Lateo.net - Flux RSS en pagaille (pour en ajouter : @ moi)

🔒
❌ À propos de FreshRSS
Il y a de nouveaux articles disponibles, cliquez pour rafraîchir la page.
Hier — 28 mars 2024Ars Technica

PyPI halted new users and projects while it fended off supply-chain attack

Par : Dan Goodin
Supply-chain attacks, like the latest PyPI discovery, insert malicious code into seemingly functional software packages used by developers. They're becoming increasingly common.

Enlarge / Supply-chain attacks, like the latest PyPI discovery, insert malicious code into seemingly functional software packages used by developers. They're becoming increasingly common. (credit: Getty Images)

PyPI, a vital repository for open source developers, temporarily halted new project creation and new user registration following an onslaught of package uploads that executed malicious code on any device that installed them. Ten hours later, it lifted the suspension.

Short for the Python Package Index, PyPI is the go-to source for apps and code libraries written in the Python programming language. Fortune 500 corporations and independent developers alike rely on the repository to obtain the latest versions of code needed to make their projects run. At a little after 7 pm PT on Wednesday, the site started displaying a banner message informing visitors that the site was temporarily suspending new project creation and new user registration. The message didn’t explain why or provide an estimate of when the suspension would be lifted.

About 10 hours later, PyPI restored new project creation and new user registration. Once again, the site provided no reason for the 10-hour halt.

Read 10 remaining paragraphs | Comments

Ubuntu will manually review Snap Store after crypto wallet scams

Man holding a piggy bank at his desk, with the piggy wired up with strange circuits and hardware

Enlarge / One thing you can say about this crypto wallet: You can't confuse it for any other. (credit: Getty Images)

The Snap Store, where containerized Snap apps are distributed for Ubuntu's Linux distribution, has been attacked for months by fake crypto wallet uploads that seek to steal users' currencies. As a result, engineers at Ubuntu's parent firm are now manually reviewing apps uploaded to the store before they are available.

The move follows weeks of reporting by Alan Pope, a former Canonical/Ubuntu staffer on the Snapcraft team, who is still very active in the ecosystem. In February, Pope blogged about how one bitcoin investor lost nine bitcoins (about $490,000 at the time) by using an "Exodus Wallet" app from the Snap store. Exodus is a known cryptocurrency wallet, but this wallet was not from that entity. As detailed by one user wondering what happened on the Snapcraft forums, the wallet immediately transferred his entire balance to an unknown address after a 12-word recovery phrase was entered (which Exodus tells you on support pages never to do).

Pope takes pains to note that cryptocurrency is inherently fraught with loss risk. Still, Ubuntu's App Center, which presents the Snap Store for desktop users, tagged the "Exodus" app as "Safe," and the web version of the Snap Store describes Snaps as "safe to run." While Ubuntu is describing apps as "Safe" in the sense of being an auto-updating container with runtime confinement (or "sandboxed"), a green checkmark with "Safe" next to it could be misread, especially by a newcomer to Ubuntu, Snaps, and Linux generally.

Read 7 remaining paragraphs | Comments

FTX fraudster Sam Bankman-Fried sentenced to 25 years in prison

FTX Founder Sam Bankman-Fried arrives at court, surrounded by photographers and other people. One man appears to be holding Bankman-Fried around the torso and escorting him.

Enlarge / FTX Founder Sam Bankman-Fried arrives at US District Court on March 30, 2023, in New York City after being hit with a criminal charge for allegedly authorizing a bribe of at least $40 million to one or more Chinese government officials. (credit: Getty Images | Michael Santiago )

Convicted FTX fraudster Sam Bankman-Fried was sentenced to 25 years in prison today, according to news reports.

The founder and ex-CEO of cryptocurrency exchange FTX was sentenced this morning by Judge Lewis Kaplan in US District Court for the Southern District of New York. Bankman-Fried had requested a sentence of 63 to 78 months (5.25 to 6.5 years), arguing that he deserved leniency because of his "charitable works and demonstrated commitment to others."

Kaplan ordered a forfeiture of $11 billion but did not order restitution "due to the complexity of the case and the number of victims," the court docket said. Kaplan instead authorized the US to compensate victims with "forfeited assets through a remission process, as restitution would be impractical in this case."

Read 17 remaining paragraphs | Comments

Yamaha and Lola pair up to enter Formula E next season

A Gen3 Formula E car with a yellow and blue livery and Lola logos on it.

Enlarge / After a 10-year gap, Lola is back developing an electric racecar, or at least the powertrain for one, as it will enter Formula E next season. (credit: Lola)

In 2022, we brought news that Lola, a once-famous racing company, was planning its renaissance. Lola never really cracked Formula 1, but it did have success in IndyCar and sports car racing with cars it designed and built from the 1960s until it ceased trading in 2012. Now, under new ownership, the company has been rebuilding its engineering facilities and expertise. And together with Yamaha as its technical partner, it has chosen Formula E for its official return to professional motorsport.

Formula E's dart-shaped electric single-seaters are getting a bit of an update before they start season 11 next year. We expect new bodywork, better tires, and perhaps the ability to use the front electric motor to send power to the wheels instead of just acting as regenerative brakes on the front axle, but those components are all spec parts, meaning every team has to use the same ones without modifying them.

That goes for the battery, too, but there is freedom when it comes to the 470 hp (350 kW) electric motor that powers the rear wheels. And then there's the software, without which the car won't go anywhere.

Read 8 remaining paragraphs | Comments

Quantum computing progress: Higher temps, better error correction

conceptual graphic of symbols representing quantum states floating above a stylized computer chip.

Enlarge (credit: vital)

There's a strong consensus that tackling most useful problems with a quantum computer will require that the computer be capable of error correction. There is absolutely no consensus, however, about what technology will allow us to achieve that. A large number of companies, including major players like Microsoft, Intel, Amazon, and IBM, have all committed to different technologies to get there, while a collection of startups are exploring an even wider range of potential solutions.

We probably won't have a clearer picture of what's likely to work for a few years. But there's going to be lots of interesting research and development work between now and then, some of which may ultimately represent key milestones in the development of quantum computing. To give you a sense of that work, we're going to look at three papers that were published within the last couple of weeks, each of which tackles a different aspect of quantum computing technology.

Hot stuff

Error correction will require connecting multiple hardware qubits to act as a single unit termed a logical qubit. This spreads a single bit of quantum information across multiple hardware qubits, making it more robust. Additional qubits are used to monitor the behavior of the ones holding the data and perform corrections as needed. Some error-correction schemes require over a hundred hardware qubits for each logical qubit, meaning we'd need tens of thousands of hardware qubits before we could do anything practical.

Read 21 remaining paragraphs | Comments

Facebook secretly spied on Snapchat usage to confuse advertisers, court docs say

Facebook secretly spied on Snapchat usage to confuse advertisers, court docs say

Enlarge (credit: NurPhoto / Contributor | NurPhoto)

Unsealed court documents have revealed more details about a secret Facebook project initially called "Ghostbusters," designed to sneakily access encrypted Snapchat usage data to give Facebook a leg up on its rival, just when Snapchat was experiencing rapid growth in 2016.

The documents were filed in a class-action lawsuit from consumers and advertisers, accusing Meta of anticompetitive behavior that blocks rivals from competing in the social media ads market.

"Whenever someone asks a question about Snapchat, the answer is usually that because their traffic is encrypted, we have no analytics about them," Facebook CEO Mark Zuckerberg (who has since rebranded his company as Meta) wrote in a 2016 email to Javier Olivan.

Read 18 remaining paragraphs | Comments

À partir d’avant-hierArs Technica

“The king is dead”—Claude 3 surpasses GPT-4 on Chatbot Arena for the first time

Two toy robots fighting, one knocking the other's head off.

