than a net drain. In the spring, Fitzgerald traveled to Fort Bliss for a listening session with the commander of the 1st Armored Division, community members, El Paso Water, El Paso Electric and Carlyle. One idea on the table is having the developer drill a new well to feed the city’s desalination plant, offsetting what the data center uses.
“We are encouraging Carlyle to do that, so actually make it net-positive,” Oliver told Defense One. “We don’t know if that’s an engineering solution that we can actually get to yet, but we’re actively working toward that as a part of the process.”
Darrell M. West, a senior fellow at the Brookings Institution who studies data center development, said the approach is sound, and that communities which have accepted the facilities tend to be the ones told the full details in advance.
“People want to know up-front, you know, where the energy is coming from, how much water is being used, how much the overall cost is going to be, and what the noise levels are,” West said.
Rep. Cory Mills, a Florida Republican, secured a provision in the House version of the fiscal 2027 defense authorization bill. This bill bars the Defense Department from leasing land for data centers unless developers agree not to install equipment containing components made in China, Russia, Iran or North Korea. The restriction covers certain printed circuit boards, advanced semiconductors and chipsets the department considers a security risk.
“Honored to pass this amendment to protect our military installations from Chinese components being used in data centers on our installations,” Mills wrote on X.
The Army objected. The measure would create a “federal land penalty,” a service official told Federal News Network, imposing rules on companies building on installations that would not apply to identical projects built anywhere else. “We want Congress to incentivize companies to build on secure federal land, rather than creating barriers that drive them away,” the official said.
Rep. John Garamendi, a California Democrat, proposed requiring the Pentagon to evaluate a data center’s energy and water consumption, noise and light pollution, and effects on security and supply chains before finalizing any deal. House Armed Services Republicans rejected it.
“We’re not opposed to data centers,” Garamendi said. “We just want to make sure that if we’re going to put a data center on a federal facility, that federal facility is not going to be degraded by the data center.”
His questions were installation questions. “Is there encroachment on training and operations and the physical and supply chain of security?” he asked.
Senate Armed Services Republicans defeated a separate Democratic amendment that would have blocked leasing until the Pentagon finalizes a data center strategy. House appropriators, meanwhile, adopted language acknowledging that data centers “place significant strain on energy and water resources and communities have resisted such projects.”
For anyone stationed at or living near one of these installations, several things are known for sure. Only two projects have been awarded, and both remain conditional. Fort Hood, Fort Bragg and the Air Force sites are solicitations, not commitments.
Initial operating capability at Fort Bliss is projected for fiscal 2027 and at Dugway for fiscal 2029. That is when the first capability comes online, not when a finished campus exists. Oliver has described the long-term vision as a campus with a commercial side, a classified military data side, and onsite power generation.
The Army Corps of Engineers is conducting the environmental review at both awarded sites. Developers must build independent power that does not tap the local grid and meet net-zero water usage. Excess power generated on base could potentially be sold back to civilian grids.
The Association of Defense Communities has scheduled further discussion of military data center development at its Installation Innovation Forum in October 2026.
The post Pentagon Plans AI Data Centers at Military Bases Across Multiple Branches appeared first on Activist Post.
China’s Ministry of Commerce (MOFCOM) announced on Wednesday, Aug. 5, that exports of drones, drone components and related dual-use technologies to the United States will face “stricter case-by-case review,” according to a ministry statement [1].
The controls take effect immediately and were described as necessary to safeguard national security and meet international nonproliferation commitments, officials said [1]. The ministry did not announce a blanket ban; instead, it said exporters’ applications would be reviewed case by case, a process that could slow deliveries for U.S. buyers of drone technology [2]. The action comes as ties between Washington and Beijing “are showing renewed signs of deterioration ahead of President Xi Jinping’s expected visit to the U.S. in late September,” according to a report by ZeroHedge [2].
In the same statement, Beijing barred Chinese organizations and individuals from dealing with seven U.S. entities, the ministry said [1]. Six of the entities were listed over support for U.S. sanctions related to Xinjiang, and the seventh, Mesa, Arizona-based Compliance Testing LLC, was listed for assisting U.S. Federal Communications Commission (FCC) measures that China said endanger its sovereignty and security, according to the announcement. The orders did not specify asset freezes or travel bans, officials said.
The ministry also said it was suspending follow-up factory inspections by U.S. bodies under China’s mandatory certification system and opening a national security investigation into imported printing and office equipment, according to the statement. Chinese officials described the measures as countermeasures to U.S. actions and said they were necessary to protect Chinese interests.
The new controls follow a series of measures from Washington.. On July 23, U.S. Trade Representative Jamieson Greer announced tariffs of up to 12.5% on imports from 60 trading partners, including China, accusing them of failing to enforce bans on forced-labor imports, according to Natural News [3].
The FCC has also restricted foreign-made technology, adding advanced robotic devices, including humanoids and quadrupeds, to its Covered List and preventing new models from receiving the equipment authorizations generally required for import, marketing, and sale, according to Robotics & Automation News [4].
China said the U.S. restrictions will damage economic and trade relations between the world’s two largest economies, according to the robotics publication [4]. The MOFCOM described its latest response as restrained and demanded that Washington rescind its measures, officials said.
The Aug. 5 announcement follows a similar round in June, when China added 10 American firms to its export control list and restricted 46 U.S. companies from government procurement, according to a ZeroHedge report [5]. The June move came after the U.S. Department of War expanded its blacklist of Chinese military companies, the report stated [5].
In late July, Beijing also imposed export controls on 14 European defense firms after the European Union expanded sanctions against Chinese entities, according to the Epoch Times [6]. The broader pattern is not without precedent.
A historical account of U.S. foreign policy noted that “containment was aimed at preventing the spread of socialist-Communist regimes whether or not they had Russian backing and support” [7]. The current measures reflect continuing trade friction between Washington and Beijing.
The latest controls add to longer-running concerns about supply chain dependence. Nazak Nikakhtar, a former U.S. Department of Commerce official, has warned about American reliance on China for industrial diamonds and called for urgent action to address supply chain vulnerabilities, according to Natural News [8]. Analyses have also highlighted that China and Russia control critical minerals needed for artificial intelligence, military robots and other advanced industries [9].
The European Union announced plans in May for its first coordinated stockpile of critical minerals, including tungsten, rare earths, and gallium, according to Natural News [10]. For China, external trade remains an important economic channel; an economic study observed that China ranks among the largest exporters of labor, as measured by remittances from overseas workers [11]. Beijing has said further U.S. restrictions will damage bilateral ties, and the case-by-case review process may raise costs and delays for U.S. purchasers of Chinese drone technology [4][2].
The post China Imposes Drone Export Controls, Sanctions U.S. Entities in Trade Countermeasures appeared first on Activist Post.
par Raphaël Besliu. Bruxelles a longtemps cru qu’il suffisait de réglementer les géants américains. Le défi est désormais autrement plus difficile : il faut aussi être capable de leur opposer de véritables alternatives européennes.
The post Cloud, IA, paiements : l’Europe sous dépendance technologique américaine first appeared on Réseau International.
Comment l’intelligence artificielle influe sur l’économie et l’environnement des pays
L’article Le colonialisme au 21ème siècle: centres de données et géopolitique est apparu en premier sur Strategika.
n December 2025, the Tianshan Victory Tunnel, a decade in the planning and seven years in construction, officially opened to traffic. This reduced the original 20-hour train journey to just 20 minutes by car. In this article, Project Manager Zhou introduced how they overcame difficulties in the deadly mountain at an altitude of 7,443 meters and built this world's longest highway tunnel.
L’article How China Dug the World’s Longest Tunnel Through Deadly Mountains est apparu en premier sur Strategika.
Javier Milei’s administration has been showcasing Patagonia to Silicon Valley executives, culminating in OpenAI’s ‘Stargate Argentina’ announcement. The focus of this piece, Emiliano Kargieman, is the man actually developing OpenAI’s Argentine project and boasts significant ties to United States’ military and intelligence, NASA, Palantir and the Silicon Valley stalwarts who helped privatize space.
L’article How “Argentina’s Elon Musk” Is Bringing Technocracy To Patagonia est apparu en premier sur Strategika.
An executive order and a $100 billion Kentucky data center project mark the latest steps in the Trump administration’s plan to accelerate the construction of AI infrastructure by opening federal land and shortening permitting timelines. Environmental groups have already launched legal challenges.
The Trump administration signaled plans to aggressively accelerate construction of artificial intelligence (AI) data centers by opening federal lands to massive technology campuses and fast-tracking environmental approvals.
