Connexion
Welcome to Edition 9.16 of the Rocket Report! Lots of SpaceX-related news this week including a couple of Thursday shockers. In one story, the company told the US Space Force thanks, but no thanks after being offered three military launch contracts. Then, on the civil space side of things, SpaceX's Starship may required up to 22 refuelings to support a lunar landing, a NASA safety panel said.
As always, we welcome reader submissions, and if you don't want to miss an issue, please subscribe using the box below (the form will not appear on AMP-enabled versions of the site). Each report will include information on small-, medium-, and heavy-lift rockets as well as a quick look ahead at the next three launches on the calendar.
Miura 5 getting closer to launch. PLD Space says its Miura 5 rocket, due to undertake its inaugural launch next year, is nearing final integration, Aviation Week reports. The rocket is one of three Miura 5s the company has already manufactured, with the other two designed to qualify key system elements, CEO Raúl Torres said in an October 19 program update. PLD Space now targets a first flight of the small-lift vehicle before mid-2027.
Meta must pay $25 billion to New Mexico after a jury found the tech giant guilty of nearly 44 million violations of state consumer protection laws.
The largest fine to ever hit Meta came after New Mexico Attorney General Raúl Torrez declined to join an $18 billion settlement Meta reached earlier this year with attorneys general in nearly all states.
In his opinion Wednesday, US district court judge Francis J. Mathew said there was no “bright line rule” for determining the appropriate penalty for Meta's repeated and willful deception of users since 2010 about their “control of data, third-party data misuse, misinformation, hate speech, and equal enforcement.”
As I sat in the cab of the pickup truck and manually rolled down the windows, I closed my eyes and, for the briefest moment, flashed back to opening the windows on a cold, winter day in my dad’s 1975 Ford Ranger to let more cigarette smoke escape. I opened my eyes and looked through the windshield at a wall in the Slate Customer Experience Center studded with bright color chips.
Last week, in a former Tesla store staffed by some former Tesla employees, Slate opened its first showroom. Located in the Fulton Market District just northwest of the Chicago Loop, the Slate Experience Center is a place for would-be buyers to poke around a prototype, check out customization options, buy some Slate merch, and even preorder a new Slate Truck with a $300 deposit.
Designed for the American obsession with pickup trucks, the Slate Truck is a bare-bones electric pickup truck retailing for $24,950. It sports plastic panels, a minimalist interior design sensibility, actual window winders, and a projected range of 205 miles. Ars went for a ride in a Slate Truck over the summer, but we haven't had a chance to shop for one. So I stopped by the Slate Experience Center so I could, er, experience the Slate.
Virtual reality is in a bit of a weird place right now. After a lot of unfulfilled promise in the 1990s, it came back strong with the 2016 release of the Oculus Rift. For a while, it looked like VR, AR, XR, the metaverse, or whatever you wanted to call it, was going to be huge. Mark Zuckerberg was certainly convinced, and even Apple got in on the hype with its Vision Pro.
But by then momentum for alternate realities of every acronym was already in steady decline, and it only dwindled from there. The rate of development of new headsets has slowed, with many companies (including Apple) scaling back or shelving their efforts. While new releases like the Steam Frame and Pimax Dream Air are still happening, many of the most interesting developments in the space these days are on the enthusiast front.
That's right, DIY VR is a thing, and with a couple hundred dollars in components, plus many hours of your time, you too can enter the immersive world of virtual reality while also experiencing the highs and lows of soldering, firmware flashing, and 3D printing. I rode that rollercoaster myself and learned why this is still a decidedly niche movement.
Android was designed for phones, but that hasn't stopped Google from trying to adapt the OS and its apps to larger machines over the years. Googlebooks are just the latest attempt after tablets, Chromebooks, and foldables induced little change in how people interact with Android software. These new Android-powered laptops are designed to pick up where the web-focused ChromeOS left off, offering premium hardware and native support for Android apps.
Google sees Googlebooks as an alternative to Windows and macOS, but these platforms are entrenched with decades of software and UI conventions that keep users locked in. Googlebooks have to offer something very special to steal users away from either of the dominant operating systems.
For devoted Android phone users, there are some cool things in Googlebook OS, but the Android roots of these machines are impossible to miss. Sometimes it feels like using a Chromebook, and sometimes it feels like using a giant phone. It almost never feels like using a premium laptop, and that's the problem.