Enlarge (credit: Getty Images / Benj Edwards)

On Tuesday, Anthropic's Claude 3 Opus large language model (LLM) surpassed OpenAI's GPT-4 (which powers ChatGPT) for the first time on Chatbot Arena, a popular crowdsourced leaderboard used by AI researchers to gauge the relative capabilities of AI language models. "The king is dead," tweeted software developer Nick Dobos in a post comparing GPT-4 Turbo and Claude 3 Opus that has been making the rounds on social media. "RIP GPT-4."

Since GPT-4 was included in Chatbot Arena around May 10, 2023 (the leaderboard launched May 3 of that year), variations of GPT-4 have consistently been on the top of the chart until now, so its defeat in the Arena is a notable moment in the relatively short history of AI language models. One of Anthropic's smaller models, Haiku, has also been turning heads with its performance on the leaderboard.

"For the first time, the best available models—Opus for advanced tasks, Haiku for cost and efficiency—are from a vendor that isn't OpenAI," independent AI researcher Simon Willison told Ars Technica. "That's reassuring—we all benefit from a diversity of top vendors in this space. But GPT-4 is over a year old at this point, and it took that year for anyone else to catch up."

Read 8 remaining paragraphs | Comments

SCOTUS mifepristone case: Justices focus on anti-abortion groups’ legal standing

Par : Beth Mole
Demonstrators participate in an abortion-rights rally outside the Supreme Court as the justices of the court hear oral arguments in the case of the <em>US Food and Drug Administration v. Alliance for Hippocratic Medicine</em> on March 26, 2024 in Washington, DC.

Enlarge / Demonstrators participate in an abortion-rights rally outside the Supreme Court as the justices of the court hear oral arguments in the case of the US Food and Drug Administration v. Alliance for Hippocratic Medicine on March 26, 2024 in Washington, DC. (credit: Getty | Anna Moneymaker)

The US Supreme Court on Tuesday heard arguments in a case seeking to limit access to the abortion and miscarriage drug mifepristone, with a majority of justices expressing skepticism that the anti-abortion groups that brought the case have the legal standing to do so.

The case threatens to dramatically alter access to a drug that has been safely used for decades and, according to the Guttmacher Institute, was used in 63 percent of abortions documented in the health care system in 2023. But, it also has sweeping implications for the Food and Drug Administration's authority over drugs, marking the first time that courts have second-guessed the agency's expert scientific analysis and moved to restrict access to an FDA-approved drug.

As such, the case has rattled health experts, reproductive health care advocates, the FDA, and the pharmaceutical industry alike. But, based on the line of questioning in today's oral arguments, they have reason to breathe a sigh of relief.

Read 11 remaining paragraphs | Comments

Missouri AG sues Media Matters over its X research, demands donor names

A photo of Elon Musk next to the logo for X, the social network formerly known as Twitter,.

Enlarge (credit: Getty Images | NurPhoto )

Missouri Attorney General Andrew Bailey yesterday sued Media Matters in an attempt to protect Elon Musk and X from the nonprofit watchdog group's investigations into hate speech on the social network. Bailey's lawsuit claims that "Media Matters has used fraud to solicit donations from Missourians in order to trick advertisers into removing their advertisements from X, formerly Twitter, one of the last platforms dedicated to free speech in America."

Bailey didn't provide much detail on the alleged fraud but claimed that Media Matters is guilty of "fraudulent manipulation of data on X.com." That's apparently a reference to Media Matters reporting that X placed ads for major brands next to posts touting Hitler and Nazis. X has accused Media Matters of manipulating the site's algorithm by endlessly scrolling and refreshing.

Bailey yesterday issued an investigative demand seeking names and addresses of all Media Matters donors who live in Missouri and a range of internal communications and documents regarding the group's research on Musk and X. Bailey anticipates that Media Matters won't provide the requested materials, so he filed the lawsuit asking Cole County Circuit Court for an order to enforce the investigative demand.

Read 22 remaining paragraphs | Comments

Genesis unveils its take on the big luxury EV—the Neolun Concept

The front half of the Genesis Neolum Concept EV

Enlarge / This concept points the way to a future Genesis flagship SUV. (credit: Jonathan Gitlin)

NEW YORK—You can always rely on Genesis to bring at least one interesting concept to the New York International Auto Show. This year, the company brought several. At a busy reveal at the brand's Genesis House in Manhattan, it showed us its high-performance ambitions with not one but four bright orange machines, plus one rather famous Belgian racing driver. Then, in a chamber reminiscent of The Barmacide Feast, we got to see the poshest Genesis yet, the brand's take on a big luxury electric vehicle inspired by Korean hospitality.

The Neolum Concept

  • The Genesis Neolum has the looks to hold its own next to any Range Rover or Bentley Bentayga. [credit: Jonathan Gitlin ]

Genesis was tight-lipped in the lead-up to Monday night's unveilings, but no one was entirely surprised to see a big electric SUV. Genesis is owned by Hyundai Group, after all, and has access to the E-GMP architecture, a thoroughly up-to-date flexible platform that keeps impressing us. Kia just used E-GMP to great effect to make the EV9, a three-row family SUV. And Hyundai's take on that form factor is due later this year in the production Ioniq 7, so an upmarket model from Genesis seemed obvious.

"The last eight years, it was about finding who we are and then discovering DNA for the Genesis," said SangYup Lee, global design head for Genesis. "Now it's time to expand."

Read 11 remaining paragraphs | Comments

Super Mario Maker’s “final boss” was a fraud all along

When good robots fall into the wrong hands, bad things can happen...

When good robots fall into the wrong hands, bad things can happen... (credit: Aurich Lawson | Nintendo)

The Super Mario Maker community and "Team 0%" have declared victory in their years-long effort to clear every user-submitted level in the original Wii U game before the servers shut off for good on April 8. That victory declaration comes despite the fact that no human player has yet to clear "Trimming the Herbs" (TTH), the ultra-hard level that gained notoriety this month as what was thought to be the final "uncleared" level in the game.

This strange confluence of events is the result of an admission by Ahoyo, the creator of Trimming the Herbs, who came clean Friday evening regarding his use of automated, tool-assisted speedrun (TAS) methods in creating the level. That means he was able to use superhuman capabilities like slow-motion, rewinding, and frame advance to pre-record the precise set of perfectly timed inputs needed to craft the "creator clear" that was necessary to upload the level in the first place.

Ahoyo's video of a "creator clear" for Trimming the Herbs, which he now admits was created using TAS methods.

"I’m sorry for the drama [my level] caused within the community, and I regret the ordeal," Ahoyo wrote on the Team 0% Discord and social media. "But at least it was interesting. However in the end the truth matters most. Congratulations to Team 0% for their well-earned achievement!"

Read 29 remaining paragraphs | Comments

Starliner’s first commander: Don’t expect perfection on crew test flight

Technicians at NASA's Kennedy Space Center in Florida prepare Boeing's Starliner spacecraft for fueling.

Enlarge / Technicians at NASA's Kennedy Space Center in Florida prepare Boeing's Starliner spacecraft for fueling. (credit: Boeing)

HOUSTON—While it doesn't have the same relevance to public consciousness as safety problems with commercial airliners, a successful test flight of Boeing's Starliner spacecraft in May would be welcome news for the beleaguered aerospace company.

This will be the first time the Starliner capsule flies into low-Earth orbit with humans aboard. NASA astronauts Butch Wilmore and Suni Williams are in the final stages of training for the so-called Crew Flight Test (CFT), a milestone running seven years behind the schedule Boeing said it could achieve when it won a $4.2 billion commercial crew contract from NASA a decade ago.

If schedules hold, Wilmore and Williams will take off inside Boeing's Starliner spacecraft aboard a United Launch Alliance Atlas V rocket after midnight May 1, local time, from Cape Canaveral Space Force Station in Florida. They will fly Starliner to the International Space Station for a stay of at least eight days, then return the capsule to a parachute-assisted, airbag-cushioned landing in the western United States, likely at White Sands, New Mexico.