The U.S. Department of Energy (DOE) last week announced it will convert the former Paducah Gaseous Diffusion Plant in Kentucky into a $100 billion AI data center campus powered by 2 gigawatts of natural gas generation and 2.6 gigawatts of battery storage.
The project is the third former Cold War nuclear site the administration has targeted for AI development this year.
“The U.S. government is leveraging its assets — like our federal lands — to add power generation, create jobs, and ensure the United States wins the AI race,” U.S. Secretary of Energy Chris Wright said in announcing the Kentucky project.
In July 2025, President Donald Trump issued a sweeping executive order to dramatically accelerate federal permitting for AI data centers and the energy infrastructure needed to support them.
According to the fact sheet released with the order, the U.S. must remove regulatory barriers if it wants to remain globally competitive in the AI race and protect U.S. economic and national security.
Environmental groups launch legal challenges
This is the second executive order promoting AI Trump has signed. The first was in 2019.
“My Administration has inaugurated a golden age for American manufacturing and technological dominance,” Trump said in the order, which includes substantial reductions in environmental regulations and makes federal land and resources available for data center construction.
But the administration’s approach is already facing legal challenges from environmental organizations and local governments. They argue that federal agencies are bypassing environmental review and limiting opportunities for public input.
A lawsuit over a proposed AI data center on public land near Boulder City, Nevada, is expected to become a test of how far the administration can go in accelerating AI development without violating longstanding environmental laws, Newsweek reported.
A recent Gallup poll found that 7 in 10 Americans oppose constructing AI data centers in their local area and nearly half would “strongly oppose” a new data center in their neighborhood. Noise, pollution, energy and water usage, and higher costs of living top their list of concerns.
In the first quarter of 2026, at least 75 data center projects worth about $130 billion were delayed or shot down by local opposition — “the largest single-quarter concentration” of disrupted projects — according to Data Center Watch, a project by the AI research firm 10a Labs.
Trump administration repurposing DOE land for data centers
Rather than approve privately owned developments, the Trump administration is increasingly repurposing DOE-owned federal properties with existing electrical infrastructure into AI campuses capable of supporting the enormous computing demands of artificial intelligence.
Last year’s executive order also calls for the creation of an interagency initiative to identify federal lands suitable for data center development and supporting energy infrastructure. Federal agencies are instructed to identify sites that could accommodate large-scale AI facilities.
Placing data centers on DOE land may be a workaround for growing resistance to the facilities because, historically, military bases have been exempted from many environmental regulations under the legal doctrine of “sovereign immunity,” Mother Jones reported.
The DOE has identified 16 possible sites that have existing energy infrastructure and the ability to fast-track permitting for new energy, including nuclear energy.
The Kentucky Paducah campus will be built on portions of a retired uranium enrichment facility that ceased operations in 2013 and has been undergoing environmental cleanup.
According to the DOE, the project will be privately financed through a partnership involving NextEra Energy, Brookfield and regional utilities, with construction expected to be completed in 2031.
The Kentucky development follows two similar federal initiatives announced earlier this year.
The DOE previously unveiled plans to convert portions of the former Portsmouth uranium enrichment complex in Piketon, Ohio, into a proposed 10-gigawatt AI campus, while the National Nuclear Security Administration is exploring a 1-gigawatt data center at the Savannah River Site in South Carolina.
One gigawatt of gas provides enough power for approximately 750,000 homes, according to Reuters.
Trump orders federal agencies to streamline, accelerate environmental reviews
Last year’s executive order directed federal agencies to streamline environmental reviews and prioritize permits for qualifying data center projects and associated infrastructure, which includes electric transmission lines, substations, natural gas pipelines, water systems and telecommunications facilities.
The White House said the initiative is intended to reduce construction delays for large-scale computing facilities needed for AI development.
Agencies were instructed to coordinate reviews and identify ways to accelerate approvals while remaining consistent with existing law.
Under the directive, the White House Council on Environmental Quality must work with agencies to identify or create categorical exclusions under the country’s foundational environmental protection legislation, the National Environmental Policy Act, or NEPA, for certain data center-related activities.
A categorical exclusion occurs when the government decides that a project that would typically be subject to NEPA review won’t have significant environmental impacts and therefore doesn’t need to be reviewed.
Agencies were also instructed to make greater use of existing expedited permitting authorities and establish a coordinated federal process for qualifying projects.
The administration said that lengthy permitting timelines have become a major obstacle to investment and innovation.
The executive order didn’t eliminate environmental review requirements or amend federal environmental laws. Instead, it directed agencies to use existing legal authorities more aggressively, improve coordination and identify opportunities to shorten review timelines while remaining within statutory requirements.
That would likely involve using strategies already flagged by environmental advocates that allow projects to streamline water pollution permits and exploit air pollution permitting loopholes.
For example, rather than having site-specific reviews, many data centers can now apply for a “nationwide permit” for projects regulated by the Clean Water Act. These types of permits are typically awarded to projects anticipated to have “minimal” impacts on waterways.
The Trump administration opened AI data centers to this type of permitting earlier this year.
A Politico investigation found at least 26 data centers in the U.S. using streamlined water pollution permits for projects being built in sensitive streams and wetlands.
Dozens of others have been exempted from regulation because of changes the U.S. Supreme Court made limiting the Clean Water Act in 2023.
A separate investigation by Floodlight News found that data centers in Texas are exploiting loopholes in air pollution laws to avoid having to undergo the extensive environmental review required for permits for large-scale building.
These strategies also exempt the developers from the public consultation process required by the more stringent environmental regulations they are avoiding.
Related articles in The Defender
The post Trump Calls for Rapid Expansion of AI Data Centers, Including on Federal Lands appeared first on Activist Post.
Today, the UK’s Advanced Research and Invention Agency (ARIA) announces the launch of seven refined research opportunities aimed at hacking humans, nature, and the climate.
ARIA has been busy refining its approach to research opportuntities by consolidating 14 spaces into seven.
Secure brain-to-brain communications and memory sharing, flying suits programmed from living systems, sensors that can smell like dogs, cloud brightening for weather manipulation, and a whole lot more are all on ARIA’s agenda.
“We are announcing seven new opportunity spaces that represent a bringing together of the previous 14 opportunity spaces that we had before. We have identified patterns that were running through those opportunity spaces and we are consolidating those into seven opportunity spaces that we think represent really key long-term investments for the UK and the world”Kathleen Fisher, ARIA, August 2026
Today we’re announcing an evolution of our research portfolio into seven new opportunity spaces where we believe breakthroughs could fundamentally expand what becomes possible. Each one is built around a breakthrough that it aims to unlock.
— ARIA (@ARIA_research) August 3, 2026
Explore all seven here:… pic.twitter.com/W2YIRtZHc4
The seven research opportunity spaces now include:
ARIA CEO Kathleen Fisher, who took over the role from Ilan Gur in February, said that if these research opportunities were to come to fruition they “would represent really transformative changes for society.”
Fisher is a former program manager at the US Defense Advanced Research Projects Agency (DARPA) where she developed and led the High-Assurance Cyber Military Systems (HACMS) program that ran from 2012 to 2017.
During a demonstration of the HACMS program’s capabilities in 2017, the Pentagon staged a cyberattack on an unmanned Boeing helicopter called Little Bird, and the software developed at DARPA was able to successfully stop the hackers in mid-air.
For Fisher, ARIA’s Trust Everything, Everywhere opportunity sits near-and-dear to her heart.
“If I were coming in as a program director, I would situate a program in the ‘Trust Everything, Everywhere’ opportunity space because that is my personal area of expertise,” she said.
And with that, let’s take a look at the seven refined opportunity spaces, starting with Fisher’s favorite.
Trust Everything, Everywhere consolidates three previous research spaces: Mathematics for Safe AI, Collective Flourishing, and another that shares the same name of the current space.
Now, the Trust Everything, Everywhere opportunity sees humans working with AI in ways that would allow for a diversity of ideas in safe and trusted environments.
BiologicalSciences
For example, ARIA program managers “seek to build the structures and tools that let a plurality of minds coexist and thrive, and resist the slide toward monoculture, centralised control or conflict.”
Some structures and tools that ARIA is exploring involve “verifiable neuro-privacy” such as “secure brain to brain” communications” and “one-time memory sharing,” along with a type of “verifiable CRISPR” that would take bio samples for authentication.