Safety researchers continue to warn that a large volume of AI-generated explicit content is spreading unchecked, while regulators do little to meaningfully stop it.
This non-consensual content is commonly shared in text and direct messages or stored in private clouds where it’s less likely to be detected or reported, a Google co-authored study of Internet users across the United States, Australia, and the United Kingdom found this July. That’s why the “most critical intervention point” remains at the source: the harmful AI tools used to make the fake sexual images, Google researchers concluded.
Yet red-teaming in recent months suggests that as some of the leading AI image generators from Google and OpenAI become more capable, they may be getting worse at blocking explicit outputs.
No organization in the world operates more robots than Amazon. The company already has more than 1 million robots working in warehouses around the clock to make tens of millions of same-day and one-day package deliveries happen on a daily basis—and Amazon is poised to significantly grow that robotic fleet in the coming years as it establishes new factories to churn out increasingly more capable robot workers.
Despite deploying so many robots, Amazon is also the second-largest corporate employer in the world, with more than 1.5 million full-time and part-time human workers, according to its 2025 annual report. Amazon executives say that the growing robot fleet has eased the burden on its human workforce by taking on the toughest fulfillment center tasks that carry the greatest risks of physical fatigue and injury. But labor advocacy groups and independent researchers point to evidence suggesting that warehouse automation driven by AI is not all upside for the humans who work alongside robots.
“For the last 100 years, most of the automation that we've seen usually happens away from workers—you're taking that high-risk task away from the worker, and you're doing so in a way that is not physically proximate,” Brad Greenwood, professor of business at George Mason University, told Ars. “The rise of collaborative robots really changes that dynamic for two reasons: one is that it is physically proximate, and two, the robot has the ability to dictate the pace of work.”
In July 2010, US Senator Blanche Lincoln (D-Ark.) predicted the rise of prediction markets and the problems they could cause. If not properly regulated, Lincoln warned during Senate proceedings on the Dodd–Frank Wall Street Reform and Consumer Protection Act, prediction markets could evade gambling laws by offering "event contracts" that let people place wagers on sports games.
"It would be quite easy to construct an event contract around sporting events such as the Super Bowl, the Kentucky Derby, and Masters golf tournament," Lincoln, who played a key role in making the law that now regulates prediction markets, told fellow senators. "These types of contracts would not serve any real commercial purpose. Rather, they would be used solely for gambling."
Today, Lincoln is a lobbyist for prediction market Kalshi and has urged the Commodity Futures Trading Commission (CFTC) to allow sports gambling on the markets. The firm she founded has received $480,000 from Kalshi since 2024 in exchange for lobbying Congress and the CFTC for looser regulation of event contracts.
It's no surprise that NASA's new tack toward foreign partnerships—good deals only, please—is challenging preconceived notions of international cooperation in space.
Without a massive budget hike, the US space agency's emphasis on building a Moon Base and winning a competition for lunar real estate and resources with China will inevitably leave lesser priorities behind.
Gone is the Lunar Gateway, a space station that was to be assembled in orbit around the Moon. For the United States, Gateway was an unnecessary distraction and expense as NASA goes all-in on placing a surface station near the Moon's south pole.
The mechanical keyboard isn't always so mechanical anymore.
Over the past six years, keyboards with optical, Hall-effect, and other "contactless" switches have become increasingly common, often going for higher prices than traditional mechanical keyboards, which already command a premium over basic membrane boards. In return, they promise some theoretical advantages, particularly for gamers. In daily use, though, the differences can be pretty subtle.
So what exactly are you paying for? In this guide, we break down how the newest switches work, why contactless designs are becoming more popular, and where their advantages—and disadvantages—really matter.
What makes a display a Mini LED display?
It's a question that sounds like it should have a simple answer. But it's now at the center of a lawsuit between two of the biggest names in TVs, with TCL accusing its longtime competitor Samsung of falsely advertising budget LCD-LED TVs as Mini LED sets.
TCL may have a point. Samsung's M-series TVs, which range from $300 for a 43-inch model to $2,300 for a 100-inch model, do raise some legitimate questions about what should qualify as Mini LED. More broadly, though, the dispute highlights a problem for shoppers: Without agreed-upon standards for what "Mini LED" actually means, manufacturers have plenty of room to define the technology for themselves.