Read 33 remaining paragraphs | Comments

macOS Sonoma 14.4.1 released to fix the stuff that the 14.4 update broke

An M3 MacBook Air running macOS Sonoma.

Enlarge / An M3 MacBook Air running macOS Sonoma. (credit: Andrew Cunningham)

Apple has just released version 14.4.1 for macOS Sonoma, a small-but-significant patch that claims to fix several issues with third-party software and accessories that cropped up in the 14.4 update. The 14.4.1 release also includes a pair of security fixes.

Apple’s release notes highlight fixes for three major problems:

  • USB hubs connected to external displays may not be recognized
  • Copy protected Audio Unit plug-ins designed for professional music apps may not open or pass validation
  • Apps that include Java may quit unexpectedly

Users and companies began noticing problems shortly after the macOS 14.4 update was released earlier this month. Reports of broken USB hubs cropped up on Reddit, the Apple Support Communities forums, and elsewhere within the first couple of days, and issues with Java and iLok audio software DRM devices were reported later on. Some users also reported broken printer drivers and deleted file revisions in iCloud Drive, though Apple's release notes don't mention those problems.

Read 3 remaining paragraphs | Comments

Testing the 2024 BMW M2—maybe the last M car with a manual transmission

A pale blue BMW M2 seen parked in the hills

Enlarge / BMW's M2 might be the last M car it builds with three pedals and a stick shift. (credit: Peter Nelson)

We're at an interesting crossroads in the high-performance enthusiast car market. Running east to west is the adoption of electric vehicles and a slow reduction in internal combustion engine car production. North to south is the progression of ICE horsepower from the factory over the years, and it's unclear how far it continues from here. Coming in diagonally is the weakening demand for manual transmissions—this is sadly where they end.

In the middle of this intersection is the 2024 BMW M2 six-speed manual, hanging its tail out in a massive controlled drift around the edges, expressing one last hurrah as BMW's final object of internal-combustion M car affection.

I recently had the opportunity to pilot BMW's latest, smallest M car through some of Southern California's most fun mountain roads, plus Willow Springs International Raceway's Streets of Willow circuit. When it comes to quickly figuring out this kind of car's powertrain and chassis, I can't think of a better mix of pavement. Here's what makes the latest—and last—six-speed-manual-equipped M2 generation an overall excellent enthusiast coupe.

Read 21 remaining paragraphs | Comments

Dragon’s Dogma 2 is gritty, janky, goofy, tough, and lots of fun

Player shooting down a griffon with circling beams of light.

Enlarge / One day I will own griffons in such spectacular fashion. But I'm currently carrying a too-heavy backpack and clipped through a hut wall.

With all due respect to the Capcom team, which poured itself into Dragon’s Dogma 2 and deserves praise, raises, and time off, let me get right to it: I love this game for how dumb it is.

I mean "dumb" in the way most heavy metal lyrics are dumb, but you find yourself rocking out nonetheless. Dumb like when you laugh uncontrollably at the sight of someone getting conked on the head and falling over backward. Dumb as in the silliest bits of Monty Python and the Holy Grail, just nowhere near as self-aware (unless, due to translation issues, this game actually is self-aware, then I apologize).

Dragon’s Dogma 2 (DD2) reminds me of playing another huge, dumb, enjoyable game: The Elder Scrolls V: Skyrim. Not the first time you play through it, though. I’m talking about the second or third run-through (or that 100-plus-hour save in which you refuse to finish the game), and your admiration of this huge, rich world gives way to utter ridiculousness. You one-shot dragons with your broken stealth-archer build, you put buckets on the heads of NPCs to rob them, and you marvel at how the most effective fast travel is horse tilting. You lunge into possibilities, choose chaos, and appreciate all the ways you can do so.

Read 24 remaining paragraphs | Comments

GM stops sharing driver data with brokers amid backlash

Scissors cut off a stream of data from a toy car to a cloud

Enlarge (credit: Aurich Lawson | Getty Images)

After public outcry, General Motors has decided to stop sharing driving data from its connected cars with data brokers. Last week, news broke that customers enrolled in GM's OnStar Smart Driver app have had their data shared with LexisNexis and Verisk.

Those data brokers in turn shared the information with insurance companies, resulting in some drivers finding it much harder or more expensive to obtain insurance. To make matters much worse, customers allege they never signed up for OnStar Smart Driver in the first place, claiming the choice was made for them by salespeople during the car-buying process.

Now, in what feels like an all-too-rare win for privacy in the 21st century, that data-sharing deal is no more.

Read 1 remaining paragraphs | Comments

Cable ISP fined $10,000 for lying to FCC about where it offers broadband

Businessman secretly crossing fingers

Enlarge / "Yes, we offer Internet at your address." (credit: Chev Wilkinson via Getty Images)

An Internet service provider that admitted lying to the Federal Communications Commission about where it offers broadband will pay a $10,000 fine and implement a compliance plan to prevent future violations.

Jefferson County Cable (JCC), a small ISP in Toronto, Ohio, admitted that it falsely claimed to offer fiber service in an area that it hadn't expanded to yet. A company executive also admitted that the firm submitted false coverage data to prevent other ISPs from obtaining government grants to serve the area. Ars helped expose the incident in a February 2023 article.

The FCC announced the outcome of its investigation on March 15, saying that Jefferson County Cable violated the Broadband Data Collection program requirements and the Broadband DATA Act, a US law, "in connection with reporting inaccurate information or data with respect to the Company's ability to provide broadband Internet access service."

Read 20 remaining paragraphs | Comments

Hackers can unlock over 3 million hotel doors in seconds

Par : WIRED
Picture of Saflok lock on hotel door

Enlarge / A Saflok branded lock. (credit: Dormakaba)

When thousands of security researchers descend on Las Vegas every August for what's come to be known as “hacker summer camp,” the back-to-back Black Hat and Defcon hacker conferences, it's a given that some of them will experiment with hacking the infrastructure of Vegas itself, the city's elaborate array of casino and hospitality technology. But at one private event in 2022, a select group of researchers were actually invited to hack a Vegas hotel room, competing in a suite crowded with their laptops and cans of Red Bull to find digital vulnerabilities in every one of the room's gadgets, from its TV to its bedside VoIP phone.

One team of hackers spent those days focused on the lock on the room's door, perhaps its most sensitive piece of technology of all. Now, more than a year and a half later, they're finally bringing to light the results of that work: a technique they discovered that would allow an intruder to open any of millions of hotel rooms worldwide in seconds, with just two taps.

Today, Ian Carroll, Lennert Wouters, and a team of other security researchers are revealing a hotel keycard hacking technique they call Unsaflok. The technique is a collection of security vulnerabilities that would allow a hacker to almost instantly open several models of Saflok-brand RFID-based keycard locks sold by the Swiss lock maker Dormakaba. The Saflok systems are installed on 3 million doors worldwide, inside 13,000 properties in 131 countries.

Read 19 remaining paragraphs | Comments

World’s first global AI resolution unanimously adopted by United Nations

The United Nations building in New York.

Enlarge / The United Nations building in New York. (credit: Getty Images)

On Thursday, the United Nations General Assembly unanimously consented to adopt what some call the first global resolution on AI, reports Reuters. The resolution aims to foster the protection of personal data, enhance privacy policies, ensure close monitoring of AI for potential risks, and uphold human rights. It emerged from a proposal by the United States and received backing from China and 121 other countries.

Being a nonbinding agreement and thus effectively toothless, the resolution seems broadly popular in the AI industry. On X, Microsoft Vice Chair and President Brad Smith wrote, "We fully support the @UN's adoption of the comprehensive AI resolution. The consensus reached today marks a critical step towards establishing international guardrails for the ethical and sustainable development of AI, ensuring this technology serves the needs of everyone."