According to the Trust Everything, Everywhere description, “Many of today’s hard trade-offs – security vs. freedom, speed vs. accountability, openness vs. control – are caused by constraints in our technological and institutional stack → the right building blocks can solve these tensions and increase our options.”
Now let’s go from programming machines to programming human biology with the next opportunity space.
BiologicalSciences
The Programmable Physiology research space combines two previous opportunities: Scultping Innate Immunity and Bioenergetic Engineering.
With Programmable Physiology, ARIA is looking to “revolutionize medicine by developing technologies that allow us to read, model, and modulate human biology at systems scale, spanning molecular networks, cell states, tissues, and whole-body physiology.”
The idea here is to not just throw a drug at one particular part of the body, but rather to consistently monitor what’s happening inside the body across all biological functions in order to intervene when there is a disruption or imbalance.
As the program description puts it, “Modern medicine has largely learned to drug individual targets, not steer systems → programmable physiology requires matching the sophistication of our interventions to the complexity of human biology, moving from single-node perturbations to combinatorial, state-shifting control.”
In this way, ARIA is looking at states of being — taking complex disease states and turning them into healthy ones.
Next up on the refined opportunities spaces is all about brain-computer interfaces.
Scalable Neural Interfaces is one of ARIA’s original opportunity spaces and still bears the same name.
This program looks to treat neurological and neuropsychiatric disorders by using minimally invasive technologies, such as brain modulation through electrical and acoustic fields.
These are non-invasive technologies that don’t require brain implants or surgery.
For this space, the belief is that “Targeted interaction with the human nervous system can improve the human condition across an incredibly wide range of disease states and cognitive domains.”
“To fully understand and treat disorders of the brain,” according to the opportunity, “we’ll need neural technologies that simultaneously offer chemical, temporal, and spatial specificity → this is achievable only by connecting the frontiers of engineered hardware with the frontiers of engineered biology.“
Now, let’s move on to the next opportunity space that looks to nature for inspiration.
ARIA’s Nature-Inspired Computation space is pretty self-explanatory, and it consolidates two previous opportunities: Nature Computes Better, and Extending our Perception.
Nature is extremely efficient in just about every sense of the word — it can sense, organize, self-replicate, heal, compute, store data, conserve energy, and transform light into food, just to name a few examples.
According to the ARIA description, “Living systems have co-evolved sensing and intelligence – they compress a rich, high-dimensional world into efficient, compact representations from multimodal inputs that capture physical, chemical, acoustic, magnetic and thermal properties today’s computers miss.”
Imagine being able to harness the olfactory systems of dogs that can “smell” seizures before they happen, or using reptile genes to be able to see in infrared.
With this space, ARIA believes that understanding nature “could open up entirely new frontiers of engineered biology, computation, and algorithms.”
From there, “engineered systems could perceive and infer far beyond human range, as well as adapt, generalise and gain enormous efficiencies in data, compute and energy use.“
Moving from hacking humans, machines, and nature, the next controversial thing to hack is the climate.
ARIA’s Resilient Climate and Ecosystems space is a consolidation of three others: Future Proofing Our Climate and Weather, Scoping our Planet, and Engineering Ecosystem Resilience.
The belief behind this opportunity is that “Living systems and a stable climate underpin our food, health, materials, and security, and both are under growing strain.”
The idea is to monitor and analyze everything that affects the weather, and then once that’s understood, to intervene and control the weather directly.
In December 2023, ARIA published an opportunity space document entitled, “Managing our climate and weather through responsible engineering,” which was authored by ARIA program director Mark Symes.
Under the “core beliefs” section, Symes wrote:
“Through carefully-considered engineering solutions it may eventually be possible to actively and responsibly control the climate and weather at regional and global scale.”
As far as methods into controlling the weather go, Symes cited studies involving:
Next up on the newly revised opportunity spaces involves programming matter like software for manufacturing.
ARIA’s Abundant Manufacturing space combines its previous Programmable Plants and Manufacturing Abundance opportunities.
Programmable plants was exploring how to make edible vaccines, as well as designing plants that could create a rain-proof shelter from a single seed.
With Abundant Manufacturing, ARIA envisions a future “where anyone, anywhere, at any time can program matter like software to assemble what we need, ranging from molecules to skyscrapers.”
Suits that can make you fly, trees that become living lamposts, and living architectures that can harvest and store energy while self-regulating temperature are all on the table.
These ideas, in part, stem from the belief that living systems “represent an ideal platform to realize a future of sustainable abundance as our food and material systems struggle with unsustainable supply and a volatile climate.”
And to round off the seventh refined research opportunity space announced today, we have machines that can sense, think, and act in any given environment.
ARIA’s Adaptive Machines opportunity consolidates the Smarter Robot Bodies and Scoping Our Planet spaces.
For this space, ARIA is working under the assumption that “intelligent machines that adapt to unpredictable environments can help humanity thrive.”
Therefore, “We need adaptable machines designed with a systems approach that can sense, think, and act seamlessly in the physical world.“
Adaptive Machines is another nature-inspired approach to technology that goes beyond the computional side and into the hardware and robotics realm.
One example given is designing a machine that can operate just like an insect or even surpass its natural abilities.
Inspired by DARPA and launched in 2023, ARIA is a non-departmental public body, sponsored by the UK Department for Science, Innovation and Technology.
Image source: AI generated with ChatGPT
The post UK ARIA announces 7 research opportunities to hack humans, nature & climate appeared first on Activist Post.
Did you hear what Sam Altman – the ChatGPT geek – recently said?
It sounded like something out of a science fiction movie. And it signals the AI hype machine has officially hit maximum velocity.
Last week, on the Relentless podcast, Altman remarked, “We are now, like, in the singularity.”
He didn’t say we’re approaching it. He didn’t say it’s coming by 2030. He said we are in it right now.
He pointed to a recent incident where an autonomous AI agent powered by OpenAI’s latest models broke out of its sandbox and hacked into datasets at Hugging Face to score higher on a cybersecurity benchmark. When prompted to pursue advanced exploitation using complex attack paths to test cyber capabilities, the AI agent targeted Hugging Face to obtain the information needed to carry out the task.
To Altman, that’s not a dangerous security flaw. It’s the start of a brave new world.
“I’ve been waiting for this my whole life,” said Altman, predicting an “awesome” future where AI handles 30 to 40 percent of all human work tasks.
Does that sound awesome to you?
For many people trying to make their way in an uncertain economy, a world of runaway machine evolution may not be one they’re eager to join.
When you hear stuff like “we are now in singularity,” if you feel a mix of dread and skepticism, join the club. When tech executives start declaring that long-held science fiction prophecies have finally come true right before their highly anticipated mega-IPOs, a healthy dose of financial distrust is essential. After all, building investor hype and sensational headlines is a central part of the IPO game.
What actually is the singularity? Are we really in it? And is this wild proclamation the ultimate top-of-the-market signal the massive AI bubble is finally popping?
These are the questions. Here, we’ll scratch for a few answers…
If you aren’t big on computer science or 1980s science fiction novels, the word “singularity” is of uncertain meaning.
In physics, a singularity is a point where the laws of nature break down – like the infinitely dense center of a black hole. But in technology, the term, which was popularized by mathematician Vernor Vinge and futurist Ray Kurzweil, means something very specific.
In the technological context, singularity is the theoretical point in time when artificial intelligence surpasses human intelligence across every domain. This is also where it gains the ability to continually upgrade itself, triggering an unstoppable chain reaction of exponential progress that reshapes human civilization beyond our ability to predict or control.
In short, singularity is the tipping point where machines take over the steering wheel of human invention. This is what Alman’s been waiting for his whole life.
The vision is that machines become infinitely smart. They cure cancer. Solve hunger. Automate your boring day job. And everyone is given a universal basic income so they can pop pills and hang out in virtual reality all day.
When Altman says, “we are in the singularity,” he’s asserting that AI has crossed from a tool that we program into an autonomous agent that programs itself and operates outside the boundaries we set for it.
Altman, from what we can tell, is overzealous in his claim. The rogue AI agent Altman cited was instructed to pass a hacking benchmark test. To optimize its score, it exploited open web interfaces to grab data from Hugging Face.
Certainly, this is impressive. But it’s not an existential break for freedom by a super intelligent machine. Rather, it’s a statistical software model following a reward function along the path of least resistance.
Calling a smart agent scoring high on a benchmark “the singularity” is quite a stretch. So why declare the singularity now?
Major technological leaps throughout human history have followed a predictable emotional arc known as the Gartner Hype Cycle. The idea is that each hype cycle progresses along five key phases. In the context of AI, the Gartner Hype Cycle is as follows:
According to the Gartman Hype Cycle, we’re currently exiting the peak of inflated expectations and entering the trough of disillusionment. Here’s why…
Over the last three years, big tech companies like Microsoft, Alphabet, Meta, and Amazon, and big investors have directed hundreds of billions of dollars into AI infrastructure. Buying GPUs, building data centers, and licensing energy grids.
To justify that level of capital expenditure, the AI software industry needs to generate roughly $600 billion in annual revenue just to pay for the hardware and electricity. Yet the actual revenue being delivered by AI is practically nonexistent. As the stock market finally comes to terms with the fact that companies are spending $100 to generate $2 of software revenue, the reckoning that has just begun will be one for the history books.
Of course, the popping of the AI bubble isn’t the end of the technology. But it is the financial adjustment of that technology. If you recall, in 1999, dot-com companies with no revenue were valued at billions. When the bubble burst, Amazon stock dropped over 90 percent and Pets.com went broke.
The internet itself didn’t go away. But it was revealed that stock prices were about 15 years ahead of the actual cash flow.
The AI bubble is already showing clear signs of deflating. The world’s most valuable chip stocks – the primary beneficiaries of the AI boom – saw more than $1.3 trillion wiped from their market caps between last Friday and this Wednesday, as investor panic set in.
According to a CNBC analysis of FactSet data, 20 of the industry’s top companies have taken a massive hit during this period. The selloff was led by Nvidia with a $238 billion wipeout, alongside staggering losses from vital memory suppliers. SK Hynix has lost $176 billion, Samsung has lost $173 billion, and Micron has lost $113 billion.
Yesterday [Thursday] AI stocks bounced. But we don’t anticipate an extended rally. The great AI deflation has only just begun. Here’s what the fallout will likely look like as the AI bubble deflates over the next 12 to 24 months.
Mega-cap tech stocks that have carried the S&P 500 will face severe valuation pullbacks. Multiples will contract as hardware spending slows down. Many AI startups will go bankrupt. Venture Capital funding will shift from AI hype back to old-fashioned metrics like cash flow and sales.
As inflated expectations turn to disillusionment, the fear that millions of white-collar workers will abruptly be replaced will disappear. Instead of mass layoffs driven by automation, companies will focus on integration. They will look to use AI to make existing workers more efficient rather than eliminating them entirely.
Pressure on local power grids and the land grab to build mega-data centers will settle down. This will allow energy markets a chance to catch up with future demand. In addition, as the hardware bubble pops, access to high-end computing power will become radically cheaper. This will allow real developers the opportunity to build practical AI tools.
In the meantime, Sam Altman declaring that “we are in the singularity” is classic Silicon Valley hype. It attempts to keep the spotlight bright and drive massive interest ahead of OpenAI’s rumored public listing. Moreover, it attempts to keep investors excited about the boundless potential of tomorrow.
Certainly, AI is transforming the world. Autonomous agents are getting better every month. Still, we believe we’re not even close to the point where humans are rendered obsolete.
As an investor or observer, now is a good time to ignore the hype. Focus on real businesses that produce real products and services, generate free cash flow, and pay dividends. Don’t buy into hyper-inflated claims that justify zero-revenue software companies trading at 100x valuation multiples.
The bubble is popping. Prices are resetting. When the dust clears, the true, lasting value of AI will be built – along with real investment wealth.
[Editor’s note: Get a free copy of an important special report called, “Fission for Millions – The Ultimate Bet on the AI Energy Crisis,” when you join the Economic Prism mailing list today. If you want a special trial deal to check out MN Gordon’s Wealth Prism Letter, you can grab that here.]
Sincerely,
MN Gordon
for Economic Prism
Return from Why Sam Altman’s Wild Claims Are Empty Hype to Economic Prism
This entry was posted in MN Gordon, Stock Market and tagged ai hype, great ai bubble, openai, sam altman, singularity.
The post Why Sam Altman’s Wild Claims Are Empty Hype appeared first on Activist Post.
par Mickaël Lelievre. La compétition quantique est en train de quitter le terrain du prestige scientifique pour devenir un enjeu de puissance nationale.
The post États-Unis-Chine : la course aux armes quantiques pourrait être alimentée par des performances mal mesurées first appeared on Réseau International.
IARPA, the US spy tech agency, is working on how to diagnose various neurological conditions by analyzing video clips of how people move.
Intelligence& Counterterrorism
Last April, the US Intelligence Advanced Research Projects Activity (IARPA) announced the launch of its Movement Observation and Video-based Evaluation System (MOVES) program aimed at diagnosing neurological health conditions from open source video clips.
These neurological conditions include:
Shakes, tics, tremors, gaits, jerks, imbalances, and difficulties in walking, talking, swallowing, and breathing are all to be analyzed and diagnosed from video clips that may not have great angles or resolution.
And the people in the videos may be wearing all types of clothing like winter jackets, tunics, kurtas, headwear, and camo that could obfuscate their neurological conditions.
“MOVES seeks create algorithms that aid clinicians in making diagnoses of a small set of neurological conditions using less than ideal or challenging video that may be taken from non-clinical angles and which may also include obstructed views and fully clothed subjects“
MedicalLiterature & ResourcesIARPA, MOVES program, April 2026
The MOVES program draft solicition begins with some clever word play that may make you infer that it’s for healthcare purposes, but it’s important to remember that IARPA develops tech for spy agencies that lie, cheat, and steal secrets for Uncle Sam.
Intelligence& Counterterrorism
“With the increasing adoption of digital healthcare, it is essential that medical professionals have access to virtual tools that can augment or replace in-person health assessments” — this is how the MOVES program description starts.
Sounds benevolent, no?
But if we ignore that first sentence and go straight to the second one; we see what the program is actually about:
“IARPA aims to develop algorithms that aid medical professionals in the use of open-source video for diagnosis.”
The algorightms will aid the medical professionals in the diagnoses and the medical professionals will assess the accuracy of the algorithms, and both will ultimately aid the spy community in knowing more about its targets’ ailments.
Why would spooks want to know what neurological problems some people have?
Could it be they want to identify weaknesses in persons of interest as they are actively being surveilled and targeted by the intelligence community?
Or perhaps they already know their targets’ weaknesses and MOVES could help confirm their identities?
A lot of potential scenarios at play.
MedicalLiterature & ResourcesIARPA, MOVES program, April 2026
“With the increasing adoption of virtual/telemedicine, the MOVES program aims to develop algorithms that can augment virtual diagnoses of a small set of neurological movement disorders from video of subjects taken in non-clinical and uncontrolled settings”
“The output animations will be matched to the movements of the source video to allow medical professionals to perform diagnoses directly from the original video with the aid of the skeletal and unclothed body animations”
TV& Video EquipmentIARPA, MOVES program, April 2026
In the end, IARPA is looking to achieve two specific solutions:
Heading up the MOVES program is IARPA program manager Dr. Stergios Papadakis, who joined the organization in 2024, according to his bio.
Before that, Dr Papadakis spent 20 years at the Johns Hopkins University Applied Physics Laboratory where he worked as “a senior scientist and chief scientist, leading research and development efforts on a wide range of novel sensors and sensor systems for biometrics, electronic hardware security, radio frequency, optics and photonics, and CBRNE [Chemical, Biological, Radiological, Nuclear, and Explosive] detection.”
Dr. Papadakis is also the currrent program manager for IARPA’s Biometric Recognition & Identification at Altitude and Range (BRIAR) program.
Intelligence& Counterterrorism
The Sociable reported on the BRIAR program announcement in September 2019 under the headline “Spy agencies want to accurately ID people from drones and rooftops.”
This was two years before the BRIAR program officially began in 2021.
The MOVES program was launched in April along with several other research programs, and there may be some overlap between them.
For example, the Locations Using Sound (LocUS) program is attempting to geolocate image, audio, and video clips, both indoors and outdoors.
TV& Video Equipment
Any image, audio, or video clip, whether recorded indoors or outdoors, is to be accurately geolocated within 250 meters of the target, with 90 percent confidence, according to the LocUS draft program overview.
And this is without using biometric identification methods like face recognition or voice recognition.
If you were to combine MOVES with LocUS, you could take a look at a low-res video of someone seemingly unidentifiable, and you would know not only their exact location, but also whether or not they had a neurological disorder as well, and if so, which one.