The resolution, titled "Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development," resulted from three months of negotiation, and the stakeholders involved seem pleased at the level of international cooperation. "We're sailing in choppy waters with the fast-changing technology, which means that it's more important than ever to steer by the light of our values," one senior US administration official told Reuters, highlighting the significance of this "first-ever truly global consensus document on AI."

Read 5 remaining paragraphs | Comments

Android 15 gets satellite messaging, starts foldable cover app support

Par : Ron Amadeo
The Android 15 logo. This is "Android V," if you can't tell from the logo.

Enlarge / The Android 15 logo. This is "Android V," if you can't tell from the logo. (credit: Google)

Android 15 continues its march toward release with the Android 15 Developer Preview 2. Android 15 won't be out until around October, but the first preview shipped a month ago. It's time for another one!

Android's satellite messaging support has been in the works for about a year now, and it sounds like Android 15 is going to launch the feature for apps. The new OS is including notifications and better status bar indicators for when you're connected to space. A "NonTerrestrialNetwork" API will let apps know when they're limited to barely there satellite connectivity. Google says Android 15 will let third-party SMS and MMS applications tap into the satellite connectivity APIs, but enhanced messaging with RCS support will be limited to "preloaded" applications only. It seems incredible that Google doesn't have public APIs for third-party RCS apps, but here's your confirmation that Android 15 will continue locking out Play Store apps from RCS.

  • Android 15's new satellite messaging UI. [credit: Google ]

Android's PDF support can be all over the place depending on what device you have, so Android 15 is including making some big improvements to the built-in PDF render. First it's going to end up as a module so it can be updated via the Play Store. Google says this Android 15 version is getting "advanced features such as rendering password-protected files, annotations, form editing, searching, and selection with copy."

Read 3 remaining paragraphs | Comments

SBF repeatedly lied to get out of “supervillain” prison term, FTX CEO alleges

FTX founder Sam Bankman-Fried (R) departs Manhattan Federal Court after an arraignment hearing on March 30, 2023, in New York City.

Enlarge / FTX founder Sam Bankman-Fried (R) departs Manhattan Federal Court after an arraignment hearing on March 30, 2023, in New York City. (credit: Drew Angerer / Staff | Getty Images News)

The CEO of FTX Trading, John Ray, sent a letter to Judge Lewis Kaplan Wednesday to correct what he called "callously" and "demonstrably false" claims that disgraced FTX founder Sam Bankman-Fried made in hopes of receiving a lighter sentence for crimes including defrauding FTX customers.

In a sentencing memo, Bankman-Fried asked the court to drastically slash his prison sentence from what he considered a "grotesque" 110-year maximum to five to six years. Prosecutors have suggested the sentence should be between 40 and 50 years, but Bankman-Fried claimed such a sentence painted him as a "depraved supervillain," Bloomberg reported.

The lightest sentence was appropriate, Bankman-Fried claimed, because the "most reasonable estimate of loss" and "harm" to customers, lenders, and investors is "zero."

Read 34 remaining paragraphs | Comments

Choose your side in a civil war with House of the Dragon’s dueling S2 trailers

This short teaser for S2 of HBO's House of the Dragon lets you choose between two full trailers.

It's been a long wait for the second season of HBO's House of the Dragon, in which House Targaryen descends into civil war over the heir to the Iron Throne. It's set to premiere in June, and HBO is ramping up its marketing with a rather clever twist: not one official trailer, but two, each presenting the perspective of one side in the bloody conflict. And we get to choose which trailer we'd like to view—although if you're like us, you'll elect to watch both.

(Spoilers for the first season below.)

As I've written previously, HBO's House of the Dragon debuted in 2022 with a solid, promising pilot episode, and the remainder of the season lived up to that initial promise. The series is set nearly 200 years before the events of Game of Thrones and chronicles the beginning of the end of House Targaryen's reign. The primary source material is Fire and Blood, a fictional history of the Targaryen kings written by George R.R. Martin. As book readers know, those events culminated in a civil war and the extinction of the dragons—at least until Daenerys Targaryen came along.

Read 7 remaining paragraphs | Comments

This stretchy electronic material hardens upon impact just like “oobleck”

This flexible and conductive material has “adaptive durability,” meaning it gets stronger when hit.

Enlarge / This flexible and conductive material has “adaptive durability,” meaning it gets stronger when hit. (credit: Yue (Jessica) Wang)

Scientists are keen to develop new materials for lightweight, flexible, and affordable wearable electronics so that, one day, dropping our smartphones won't result in irreparable damage. One team at the University of California, Merced, has made conductive polymer films that actually toughen up in response to impact rather than breaking apart, much like mixing corn starch and water in appropriate amounts produces a slurry that is liquid when stirred slowly but hardens when you punch it (i.e., "oobleck"). They described their work in a talk at this week's meeting of the American Chemical Society in New Orleans.

"Polymer-based electronics are very promising," said Di Wu, a postdoc in materials science at UCM. "We want to make the polymer electronics lighter, cheaper, and smarter. [With our] system, [the polymers] can become tougher and stronger when you make a sudden movement, but... flexible when you just do your daily, routine movement. They are not constantly rigid or constantly flexible. They just respond to your body movement."

As we've previously reported, oobleck is simple and easy to make. Mix one part water to two parts corn starch, add a dash of food coloring for fun, and you've got oobleck, which behaves as either a liquid or a solid, depending on how much stress is applied. Stir it slowly and steadily and it's a liquid. Punch it hard and it turns more solid under your fist. It's a classic example of a non-Newtonian fluid.

In an ideal fluid, the viscosity largely depends on temperature and pressure: Water will continue to flow regardless of other forces acting upon it, such as being stirred or mixed. In a non-Newtonian fluid, the viscosity changes in response to an applied strain or shearing force, thereby straddling the boundary between liquid and solid behavior. Stirring a cup of water produces a shearing force, and the water shears to move out of the way. The viscosity remains unchanged. But for non-Newtonian fluids like oobleck, the viscosity changes when a shearing force is applied.

Read 6 remaining paragraphs | Comments

Vernor Vinge, father of the tech singularity, has died at age 79

A photo of Vernor Vinge in 2006. https://en.wikipedia.org/wiki/Vernor_Vinge#/media/File:Vernor_Vinge.jpg

Enlarge / A photo of Vernor Vinge in 2006. (credit: Raul654)

On Wednesday, author David Brin announced that Vernor Vinge, sci-fi author, former professor, and father of the technological singularity concept, died from Parkinson's disease at age 79 on March 20, 2024, in La Jolla, California. The announcement came in a Facebook tribute where Brin wrote about Vinge's deep love for science and writing.

"A titan in the literary genre that explores a limitless range of potential destinies, Vernor enthralled millions with tales of plausible tomorrows, made all the more vivid by his polymath masteries of language, drama, characters, and the implications of science," wrote Brin in his post.

As a sci-fi author, Vinge won Hugo Awards for his novels A Fire Upon the Deep (1993), A Deepness in the Sky (2000), and Rainbows End (2007). He also won Hugos for novellas Fast Times at Fairmont High (2002) and The Cookie Monster (2004). As Mike Glyer's File 770 blog notes, Vinge's novella True Names (1981) is frequency cited as the first presentation of an in-depth look at the concept of "cyberspace."

Read 8 remaining paragraphs | Comments

Super Mario Bros. Wonder devs created 2,000 game-altering “Wonder Effect” ideas

Just some of the unused Wonder Effect ideas submitted via sticky note by the development team.

Enlarge / Just some of the unused Wonder Effect ideas submitted via sticky note by the development team. (credit: Kyle Orland)

SAN FRANCISCO—When thinking about what makes 2D Mario games special, Super Mario Bros. Wonder director Shiro Mouri recalled the excitement he felt playing the original Super Mario Bros., discovering things like the warp zone and hidden vine blocks for the first time. Across decades of 2D Mario games with similar designs, though, it has been harder and harder to make a game that feels like it's "full of secrets and mysteries," as he said during a Game Developers Conference presentation this week.