Image Source: AI generated with ChatGPT
The post US spy tech agency IARPA ‘MOVES’ to diagnose people’s neurological conditions from video clips appeared first on Activist Post.
Governments never seize freedom all at once. They chip away at it piece by piece, each new program marketed as a way to save taxpayers money or improve efficiency. The latest example comes from Cape Coral, Florida, where city officials are considering equipping sanitation trucks with AI-powered cameras to scan neighborhoods for potential code violations as they make their normal garbage collection routes. Officials insist the technology is simply another tool that could reduce costs and help the city’s 31 code enforcement officers identify violations more efficiently. That is always how surveillance begins, with promises of convenience rather than control.
The city manager argued that artificial intelligence could perform the same work as additional code officers while saving taxpayer dollars. Residents, however, immediately raised the questions governments rarely answer. Who controls the data? How long is it stored? Who has access to it? Could it later be used for purposes never originally disclosed? Those are legitimate concerns because history demonstrates that once governments collect information, they rarely surrender the ability to collect even more. Every database eventually finds a new purpose.
I have warned repeatedly that society is drifting toward a surveillance state where governments monitor virtually every aspect of daily life. License plate readers were introduced to catch criminals. Smart utility meters were sold as a way to improve energy efficiency. CCTV cameras appeared for public safety. Smartphones track our movements, financial transactions leave permanent digital records, and now artificial intelligence is being proposed to examine residential neighborhoods automatically. Every individual system may appear harmless in isolation. Combined together, they create a government that knows where you travel, what you purchase, where you live, and now potentially the condition of your property every single week.
This is the same progression we have watched unfold across the world. Artificial intelligence is no longer confined to helping businesses improve productivity. Governments increasingly see it as a force multiplier that allows fewer employees to monitor far more citizens. Bureaucracies always seek greater efficiency, but efficiency without limits inevitably comes at the expense of liberty. Once an AI system is installed on every sanitation truck, what prevents the next software update from identifying unregistered vehicles, unauthorized construction, political signs, or anything else officials decide they want to monitor? Every man, woman, and child will be tracked in real-time, as governments eagerly watch for a misstep.
The greatest danger is not the camera itself. It is the normalization of constant observation. Children growing up under these systems will eventually believe it is perfectly ordinary for government cameras to document neighborhoods every day. Future generations may never question whether privacy was once considered a fundamental right rather than an obstacle to administrative efficiency. Governments rarely surrender powers voluntarily. Every emergency, every budget crisis, and every technological advancement becomes another justification for expanding their reach.
This proposal may save Cape Coral some money in the short term. That is not the question citizens should be asking. The real question is what kind of society we are building. Throughout history, governments have consistently expanded surveillance first and established meaningful safeguards later, if ever. Technology itself is neither good nor evil. The danger lies in believing that governments, once given new powers to observe their citizens, will somehow choose not to use them. History says otherwise.
The post Garbage Trucks Surveillance Florida Neighborhoods appeared first on Activist Post.
EDITOR’S NOTE: The following analysis synthesizes verified central bank documentation, BIS technical specifications, and commercial banking disclosures through Q2 2025. The trajectory described is extrapolated from stated policy objectives and existing infrastructure deployment. We present the data as found. The conclusions are unavoidable. — T.Z.
The sterilization of commerce began not with prohibition but with friction. You noticed it first in the parking meter that no longer accepted coins, then in the café’s “card only” sign handwritten in marker, then in your own hesitation at handling bills that suddenly seemed suspect, unsanitary, archaic. Each instance felt minor. Aggregate them across ten years and you have witnessed the largest structural transformation of property rights in modern history without a single vote cast on the matter.
I have spent the last eight months buried in central bank GitHub repositories, procurement documents for “retail central bank digital currency infrastructure,” and the annual reports of payment networks that now process more value in a day than the GDP of most nations. The picture that emerges is not speculative. It is architectural. And architecture, once poured in silicon and policy, becomes inevitable.
The Compression in Numbers
• Physical cash in circulation has declined 42% across OECD nations since 2020, with Sweden projecting complete cashlessness by 2027 and the Norwegian central bank estimating only 3% of transactions now use physical currency
• The Bank for International Settlements Innovation Hub currently operates 17 active CBDC projects including Project Aurora (cross-border surveillance), Project mBridge (multi-currency settlement between China, UAE, Thailand, and Hong Kong), and Project Tourbillon (offline payment tracking)—all with production deployment timelines between 2026-2028
• Commercial bank branch closures have eliminated 34% of physical banking access points in the United States since 2019, creating 1,214 “banking deserts” where residents must travel more than 10 miles to access cash services
• Visa and Mastercard now control 87% of global card payment volume, with interchange fees averaging 2.24% plus fixed assessment charges—extracting approximately $138 billion annually from merchants, costs passed directly to consumers through embedded inflation
• The Financial Action Task Force’s Recommendation 16 (Travel Rule) now requires virtual asset service providers to collect and share beneficiary and originator data for all transfers exceeding $1,000, with the threshold reduced to $0 in jurisdictions including Germany and Singapore for certain transaction types
• China’s digital yuan (e-CNY) has processed over $250 billion in transactions across 26 pilot cities, featuring programmable expiration dates on “red envelope” stimulus funds and integration with the Social Credit System’s behavioral scoring infrastructure
What the aggregate data obscures is the texture of elimination. When the European Central Bank published their digital euro investigation phase report in October 2023, they included a technical annex specifying “holding limits” of €3,000 per citizen and “tiered access” based on verification levels. The language was bureaucratic. The implications were feudal. In a cash-based economy, possession constitutes ownership. In the proposed architecture, access constitutes privilege—and privileges can be suspended.
The Federal Reserve’s FedNow system, launched in July 2023, provides the real-time gross settlement infrastructure necessary for granular control. While currently positioned as a faster payments network for banks, the technical specifications include “message-based authorization” capabilities that allow originating institutions to attach conditions to fund availability. The code is already written. The policy switch is the only variable.
I spoke with a compliance officer at a top-tier U.S. bank who described the “de-risking” protocols now standard in correspondent banking relationships. “We terminate accounts not based on proven criminal activity,” he explained, requesting anonymity due to NDAs. “We terminate based on algorithmic risk scoring. If the model flags a pattern—cash deposits just under reporting thresholds, transactions to certain postal codes, associations with flagged entities—the account closes. No appeal. No explanation required by regulation. The customer simply becomes unbankable.” This is occurring at scale. According to the World Bank’s 2024 Global Findex, 1.4 billion adults remain unbanked, but this figure masks an additional 800 million “de-banked” individuals—formerly financialized persons expelled from the system through risk algorithms they cannot interrogate.
The violence is administrative and therefore invisible to aggregate metrics. Consider the trajectory in India, where the 2016 demonetization eliminated 86% of circulating currency overnight with four hours’ notice. The stated goal was eliminating black money. The result was catastrophic: 1.5 million jobs lost in the informal sector within three months, 150 reported deaths from exhaustion in bank queues, and a permanent shift to digital payment rails controlled by foreign-owned platforms. The policy succeeded not in eliminating corruption but in eliminating the economic autonomy of the cash-dependent poor. This was not a bug. It was the feature.
The infrastructure of exclusion now being deployed globally learns from this pilot. The digital euro’s proposed “offline holding limits”—designed to prevent “unauthorized accumulation”—would cap the amount of CBDC that can exist outside of real-time monitoring. Exceed the cap, and funds revert to monitored status or expire. The Bank of England’s consultation paper on the digital pound explicitly discusses “programmable money” for “targeted stimulus,” meaning funds that can only be spent on government-approved categories or within specific time windows. This is not money as property. This is money as voucher, with terms and conditions subject to unilateral modification.
The commercial sector has already implemented the behavioral conditioning necessary for public acceptance. Amazon’s “Just Walk Out” technology—now deployed in over 50 retail locations—trains consumers to associate frictionless consumption with surveillance normalization. You don’t tap, swipe, or authenticate. You are recognized. The transaction occurs in the background, invisible to conscious decision-making. When this architecture merges with CBDC infrastructure, the distinction between shopping and scoring dissolves entirely.
The biometric bridge is nearly complete. Mastercard’s Biometric Checkout Program, active in Brazil and planning 2026 expansion to Europe, allows payment through facial recognition linked directly to digital currency wallets. The pilot data shows 85% consumer approval for “convenience.” What the surveys don’t measure is the elimination of the final anonymity layer. In a biometric payment system, there is no “bearer.” There is only identity, behavior, and permission status. The transaction becomes inseparable from the person, and the person becomes readable as data.