"At some point, all of this has become normal," Mouri said of once-fantastical Mario game elements like mushrooms and coin blocks that have now become staples of the games.

Recapturing a world full of "secrets and mysteries" was the guiding principle for the development of Super Mario Bros. Wonder, Mouri said, but it took a while to figure out the new perspective necessary to get to that point. When Mouri prototyped an item that warped Mario to a new location, for instance, producer Takashi Tezuka said the effect "isn't so different from how it's always been. What if we changed the environment instead?"

Read 7 remaining paragraphs | Comments

GPT-5 might arrive this summer as a “materially better” update to ChatGPT

A glowing OpenAI logo on a blue background.

Enlarge

When OpenAI launched its GPT-4 AI model a year ago, it created a wave of immense hype and existential panic from its ability to imitate human communication and composition. Since then, the biggest question in AI has remained the same: When is GPT-5 coming out? During interviews and media appearances around the world, OpenAI CEO Sam Altman frequently gets asked this question, and he usually gives a coy or evasive answer, sometimes coupled with promises of amazing things to come.

According to a new report from Business Insider, OpenAI is expected to release GPT-5, an improved version of the AI language model that powers ChatGPT, sometime in mid-2024—and likely during the summer. Two anonymous sources familiar with the company have revealed that some enterprise customers have recently received demos of GPT-5 and related enhancements to ChatGPT.

One CEO who recently saw a version of GPT-5 described it as "really good" and "materially better," with OpenAI demonstrating the new model using use cases and data unique to his company. The CEO also hinted at other unreleased capabilities of the model, such as the ability to launch AI agents being developed by OpenAI to perform tasks automatically.

Read 7 remaining paragraphs | Comments

Lifesaving gene therapy for kids is world’s priciest drug at $4.25M

Par : Beth Mole
A mother with her twin 6-year-old boys who have metachromatic leukodystrophy, a genetic disease that leaves them unable to move. Photo taken on September 3, 2004.

Enlarge / A mother with her twin 6-year-old boys who have metachromatic leukodystrophy, a genetic disease that leaves them unable to move. Photo taken on September 3, 2004. (credit: Getty | John Ewing/Portland Press Herald)

In a medical triumph, the US Food and Drug Administration on Monday approved a gene therapy that appears to trounce a rare, tragic disease that progressively steals children's ability to talk, move, and think, leading to a vegetative state and death. For those who begin to slip away in infancy, many die by age 5. But, with the new therapy, 37 children in an initial trial were all still alive at age 6. Most could still talk, walk on their own, and perform normally on IQ tests, which was unseen in untreated children. Some of the earliest children treated have now been followed for up to 12 years—and they continue to do well.

But, the triumph turned bittersweet today, Wednesday, as the company behind the therapy, Lenmeldy, set the price for the US market at $4.25 million, making it the most expensive drug in the world. The price is $310,000 higher than what experts calculated to be the maximum fair price for the lifesaving drug; the nonprofit Institute for Clinical and Economic Review, or ICER, gave a range last October of between $2.29 million to $3.94 million.

The price raises questions about whether state, federal, and private health insurance plans will be able to shoulder the costs. "Unless states have allocated appropriately for it, and looked at the drug pipeline, they may not be prepared for what could be significant cost spikes," Edwin Park, a research professor at the McCourt School of Public Health at Georgetown University, told CNN.

Read 8 remaining paragraphs | Comments

Nvidia announces “moonshot” to create embodied human-level AI in robot form

An illustration of a humanoid robot created by Nvidia.

Enlarge / An illustration of a humanoid robot created by Nvidia. (credit: Nvidia)

In sci-fi films, the rise of humanlike artificial intelligence often comes hand in hand with a physical platform, such as an android or robot. While the most advanced AI language models so far seem mostly like disembodied voices echoing from an anonymous data center, they might not remain that way for long. Some companies like Google, Figure, Microsoft, Tesla, Boston Dynamics, and others are working toward giving AI models a body. This is called "embodiment," and AI chipmaker Nvidia wants to accelerate the process.

"Building foundation models for general humanoid robots is one of the most exciting problems to solve in AI today," said Nvidia CEO Jensen Huang in a statement. Huang spent a portion of Nvidia's annual GTC conference keynote on Monday going over Nvidia's robotics efforts. "The next generation of robotics will likely be humanoid robotics," Huang said. "We now have the necessary technology to imagine generalized human robotics."

To that end, Nvidia announced Project GR00T, a general-purpose foundation model for humanoid robots. As a type of AI model itself, Nvidia hopes GR00T (which stands for "Generalist Robot 00 Technology" but sounds a lot like the famous Marvel character) will serve as an AI mind for robots, enabling them to learn skills and solve various tasks on the fly. In a tweet, Nvidia researcher Linxi "Jim" Fan called the project "our moonshot to solve embodied AGI in the physical world."

Read 11 remaining paragraphs | Comments

Carmakers’ shady data sharing takes spotlight in GM connected car scandal

A cartoon of a car, with a straw coming out of its roof, and a cloud coming out of the straw

Enlarge (credit: Aurich Lawson | Getty Images)

Few Ars readers will have been surprised by the news from last week concerning General Motors' connected cars. As The New York Times reported, some owners of vehicles made by General Motors have been having a hard time getting car insurance. The reason? They unwittingly agreed to share their driving data with a third party. Now, at least one driver is suing. If more follow suit, this could be the push the industry needs to do better.

The heart of the problem is one of GM's OnStar connected-car services, called Smart Driver. We've tested it out in the past—it monitors things like how fast you drive, how hard you accelerate and brake, how often you drive at night, and your fuel economy, then uses that data to generate a numerical score from 0 to 100, with a higher number indicating that you're a safer driver.

These kinds of services can be useful—most people think they're great drivers until they start getting independent feedback. And the data that Smart Driver collects really can help you drive more economically and with less risk. But as I noted at the time, I was glad my insurance rates weren't at risk via data sharing with an insurer.

Read 20 remaining paragraphs | Comments

Intel receives $8.5 billion from US for expanding high-end fab capacity

Intel sign

Enlarge (credit: Bloomberg)

Intel will receive $8.5 billion in direct funding and $11 billion in loans from the US government to expand its capacity to make high-end chips as it seeks to reinvent itself as a national champion in the sector and compete with the likes of Taiwan’s TSMC and South Korea’s Samsung.

US President Joe Biden will travel to Intel’s site in Chandler, Arizona, on Wednesday to announce the package, which will go toward building new facilities for the company in the south-western state, as well as in Ohio, New Mexico and Oregon.

Biden’s intervention in Arizona—one of a handful of swing states that will decide the US presidential election pitting him against Donald Trump—comes as the Democratic president is trying to boost his languishing approval ratings on the economy.

Read 14 remaining paragraphs | Comments

Rocket launch marks big step in building China’s lunar infrastructure

A Long March 8 rocket, standing 165 feet (50 meters) tall, rolled out of its assembly building to its launch pad Sunday at the Wenchang Space Launch Site.

Enlarge / A Long March 8 rocket, standing 165 feet (50 meters) tall, rolled out of its assembly building to its launch pad Sunday at the Wenchang Space Launch Site. (credit: Luo Yunfei/China News Service/VCG via Getty Images)

The next phase of China's Moon program began with the launch of a new data relay satellite Monday to link lunar landers and rovers on the far side of the Moon with ground controllers back on Earth.

This launch sent China's Queqiao-2 relay spacecraft toward the Moon, where it will enter an elliptical orbit and position itself for the arrival of China's next robotic lunar lander, Chang'e 6, later this year.

A medium-lift Long March 8 rocket carried the Queqiao-2 spacecraft aloft from the Wenchang launch base, located on Hainan Island in southern China. This was the third flight of the kerosene-fueled Long March 8, one of a new generation of Chinese rockets designed to replace older Long March launcher designs burning toxic propellant.

Read 19 remaining paragraphs | Comments

Entirely accurate 3D-printed Mac Plus built in these 29 painstaking steps

Booted Mac replica with MacPaint open,

Enlarge (credit: Kevin Noki)

Have you ever worked on a hobby project where modifying and compiling the source code for a Linux-based emulator was possibly the easiest and most straightforward part of the whole thing?

Kevin Noki really, really wanted a functioning Macintosh Plus, complete with a functioning, auto-ejecting disk drive that it could boot from. The German maker already had a Mac Plus (1Mb) from eBay, but it had both a busted power supply and floppy drive. Rather than carve out the busted Plus' one-of-a-kind internals and slap a Raspberry Pi in there like some DIY slacker, Noki went… a different path.

47 minutes and 25 seconds of a tour-de-force of modern maker technology.

Noki 3D-printed his own Macintosh, the "Brewintosh." I would like you to consider what you think that last sentence means and then wipe your expectations clean. I have watched the entire 48-minute journey of Noki's Brewintosh, which is both very soothing on some ASMR-adjacent gut level and also low-key maddening for the way it plays down all the individual accomplishments along the way. Any one of the Brewintosh's pieces would be my entire weekend, and my spouse would not enjoy my mood while I was sunk into it.

Read 5 remaining paragraphs | Comments

We’ve got a new trailer for Furiosa: A Mad Max Saga. Verdict: Not mediocre

Check out the latest trailer for Furiosa: A Mad Max Saga, starring Anya Taylor-Joy.

We got the first trailer for the spinoff prequel film Furiosa: A Mad Max Saga in December, starring Anya Taylor-Joy as the younger incarnation of the character immortalized by Charlize Theron in Mad Max: Fury Road. We're now just a couple of months away from the film's much-anticipated release—i.e., the perfect time to drop a second trailer to keep that anticipation high.

(Spoilers for Fury Road below.)

As previously reported, we met Furiosa early on in Fury Road, working logistics for Immortan Joe (the late Hugh Keays-Byrne), who charged her with ferrying oil from Gas Town to his Citadel with the help of a small crew of War Boys and one of the war rigs—basically tractor trailer trucks souped up with armor and novel weaponry. Furiosa stole the war rig instead, taking Joe's five wives with her.

Read 5 remaining paragraphs | Comments

Facebook, Instagram may cut fees by nearly 50% in scramble for DMA compliance

Facebook, Instagram may cut fees by nearly 50% in scramble for DMA compliance

Enlarge (credit: NurPhoto / Contributor | NurPhoto)

Meta is considering cutting monthly subscription fees for Facebook and Instagram users in the European Union nearly in half to comply with the Digital Market Act (DMA), Reuters reported.

During a day-long public workshop on Meta's DMA compliance, Meta's competition and regulatory director, Tim Lamb, told the European Commission (EC) that individual subscriber fees could be slashed from 9.99 euros to 5.99 euros. Meta is hoping that reducing fees will help to speed up the EC's process for resolving Meta's compliance issues. If Meta's offer is accepted, any additional accounts would then cost 4 euros instead of 6 euros.

Lamb said that these prices are "by far the lowest end of the range that any reasonable person should be paying for services of these quality," calling it a "serious offer."

Read 22 remaining paragraphs | Comments

Nvidia unveils Blackwell B200, the “world’s most powerful chip” designed for AI

The GB200 "superchip" covered with a fanciful blue explosion.

Enlarge / The GB200 "superchip" covered with a fanciful blue explosion. (credit: Nvidia / Benj Edwards)

On Monday, Nvidia unveiled the Blackwell B200 tensor core chip—the company's most powerful single-chip GPU, with 208 billion transistors—which Nvidia claims can reduce AI inference operating costs (such as running ChatGPT) and energy consumption by up to 25 times compared to the H100. The company also unveiled the GB200, a "superchip" that combines two B200 chips and a Grace CPU for even more performance.

The news came as part of Nvidia's annual GTC conference, which is taking place this week at the San Jose Convention Center. Nvidia CEO Jensen Huang delivered the keynote Monday afternoon. "We need bigger GPUs," Huang said during his keynote. The Blackwell platform will allow the training of trillion-parameter AI models that will make today's generative AI models look rudimentary in comparison, he said. For reference, OpenAI's GPT-3, launched in 2020, included 175 billion parameters. Parameter count is a rough indicator of AI model complexity.

Nvidia named the Blackwell architecture after David Harold Blackwell, a mathematician who specialized in game theory and statistics and was the first Black scholar inducted into the National Academy of Sciences. The platform introduces six technologies for accelerated computing, including a second-generation Transformer Engine, fifth-generation NVLink, RAS Engine, secure AI capabilities, and a decompression engine for accelerated database queries.

Read 8 remaining paragraphs | Comments

Thomas Stafford, who flew to the Moon and docked with Soyuz, dies at 93

Apollo commander Tom Stafford (left) with Soyuz commander Alexei Leonov during the Apollo-Soyuz mission in July 1975.

Enlarge / Apollo commander Tom Stafford (left) with Soyuz commander Alexei Leonov during the Apollo-Soyuz mission in July 1975. (credit: NASA)

Former NASA astronaut Thomas Stafford, a three-star Air Force general known for a historic handshake in space with a Soviet cosmonaut nearly 50 years ago, died Monday in Florida. He was 93.

Stafford was perhaps the most accomplished astronaut of his era who never walked on the Moon. He flew in space four times, helping pilot the first rendezvous with another crewed spacecraft in orbit in 1966 and taking NASA's Apollo lunar landing craft on a final test run before Neil Armstrong and Buzz Aldrin set foot on the Moon in 1969.

By his own account, one of the greatest moments in Stafford's career came in 1975, when he commanded the final Apollo mission—not to the Moon but to low-Earth orbit—and linked up with a Russian Soyuz spacecraft carrying two Soviet cosmonauts. The Apollo-Soyuz Test Project (ASTP) planted the seeds for a decades-long partnership in space between the United States and Russia, culminating in the International Space Station, where US and Russian crews still work together despite a collapse in relations back on Earth.

Read 42 remaining paragraphs | Comments

The Super Mario Maker community faces its final boss

"Trimming the Herbs," mapped above, is all that stands between  "Team 0%" and its ultimate goal of clearing every <em>Super Mario Maker</em> level.

Enlarge / "Trimming the Herbs," mapped above, is all that stands between "Team 0%" and its ultimate goal of clearing every Super Mario Maker level. (credit: Is SMM Beaten Yet?)

As of late 2017, there were almost 85,000 "uncleared" levels in the original Wii U Super Mario Maker (SMM)—levels that had never been beaten by anyone except for their original uploaders. As of this writing, a group of persistent players gathered under the banner of "Team 0%" has spent years narrowing the list of uncleared levels to a single entry—a devious, Super Mario World-styled Bob-omb bounce-and-throw gauntlet named "Trimming the Herbs" (the second-to-last uncleared level went down on Thursday, March 14, as noted on the excellent "Is SMM Beaten Yet?" tracker).

Given enough time, Team 0% would undoubtedly be able to bring down SMM's "final boss," as it were. But the collective effort to finally and completely "beat" SMM has an external deadline: April 8, the day Nintendo has announced that it plans to finally shut down the aging Wii U's gameplay servers.

The next three weeks will determine whether Team 0% can live up to its moniker or if this one final level will leave the team just short of its ultimate achievement. "I’d never think we would be this close to actually achieving this goal," Team 0% founder Jeffie told Ars Technica recently. "How often does a community of gamers do something like this?"