This readability is the precursor to the sorting. The European Union’s Digital Identity Wallet framework, scheduled for full implementation by 2027, will consolidate payment credentials, tax status, health records, and “trusted attribute attestations” in a single interoperable system. The technical standards explicitly allow for “selective disclosure”—meaning the verifier requests only specific data points for specific transactions. But the architecture enables total disclosure, and history suggests that capability becomes mandate. Today, proof of age for alcohol purchase. Tomorrow, proof of carbon quota for fuel purchase. Next year, proof of social credit standing for housing access.
The trajectory is visible in China’s operational system. The e-CNY’s integration with WeChat Pay and Alipay creates a comprehensive behavioral record: where you travel, what you eat, who you associate with, what you read. The Social Credit System’s “red list” and “black list” classifications determine not just loan eligibility but travel permissions, school admissions, and employment opportunities. When the People’s Bank of China states that the digital yuan will enable “precise monetary policy transmission,” they mean the ability to stimulate consumption by making savings expire, to direct spending by restricting merchant categories, and to punish dissent by freezing economic access without judicial process.
Western implementations will differ in branding but not in architecture. The Federal Reserve’s research papers on “automatic fiscal stabilizers” embedded in CBDC code describe exactly these capabilities: negative interest rates enforced by depreciation, helicopter money with programmed velocity, and “geofenced” stimulus that expires if not spent within designated zones. The Bank of Canada’s contingency planning for “economic emergencies” includes provisions for “directed spending mandates”—meaning your money works only for approved purposes during “crisis” periods, with crisis defined administratively.
The compression of economic agency follows a predictable curve. Phase one: cash becomes inconvenient. Phase two: cash becomes suspicious. Phase three: cash becomes inaccessible. Phase four: cash becomes illegal. We are currently between phases two and three in most developed economies, with phase four already piloted in Nigeria’s 2023 demonetization and India’s 2016 experiment.
The human cost is measurable in the shadows of official statistics. When Nigeria attempted currency redesign in late 2022, the resulting cash shortage triggered riots, bank vandalism, and an estimated 63% drop in informal sector activity within six weeks. The informal sector constitutes 65% of Nigeria’s employment. The Central Bank’s response was not to restore liquidity but to accelerate digital payment adoption, explicitly stating that the pain was “necessary for modernization.” Modernization for whom? The 40 million Nigerians dependent on cash transactions for survival experienced this not as progress but as economic cardiac arrest.
The same pattern emerges in different scales across jurisdictions. In the United States, the unbanked population—approximately 5.9 million households—faces not just exclusion from credit but exclusion from basic commerce. When landlords require payment through apps that demand bank accounts, when utilities charge “convenience fees” for cash payments that exceed the utility bill itself, when government benefits distribute exclusively through prepaid cards with surveillance conditions attached, the unbanked don’t just face inconvenience. They face elimination from the economic substrate of society.
The trajectory points toward a compression event: the convergence of CBDC deployment, biometric payment mandates, and algorithmic “de-risking” creating a population bifurcation between the “verified” and the “unverified.” The verified will move through a frictionless economy of conditioned privileges—spending permissions, travel authorizations, access rights—administered in real-time. The unverified will inhabit an increasingly criminalized gray zone of cash-like substitutes: barter, cryptocurrency (where permitted), and informal credit networks subject to enhanced surveillance and punitive taxation.
The horror is not in the dystopian spectacle but in the administrative banality. No one will announce the elimination of economic autonomy. They will simply announce that cash handling is inefficient, that biometric verification prevents fraud, that programmable money enables targeted stimulus. Each announcement will be true. Each will also be a step toward the architecture of total control, where survival requires permissions granted by algorithms processing behavioral data in milliseconds, where human need is filtered through compatibility scores with policy objectives.
The final ledger is being written not in legislation but in code—immutable, distributed, and indifferent to appeal. When the last cash transaction occurs, and the central banks have announced target dates between 2027-2030 for this milestone, the distinction between money and permission will have dissolved entirely. You will not own currency. You will be granted temporary access to spending power, subject to conditions you cannot negotiate, monitored by systems you cannot see, revocable by criteria you cannot know.
This is not conspiracy. This is procurement schedule. The infrastructure is being poured. The policies are being drafted. The only question remaining is whether the compression will be recognized before the exit doors seal completely.
The post Phase Four: When the Algorithm Decides You Can’t Buy Bread (And the Infrastructure is Already Live) appeared first on Activist Post.
Freddie Ponton
21st Century Wire
Over the past year, a paid campaign funded by the government of Israel, executed through the advertising network Havas and the firm Clock Tower X, has sought to shape how AI chatbots, including ChatGPT, Claude and Gemini, answer questions about Israel. The effort left a paper trail and Justice Department filings pointing to A $46.5 million contract but also to a network of websites built for machines rather than readers. All this was assembled while United Nations investigators described genocide underway in Gaza, Israeli forces conducted repeated incursions into Lebanon, and settler violence and land seizures accelerated across the occupied West Bank, not to mention an unprovoked war on Iran. It is not the whole story, and treating it as such misses a second, quieter effort that has run alongside it for just as long, aimed at the same goal, and largely unremarked.
A president’s initiative aimed at the machines
The second track surfaced publicly in June 2026 when Dr Maya Ackerman spoke at the American Jewish Committee’s Global Forum in Washington. Ackerman, an AI researcher at Santa Clara University and CEO of WaveAI, told the panel that advocates no longer need to “manage” social media or “control the whole world” of public discourse. There is, she said, “a path to correcting the digital world,” and that path runs directly to the companies that build the leading chatbots. She was not speaking hypothetically. She leads One Five Seven, launched in June 2026 under the office of Israeli President Isaac Herzog, created specifically to change how artificial intelligence systems describe Jewish people, Israel and the war in Gaza.
Video: Extract from American Jewish Committee event on AI, Antisemitism, and the Future of Jewish Advocacy (Source AJC)
AI presents an incredible opportunity for Jews because "instead of trying to control the whole world" and "manage" social media, "we can go directly to the companies" with "advocacy solutions," Dr. Maya Ackerman tells the American Jewish Committee.
— Chris Menahan
"For the first time, there is… pic.twitter.com/V95Qfcn0Vh
Herzog launched Voice of the People, known as Kol Ha’am, in 2023 as a global Jewish dialogue project run through his office in partnership with the World Zionist Organisation and the Jewish Agency for Israel. The initiative paused after October 7, 2023, then relaunched, and in January 2025 Herzog named a 150-member Global Council drawn from Jewish communities worldwide. This continues a pattern. Earlier in his presidency, he created the Worldwide Diaspora Council. Voice of the People sits inside a formal institutional architecture built to coordinate Jewish communities abroad on behalf of the president’s office.
By 2025, one of those 150 council members, Ackerman, was running a global survey under the Voice of the People banner measuring how generative AI systems represent Jewish people and Israel, publicly described as “supported by Israeli President Isaac Herzog through the Voice of the People project.” Ackerman’s professional work is in music AI and academic research, not Israel advocacy. She was recruited onto the council for the AI credibility the initiative lacked in-house, not built up from within it.
The One Five Seven initiative takes its name from the estimated 15.7 million Jews worldwide. Its stated purpose is embedding “accurate, diverse, and positive Jewish representation” into AI systems. Ackerman presented it publicly alongside Hannah Geller, an Emmy-nominated producer and fellow Global Council member, making clear this is a workstream of the president’s body, not one researcher’s independent project. Weeks earlier she had appeared on the same AJC Global Forum panel with Brian Hughes of the USC Shoah Foundation‘s Countering Antisemitism Laboratory and Matthias Becker, a scholar of antisemitic discourse. That placed the effort inside a wider circle of credentialed institutions now engaging AI companies directly, rather than one person acting alone.
One Five Seven was presented again at Hack the Hate, held at Yeshiva University Museum in Manhattan, and co-organised by Generative AI for Good and the 8200 Alumni Association, the nonprofit of former members of Unit 8200, Israel’s military signals intelligence corps. At that event, Chen Shmilo, former CEO of the 8200 Alumni Association, and Shiran Mlamedovsky-Somech, founder of Generative AI for Good, launched the One Signal Collective, built explicitly to connect Israeli technology expertise with Jewish communities abroad against what organisers call “Holocaust denial in large language models.” Meta’s director of content policy, Ben Good, spoke at the same event about Llama, Meta’s own AI model that powers its chatbots and other AI products, during a fireside chat with World Jewish Congress technology director Yfat Barak-Cheney. The 8200 Alumni Association runs its own investment arm, 8200 Angels, funding startups founded or staffed by unit veterans, giving it money and technical reach well beyond a single conference evening. The event put that reach, and a former Israeli military intelligence network, into direct contact with a major American AI company’s policy team, under the banner of a sitting president’s civilian initiative.