Read 22 remaining paragraphs | Comments

USB hubs, printers, Java, and more seemingly broken by macOS 14.4 update

USB hubs, printers, Java, and more seemingly broken by macOS 14.4 update

Enlarge

A couple of weeks ago, Apple released macOS Sonoma 14.4 with the usual list of bug fixes, security patches, and a couple of minor new features. Since then, users and companies have been complaining of a long list of incompatibilities, mostly concerning broken external accessories like USB hubs and printers but also extending to software like Java.

MacRumors has a good rundown of the list of issues, which has been steadily getting longer as people have run into more problems. It started with reports of malfunctioning USB hubs, sourced from users on Reddit, the Apple Support Communities forums, and elsewhere—USB hubs built into various displays stopped functioning for Mac users after the 14.4 update.

Other issues surfaced in the days after people started reporting problems with their USB hubs, including some instances of broken printer drivers, unexpected app crashes for some Java users, and problems launching apps that rely on the PACE anti-piracy software (and iLok hardware dongles) to authenticate.

Read 4 remaining paragraphs | Comments

Apple may hire Google to power new iPhone AI features using Gemini—report

A Google

Enlarge (credit: Benj Edwards)

On Monday, Bloomberg reported that Apple is in talks to license Google's Gemini model to power AI features like Siri in a future iPhone software update coming later in 2024, according to people familiar with the situation. Apple has also reportedly conducted similar talks with ChatGPT maker OpenAI.

The potential integration of Google Gemini into iOS 18 could bring a range of new cloud-based (off-device) AI-powered features to Apple's smartphone, including image creation or essay writing based on simple prompts. However, the terms and branding of the agreement have not yet been finalized, and the implementation details remain unclear. The companies are unlikely to announce any deal until Apple's annual Worldwide Developers Conference in June.

Gemini could also bring new capabilities to Apple's widely criticized voice assistant, Siri, which trails newer AI assistants powered by large language models (LLMs) in understanding and responding to complex questions. Rumors of Apple's own internal frustration with Siri—and potential remedies—have been kicking around for some time. In January, 9to5Mac revealed that Apple had been conducting tests with a beta version of iOS 17.4 that used OpenAI's ChatGPT API to power Siri.

Read 5 remaining paragraphs | Comments

Fujitsu says it found malware on its corporate network, warns of possible data breach

Par : Dan Goodin
Fujitsu says it found malware on its corporate network, warns of possible data breach

Enlarge (credit: Getty Images)

Japan-based IT behemoth Fujitsu said it has discovered malware on its corporate network that may have allowed the people responsible to steal personal information from customers or other parties.

“We confirmed the presence of malware on several of our company's work computers, and as a result of an internal investigation, it was discovered that files containing personal information and customer information could be illegally taken out,” company officials wrote in a March 15 notification that went largely unnoticed until Monday. The company said it continued to “investigate the circumstances surrounding the malware's intrusion and whether information has been leaked.” There was no indication how many records were exposed or how many people may be affected.

Fujitsu employs 124,000 people worldwide and reported about $25 billion of revenue in its fiscal 2023, which ended at the end of last March. The company operates in 100 countries. Past customers include the Japanese government. Fujitsu’s revenue comes from sales of hardware such as computers, servers, and telecommunications gear, storage systems, software, and IT services.

Read 3 remaining paragraphs | Comments

Elon Musk’s xAI releases Grok source and weights, taunting OpenAI

An AI-generated image released by xAI during the launch of Grok

Enlarge / An AI-generated image released by xAI during the open-weights launch of Grok-1. (credit: xAI)

On Sunday, Elon Musk's AI firm xAI released the base model weights and network architecture of Grok-1, a large language model designed to compete with the models that power OpenAI's ChatGPT. The open-weights release through GitHub and BitTorrent comes as Musk continues to criticize (and sue) rival OpenAI for not releasing its AI models in an open way.

Announced in November, Grok is an AI assistant similar to ChatGPT that is available to X Premium+ subscribers who pay $16 a month to the social media platform formerly known as Twitter. At its heart is a mixture-of-experts LLM called "Grok-1," clocking in at 314 billion parameters. As a reference, GPT-3 included 175 billion parameters. Parameter count is a rough measure of an AI model's complexity, reflecting its potential for generating more useful responses.

xAI is releasing the base model of Grok-1, which is not fine-tuned for a specific task, so it is likely not the same model that X uses to power its Grok AI assistant. "This is the raw base model checkpoint from the Grok-1 pre-training phase, which concluded in October 2023," writes xAI on its release page. "This means that the model is not fine-tuned for any specific application, such as dialogue," meaning it's not necessarily shipping as a chatbot. But it will do next-token prediction, meaning it will complete a sentence (or other text prompt) with its estimation of the most relevant string of text.

Read 9 remaining paragraphs | Comments

Once “too scary” to release, GPT-2 gets squeezed into an Excel spreadsheet

An illustration of robots sitting on a logical block diagram.

Enlarge (credit: Getty Images)

It seems like AI large language models (LLMs) are everywhere these days due to the rise of ChatGPT. Now, a software developer named Ishan Anand has managed to cram a precursor to ChatGPT called GPT-2—originally released in 2019 after some trepidation from OpenAI—into a working Microsoft Excel spreadsheet. It's freely available and is designed to educate people about how LLMs work.

"By using a spreadsheet anyone (even non-developers) can explore and play directly with how a 'real' transformer works under the hood with minimal abstractions to get in the way," writes Anand on the official website for the sheet, which he calls "Spreadsheets-are-all-you-need." It's a nod to the 2017 research paper "Attention is All You Need" that first described the Transformer architecture that has been foundational to how LLMs work.

Anand packed GPT-2 into an XLSB Microsoft Excel binary file format, and it requires the latest version of Excel to run (but won't work on the web version). It's completely local and doesn't do any API calls to cloud AI services.

Read 12 remaining paragraphs | Comments

Security footage of Boeing repair before door-plug blowout was overwritten

NTSB Chair Jennifer Homendy sitting in front of a microphone while testifying at a Senate hearing.

Enlarge / National Transportation Safety Board Chair Jennifer Homendy testifies about the Boeing door-plug investigation before the Senate Commerce, Science, and Transportation Committee on March 6, 2024, in Washington, DC. (credit: Getty Images | Kevin Dietsch )

A government investigation into a Boeing 737 Max 9 plane's door-plug blowout has been hampered by a lack of repair records and security camera footage, the National Transportation Safety Board's chair told US senators. Boeing was "unable to find the records" and told the NTSB that the security camera footage was overwritten.

"To date, we still do not know who performed the work to open, reinstall, and close the door plug on the accident aircraft," NTSB Chair Jennifer Homendy wrote Wednesday in a letter to leaders of the Senate Commerce, Science, and Transportation Committee. "Boeing has informed us that they are unable to find the records documenting this work. A verbal request was made by our investigators for security camera footage to help obtain this information; however, they were informed the footage was overwritten. The absence of those records will complicate the NTSB's investigation moving forward."

A Boeing spokesperson told Ars today that under the company's standard practice, "video recordings are maintained on a rolling 30-day basis" before being overwritten. The NTSB's preliminary report on the investigation said the airplane was delivered to Alaska Airlines on October 31, 2023, after a repair in a Boeing factory. On January 5, the plane was forced to return to Portland International Airport in Oregon when a passenger door plug blew off the aircraft during flight.

Read 10 remaining paragraphs | Comments

Lawsuit opens research misconduct report that may get a Harvard prof fired

Image of a campus of red brick buildings with copper roofs.

Enlarge / Harvard's got a lawsuit on its hands. (credit: Glowimages)

Accusations of research misconduct often trigger extensive investigations, typically performed by the institution where the misconduct allegedly took place. These investigations are internal employment matters, and false accusations have the potential to wreck someone's career needlessly. As a result, most of these investigations are kept completely confidential, even after their completion.

But all the details of a misconduct investigation performed by Harvard University became public this week through an unusual route. The professor accused of misconduct, Francesca Gino, had filed a multi-million dollar lawsuit, targeting both Harvard and a team of external researchers who had accused her of misconduct. Harvard submitted its investigator's report as part of its attempt to have part of the suit dismissed, and the judge overseeing the case made it public.

We covered one of the studies at issue at the time of its publication. It has since been retracted, and we'll be updating our original coverage accordingly.

Read 15 remaining paragraphs | Comments

Walmart resurrects the M1 MacBook Air as an entry-level $699 laptop

The M1 MacBook Air returns as a Walmart budget laptop.

Enlarge / The M1 MacBook Air returns as a Walmart budget laptop. (credit: Walmart)

Apple no longer sells the M1 MacBook Air as of earlier this month, discontinuing it and offering the M2 version of the Air as its entry-level model instead. But it looks like the M1 Air may live on, at least for a while—US retailer Walmart made a point of announcing today that it would carry and sell the M1 Air in its online store and at “select” retail locations for a much-lowered price of $699.

This is lower than the $999 that Apple was asking for the laptop just a few weeks ago, and it's lower than the $759 that the M1 Air goes for in Apple’s refurbished store. These prices are all for the base configuration of the M1 Air, with 8GB of memory and 256GB of storage. Walmart offers all three color finishes for the M1 Air—silver, gold, and space gray—but doesn’t directly sell any versions with more RAM or storage.

This isn’t the Air config we’d recommend to most enthusiasts—for them, an M3 Air or a refurbished M2 model with more RAM and storage come with enough benefits to be worth the extra cost. But it is a surprisingly low price for what remains a solid entry-level laptop, especially given that Walmart doesn’t offer any other Macs in its stores (other Macs on Walmart’s website are available from third-party sellers).

Read 3 remaining paragraphs | Comments

“Really bad timing”: Meta is killing misinformation analysis tool on August 14

The Facebook app logo is being displayed on a mobile phone in Colombo, Sri Lanka, on March 5, 2024.

Enlarge (credit: Getty)

Meta is discontinuing data analysis tool CrowdTangle on August 14. The closure will come three months ahead of the next US presidential election and three years after it was reported that the platform used for spotting misinformation on Facebook and Instagram was causing internal strife.

Meta acquired CrowdTangle in 2016. CrowdTangle has been used by researchers, reporters, and government officials to identify trends about conspiracies and other forms of misinformation spreading through Facebook. Meta is going to replace CrowdTangle with a technology currently under development called Meta Content Library, but it will only be available to academic and nonprofit researchers. For-profit organizations, like many news organizations, will lose access, as The Wall Street Journal points out.

Previously, CrowdTangle had some features available to the public, like Live Displays, which tracked how people discussed trending topics on certain social media channels like Facebook Pages. Journalists working at for-profit news outlets were able to apply for access to the full CrowdTangle service, as were publishers, including music labels, content creators, and public figures.

Read 13 remaining paragraphs | Comments

Raspberry Pi-powered AI bike light detects cars, alerts bikers to bad drivers

Copilot mounted to the rear of a road bike

(credit: Velo AI)

Whether or not autonomous vehicles ever work out, the effort put into using small cameras and machine-learning algorithms to detect cars could pay off big for an unexpected group: cyclists.

Velo AI is a firm cofounded by Clark Haynes and Micol Marchetti-Bowick, both PhDs with backgrounds in robotics, movement prediction, and Uber's (since sold-off) autonomous vehicle work. Copilot, which started as a "pandemic passion project" for Haynes, is essentially car-focused artificial intelligence and machine learning stuffed into a Raspberry Pi Compute Module 4 and boxed up in a bike-friendly size and shape.

A look into the computer vision of the Copilot.

While car-detecting devices exist for bikes, including the Garmin Varia, they're largely radar-based. That means they can't distinguish between vehicles of different sizes and only know that something is approaching you, not, for example, how much space it will allow when passing.

Read 5 remaining paragraphs | Comments

New iPads may be coming soon, but they won’t change the awkward spot the iPad is in

Apple's $329 9th-gen iPad is over 2 years old and the last model to use Lightning or the old Apple Pencil.

Enlarge / Apple's $329 9th-gen iPad is over 2 years old and the last model to use Lightning or the old Apple Pencil. (credit: Andrew Cunningham)

After leaving the iPad lineup untouched for the entirety of 2023, Apple is reportedly preparing to overhaul all of its tablets within the next few weeks, according to Bloomberg's Mark Gurman. We should see major iPad Pro and iPad Air refreshes "around the end of March or in April," says Gurman, along with a special build of iOS 17.4 that adds support for the new hardware.

We'll talk about the specifics of these iPad rumors momentarily, but reading about them got me thinking about what it would take to make me consider an upgrade for either of the iPads currently rolling around my house—a third-generation iPad Air that is currently used mostly for watching Octonauts and assembling Super Mario Lego sets, and a fifth-generation M1 Air that I use mostly for reading and browsing.

At least for me, the answer isn't "new hardware." After a brief stint a few years ago using the iPad as a focused writing device, I've mostly relegated it to tablet-y content consumption, leaving behind the cottage industry of enthusiasts who keep trying to come up with workarounds to make the iPad into a Mac. To replace an iPad at this point, I would either need one of them to break or for Apple to dramatically change what the high-end iPads are capable of.

Read 12 remaining paragraphs | Comments

Apple to allow iOS app installs from websites, but small devs don’t qualify

App icons displayed on an iPhone screen.

Enlarge (credit: Getty Images | NurPhoto )

Apple will let European app developers distribute iPhone and iPad applications to users directly from a website, instead of through an app store. It's the latest app-installation option announced by Apple as it seeks to comply with new European rules, but this one will only be available to developers who had an app installed by over 1 million users the previous year.

In an announcement today, Apple said it plans to introduce "a new way to distribute apps directly from a developer's website." The Web Distribution option will become available after a software update "later this spring," letting developers "distribute their iOS apps to EU users directly from a website owned by the developer."

"Apple will provide authorized developers access to APIs that facilitate the distribution of their apps from the web, integrate with system functionality, back up and restore users' apps, and more," the company said.

Read 13 remaining paragraphs | Comments

Cicadas pee in jets, not droplets. Here’s why that’s kinda weird.

Cicadas' unique urination unlocks new understanding of fluid dynamics. Credit: Georgia Tech (Saad Bhamla/Elio Challita).

Cicadas might be a mere inch or so long, but they eat so much that they have to pee frequently, emitting jets of urine, according to a new paper published in the Proceedings of the National Academy of Sciences. This is unusual, since similar insects are known to form more energy-efficient droplets of urine instead of jets. Adult cicadas have even been known to spray intruders with their anal jets—a thought that will certainly be with us when "double brood" cicada season begins in earnest this spring.

The science community has shown a lot of interest in the fluid dynamics of sucking insects but not as much in how they eliminate waste, according to Georgia Tech's Saad Bhamla (although Leonardo da Vinci was fascinated by jet behavior and the role of fluid cohesion in drop formation). Yet, this is a critical function for any organism's ecological and metabolic regulation. So Bhamla's research has focused on addressing that shortcoming and challenging what he believes are outdated mammal-centric paradigms that supposedly govern waste elimination in various creatures.

For instance, last year, his team studied urination in the glassy-winged sharpshooter. The sharpshooter drinks huge amounts of water, piercing a plant's xylem (which transports water from the roots to stems and leaves) to suck out the sap. So sharpshooters pee frequently, expelling as much as 300 times their own body weight in urine every day. Rather than producing a steady stream of urine, sharpshooters form drops of urine at the anus and then catapult those drops away from their bodies at remarkable speeds, boasting accelerations 10 times faster than a Lamborghini.

Read 9 remaining paragraphs | Comments

❌