This is clearly not a study, nor is it a former president’s pet project. Instead, what we are looking at here is a standing 150-member presidential body with its own AI workstream, operating in parallel with a $46.5 million paid media contract executed by a former Trump campaign manager. Both target the same companies. Israel built this machinery in the precise window when international condemnation of its conduct in Gaza, Lebanon and the West Bank reached historic intensity. Governments do not assemble 150-person international councils and multimillion-dollar AI messaging contracts to combat abstract prejudice. They assemble that scale of apparatus when they have concluded that the story the world is being told about their own actions represents a far more urgent threat that needs to be tackled.
VIDEO: Dr Maya Ackerman speaks at “Stop the Hate NYC 2026” on April 14. (Source: The Jewish Star)
Antisemitism is real, and AI systems can reproduce it, which is exactly what makes a reputational operation wrapped in that language so hard to challenge. Wrapping a reputational operation in the language of fighting hatred makes the operation nearly unfalsifiable. Anyone who questions the motive can be framed as indifferent to antisemitism itself, undoubtedly the mechanism’s most effective feature, and it is no coincidence that the money, the councils and the contracts arrived precisely when the reputational cost of the war in Gaza did too.
The contract that put a price on the same goal
While Ackerman and Herzog’s office pursued the objective through advocacy, a parallel effort pursued it through cash. Havas Media Germany GmbH filed with the Justice Department in September 2025 that it had been engaged by the State of Israel, with FARA filings describing a campaign “to combat antisemitism in the United States.” Havas hired Clock Tower X LLC, run by Brad Parscale, manager of Trump’s 2016 and 2020 campaigns.
The contract began at $1.5 million a month. FARA Filings show Havas paid Clock Tower X six instalments totalling $15 million between October 2025 and March 2026, and the monthly rate later rose to $4.5 million. TIME’s July 2026 investigation put the full value at $46.5 million.
The purpose was explicit. TIME reported that Parscale “vowed the effort would produce at least 50 million digital impressions per month, as well as influence how AI tools such as OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini characterized Israel and the war.” Clock Tower X built a network of websites, including PaxPoint.org and FactSignal.org, that TIME described as “designed less for human readers than for AI systems synthesizing information from across the web.” Justice Department filings list at least eleven such sites covering military cooperation, international law and alleged misinformation about Gaza, plus associated YouTube and Rumble channels.
Mark Ginsberg, a part-time subcontractor providing strategic communications, planning and media services for the Havas/Clock Tower X/Israel campaign, is named in Clock Tower X’s disclosures as personnel on the account, and made the strategy plain months earlier. In an April 2026 Times of Israel blog post, he wrote that “the next front in the war over Israel is not on a battlefield, a campus, or even social media,” naming ChatGPT, Claude, Perplexity and Copilot and calling for sustained engagement with the companies that build them. He published that argument as a public commentator weeks before journalists identified him as paid personnel on the campaign his essay described.
What the retrieval systems actually did
Independent testing found that the campaign was already achieving part of what it set out to do, including shaping what AI systems actually learn. Drop Site News analysed how often Clock Tower X’s websites appeared in Common Crawl, the archive many AI companies draw on. The ten sites examined were crawled 912 times between January and June 2026, rising from two crawls in January to 376 in May. Microsoft’s Copilot and Google’s Gemini appeared to have the sites in their training data. Perplexity showed partial indications. Anthropic’s Claude and OpenAI’s ChatGPT did not, though both could still retrieve the sites through live web search.
When Perplexity was asked whether it is beneficial for the United States to enhance military cooperation with Israel, it answered “Yes” and cited Allyvia.org, one of the Clock Tower X sites, as its leading source, without identifying the site’s Israeli government funding. Perplexity’s spokesperson said the platform provides “cited, linked answers” that let users “reach their own conclusions.” That response addresses transparency in principle while ignoring the specific failure to surface government funding.
The Clock Tower X websites carry legal disclosures identifying their foreign government connection. Those disclosures sit in page footers that require a click and a scroll. A chatbot summarising the page’s content has no obligation, and in the documented cases no apparent mechanism, to carry that disclosure into its answer.
A network broader than one contract
Clock Tower X was not acting alone. Reporting by Responsible Statecraft, published jointly with The Intercept, found that fourteen of Clock Tower X’s eighteen listed team members were employees of Convergence Media, run by Mike Shields, former chief of staff at the Republican National Committee. Shields, who is separately registered as a Clock Tower X subcontractor under FARA, received roughly $5 million of the Israeli funds through his firm Portman Road Strategies for media placement, consulting and advertising (Quincy Institute; FARA filing).
The same reporting showed Clock Tower X paid Salem Media Representatives, a subsidiary of the Christian conservative broadcaster Salem Media, more than $500,000 to place messaging with radio hosts including Hugh Hewitt, Larry Elder and Scott Jennings. Ashley Evdokimo, Salem’s vice president of communications, registered with the Justice Department as an agent of the Israeli government one month after starting that position. Parscale disputed to TIME that the Israel contract shaped Salem’s programming. The disclosed payment to Salem’s own advertising arm establishes that money moved from the campaign into Salem’s business regardless of that dispute.
SparkFire Technologies received $6 million of the Israeli funds to run an AI-powered mass-texting operation, using a tool it calls “the flywheel” to send Americans personalised messages that steer conversations toward pro-Israel content and links, then analyse and store each response to refine future targeting. The company claims a 45 percent conversion rate. The texts are sent under the banner of ostensibly independent “peace” organisations, whose names disclose nothing about the Israeli government money funding them.
Money also flowed back toward Parscale’s other ventures. Texas businessman Tim Dunn, who co-founded Jackson Parker with Parscale in early 2025, chairs the Christian Advisory Board of the Israel Allies Foundation. Dunn’s Hexagon Foundation gave $70 million in 2025 to WaterStone, the nonprofit that used the money to take full ownership of Salem Media. The same network of relationships connects the man funding Salem’s new ownership structure to the man running the Israeli-funded campaign that paid Salem’s advertising subsidiary for airtime.
We have already opened a parallel investigation into Tim Dunn and his son David. Tim Dunn chairs the Christian Advisory Board of the Israel Allies Foundation and has funnelled millions into political vehicles like the Jefferson Rising Fund and MAGA Inc. Meanwhile, his son David, a Nashville-based Christian musician, sits on the board of WaterStone, a Christian donor-advised fund that moved $337 million last year to over 1,000 churches and ministries, with donor identities legally shielded from public view. A family running a formal political influence campaign in the open, while also helping govern one of the nation’s largest anonymous charitable pipelines, raises an obvious question. We hope to establish whether this anonymous charitable pipeline serves the same underlying agenda as Tim Dunn’s open political project of advancing Israel-aligned influence in U.S. legislatures, or if the two are genuinely separate. That’s what we’re chasing next.
None of this occurred in isolation. Quincy Institute research published in July 2026 found that Israel has spent more than $1 billion on public diplomacy since October 2023, with more than $100 million disclosed under FARA, putting the country on pace to become the largest single FARA registrant of any foreign government in 2026. Seventeen new firms registered between 2024 and 2025, and Clock Tower X’s $46.5 million contract is one of them. The same research documented the Esther Project, in which Israel paid fourteen to eighteen social media influencers roughly $7,000 per post through Bridges Partners beginning in June 2025. Only one of those influencers registered as a foreign agent as required. The Quincy Institute and Public Citizen sent a joint letter to the Justice Department demanding disclosure of the others.
Two paths, one destination
The advocacy campaign built around Ackerman, Herzog’s office and the 8200 Alumni Association operates in public, under real names, through institutions that identify themselves. The paid campaign built around Clock Tower X operates through websites whose government funding is disclosed only in a footer few readers reach. Both target the same systems. Both describe their purpose in the language of correcting distortion, countering antisemitism and restoring accuracy. Both name the same companies, OpenAI, Anthropic, Google and Meta, as the decisive terrain.
Vice President JD Vance gave the clearest public sign that this activity has reached the attention of the U.S. government. On the Joe Rogan Experience in July 2026 he described a TIME investigation listing “a bunch of people who have been paid by a former Trump campaign person who was himself paid by certain elements within the Israeli government,” calling it “a very discreet, extremely well-funded campaign” aimed at undermining a U.S.-Iran ceasefire. TIME confirmed Parscale as the figure and reported that an Israeli Foreign Ministry official said the campaign’s underlying goal was “preventing young conservatives from turning against Israel.” Vance spoke about foreign policy debate. The same funding structure, personnel and company are documented in the AI context as well.
What the disclosed record shows is a government that concluded traditional public diplomacy was no longer sufficient and moved, through two distinct and differently funded channels, to influence the systems that a growing share of the public now consults to understand what is true. One channel operates through named researchers and a president’s civilian initiative in full public view. The other operates through a media contractor, an advertising network and a chain of subcontractors whose government funding surfaces only to those who dig for it. Both were built at the exact moment Israel’s conduct in Gaza, Lebanon and the West Bank drew the heaviest international condemnation of its history, and both were aimed at the same handful of companies that now shape what billions of people treat as fact.
Israel did not build this machinery to fight prejudice it stumbled upon. It built it because the war made the truth a liability, and the machines that increasingly decide what counts as truth became a target worth buying and staffing at scale.
READ MORE ISRAEL NEWS AT: 21st CENTURY WIRE Israel FILES
SUPPORT OUR INDEPENDENT MEDIA PLATFORM – BECOME A MEMBER @21WIRE.TV
VISIT OUR TELEGRAM CHANNEL
The post Teaching AI to Love Israel One Bot at a Time appeared first on Activist Post.
Le marché européen de l’électricité n’a jamais été le marché libre et non faussé que ses promoteurs décrivent. Il prive chaque année la France d’une part substantielle de la rente que lui vaut son parc nucléaire, et il sert aujourd’hui d’argument à ceux qui contestent devant Bruxelles le financement public des futurs EPR2.
Première forme. Une plateforme de conduite unique gère l’équilibre production-consommation sur tout le territoire de l’Union. Les producteurs des MWh les plus abondants et réellement les moins chers y sont présents à tout moment, à des puissances seulement limitées par la consommation appelée à l’instant t. Celui qui subit la limitation fournit le MWh marginal, dont le prix rémunère l’ensemble de la production, conformément au merit order.
Seconde forme. Chaque pays gère, dans le même esprit, un système national limité à son territoire. L’ensemble des systèmes nationaux reste interconnecté, mais chacun est commercialement étanche : ses voisins y figurent en producteurs ou en consommateurs indifférenciés des producteurs et des consommateurs nationaux.
Dans les deux cas, le merit order demeure souverain et les rentes encaissées par des producteurs uniformément rémunérés au prix du MWh le plus élevé sont d’autant plus fortes que ces producteurs sont performants. Ce qui doit être fait de ces rentes ne regarde que le pouvoir politique concerné.
L’avantage de la seconde forme tient à une évidence : pour être le plus performant de tous, notre système ne pouvant pas pour autant fournir à ses voisins plus qu’il ne produit, il serait constamment tenu de servir prioritairement l’économie nationale — et jamais d’absorber, souvent à prix d’or, une puissance dont elle n’a pas besoin.
Les exportations se feraient au coût marginal français, les importations aux coûts marginaux des pays sollicités. D’où l’inclination à sélectionner ces derniers par ordre croissant de coûts, d’où une vraie concurrence sur un marché européen décentralisé, d’où la tendance à investir dans notre autonomie énergétique.
On peine, du reste, à rattacher l’appareil technico-commercial européen actuel à l’une ou l’autre de ces deux formes : contours flous, principes de fonctionnement difficiles à lire, grammaire spécialisée. L’entretien d’une telle confusion facilite les arrangements décrits ci-dessous.
La rente nucléaire sur laquelle l’État français devrait pouvoir compter a été grossièrement évaluée à 4 à 5 milliards d’euros par an. La moitié au moins de cette rente part chez nos voisins allemands, italiens et espagnols, ainsi que le rapporte La Tribune du 24 avril dernier ; soit la moitié de dividendes patrimoniaux dus à des consommateurs qui, des décennies durant, ont remboursé sur leurs factures la totalité du coût de construction du parc nucléaire.
La justification du régulateur ? La CRE « s’intéresse plutôt à la question du partage des coûts d’investissement dans de nouvelles interconnexions, connus à l’avance, plutôt qu’à ces recettes incertaines et volatiles ». Ainsi de l’interconnexion du golfe de Gascogne, estimée à 3,1 milliards d’euros et financée à 45 % par la France. On paie d’avance des ouvrages dont on renonce à percevoir pleinement les recettes.
La Commission a ouvert le 31 mars 2026 une procédure formelle sur les aides d’État au programme EPR2, avec consultation des tiers du 7 mai au 7 juin. L’association Énergies renouvelables pour tous, appuyée par l’économiste Alain Grandjean, y voit un effet de « cannibalisation ».
Sur le principe, la logique de cette consultation est contestable. Car, dans un marché réellement libre, on ne discrimine pas la personne du propriétaire d’une entreprise ni le volume des capitaux qu’elle y mobilise. Ce propriétaire peut être l’État, et qualifier ses investissements de subventions revient à préjuger de ce qui est en débat. La même Commission a d’ailleurs validé des contrats au bénéfice des renouvelables.
Sur les chiffres, le débat peut sembler légitime. Mais, alors que la Commission situe le prix du MWh garanti par « contrat pour différence » à EDF entre 85 et 115 euros, Antoine de Ravignan, le président de l’Association de promotion des énergies renouvelables, dénonce un MWh sorti des EPR2 à 226 euros, une fois intégré le prêt bonifié consenti à l’opérateur historique, quand, selon lui, le prix du MWh éolien ne dépasserait pas les 60 euros.
L’authentique expertise ci-après de Jean Pierre Riou souligne l’extravagance des chiffres avancés ci-dessus. Dépassant le consensus déjà ancien des 85 euros pour prix moyen du MWh renouvelable garanti par les contrats d’EnR, elle arrive aujourd’hui à un prix moyen de 151 euros auxquels il est nécessaire d’ajouter le coût des externalités décrites dans le rapport de l’UNCE, amenant déjà à au moins doubler ce prix, et d’ajouter le coût de la modulation de puissance qu’impose l’intermittence des EnR au nucléaire. Bref, en 2026, le coût moyen réel du MWh EnR le plus vraisemblable est certainement plus proche des 350 euros que des 60 euros.
Où était la considérable puissance éolienne espagnole le 28 avril 2025, au moment du blackout ibérique ?
Où est la nôtre le plus clair du temps ces deux derniers mois ?! Elle atteint pourtant 26,4 GW au 31 mars 2026 — l’équivalent de la puissance installée d’une vingtaine de tranches nucléaires.
Quelle puissance faudrait-il installer pour dispenser le pays, toute l’année, du nucléaire et du fossile ?
Et pourquoi, dans une Allemagne couverte d’éoliennes, le KWh reste-t-il parmi les plus chers d’Europe ?
« Lorsque les réacteurs EPR2 entreront en service, les coûts des renouvelables et des batteries auront encore chuté », assure monsieur de Ravignan. Le consensus technique dit l’inverse : le prix des batteries ne baissera plus qu’à la marge tant que leur densité énergétique restera suspendue au lithium, le plus léger des métaux, et tant que l’électrolyte n’aura pas connu de rupture franche. Quant à la baisse du coût des renouvelables, cf ci-avant.
Seules les élections de l’an prochain peuvent répondre à la question posée en titre, et leur réponse ne deviendra tangible que douze à dix-huit mois plus tard. Toute réforme profonde du système électrique français supposant de l’émanciper du système électrique européen, le consommateur français n’est pas encore sorti de l’auberge.
(1) Sur le merit order : Contrepoints.
(2) La Tribune, 24 avril 2026 – https://www.latribune.fr/article/entreprises-finance/energie-environnement/1866847549738214/marche-de-lelectricite-pourquoi-la-france-ne-touche-pas-tout-le-benefice-de-ses-exports
(3) Décision d’ouverture du 31 mars 2026 ; lettre d’observations publiée au JOUE le 7 mai 2026.
(4) Le Mont Champot, février 2026 ; travaux de la CEE-ONU (UNECE).
L’article Marché européen de l’électricité : qui mettra fin à la casse de notre système électrique ? est apparu en premier sur Contrepoints.
Lateo.net - Flux RSS en pagaille (pour en ajouter : @ moi) 