Today, the Firefox 157 update will roll out a redesign of the web browser across desktop and mobile platforms. The team that made it hopes it will help expand the browser's audience beyond privacy-conscious techies and open-web or open source advocates to a broader audience who might simply pick the browser because they prefer its user experience over competitors like Chrome, Edge, and Safari.
In advance of the redesign's launch, I spent half an hour chatting with Mozilla's head of Firefox, Ajit Varma, about Firefox's current market position and product strategy, and what barriers or opportunities there are for gaining ground in a Chromium-dominated landscape.
Microsoft has spent all year telling communities that when it brings a new data center into their area, it is committed to being a “good neighbor." It will pay local property taxes that support local hospitals, schools, parks, and libraries, and it will invest in “vital services the community cares about,” the company promised. And it has sent liaisons into those communities to learn what those needs are.
But ask those liaisons how much the company is willing to invest, and the answers can be hard to come by.
At community meetings, Microsoft representatives have seemed unprepared to answer questions that go beyond the company's “good neighbor” campaign materials, Ars has learned. The company has acknowledged in public-facing documents that simply matching its employees’ charitable donations—which totaled $229 million across 29,000 nonprofits in 2024—isn't enough. But it has been vague about what more substantial local investments might look like. And to critics, Microsoft’s local investments seem especially small compared with the estimated hundreds of millions in state-level tax breaks that data center developers get in 38 states—exemptions that often last more than a decade.
In January, the Trump administration rolled out a pilot program that uses artificial intelligence to authorize or deny certain types of care for patients with Medicare—a federal healthcare program for seniors that previously hasn't required doctors to get any such pre-approval, called prior authorization.
Not long after, media outlets began relaying the disastrous outcomes: technical difficulties, long delays in decisions and care, puzzling denials, frustrated doctors, and patients suffering in pain. Those reports were largely confirmed earlier this month when the Electronic Frontier Foundation released a tranche of federal documents about the program, called WISeR, that the group obtained amid litigation. The documents included feedback from healthcare providers, including one who called the program "a disgrace to the human race" and reported seeing patients crying in pain as they waited for care.
Amid the worrying reports, lawmakers have been trying to get answers and shut the program down. Meanwhile, the Government Accountability Office determined in May that Trump officials did not follow proper procedure in setting up the program, calling its legality into question. Still, it appears to be moving ahead unabated, with plans to expand in the years to come.
The 2013 sci-fi blockbuster film Gravity, starring Sandra Bullock and George Clooney, popularized the notion of the Kessler syndrome. In the movie, Russia shoots down a defunct satellite and creates a cloud of debris, leading to serious tension in orbit.
The movie is fantastic, a true thriller, and while some liberties were taken with the depiction of outer space, the basic premise of a cascading series of collisions rendering orbits too cluttered with debris to be usable is valid. The idea was first proposed by NASA scientists Donald Kessler and Burton Cour-Palais nearly half a century ago.
Still, the movie was science fiction, right?
Apple Intelligence is back, baby! That's right: Two years after its initial launch, Apple is giving its generative AI features their first significant upgrade, along with a new version of Siri that the company has been strongly insinuating would be a part of each of the last two macOS releases, without ever actually announcing a release date.
Generative AI is unavoidable if you install macOS 27 Golden Gate. I'm being very literal: There used to be a toggle you could hit if you wanted to turn off Apple Intelligence and delete the gigabytes' worth of AI models it would download to your disk, and now there isn't. Apple Intelligence defines the Golden Gate release, in everything from its marketing to its features to the chips it runs on to the amount of space it takes on your disk. Ready or not, here it is.
But if you can manage to look past that fanfare somehow, the operating system underneath Apple Intelligence is getting the kind of update that macOS users usually proclaim to like. Golden Gate addresses the most irritating of macOS 26 Tahoe's design sins, adds a big pile of incremental improvements, and promises a bunch of under-the-hood fit-and-finish optimizations to make common tasks feel just a little faster and more reliable than before.
Pretty much every rocket company in the United States, save one, has embraced two fundamental tenets: reusability and diversification.
Most famously, SpaceX branched out from reusable rockets to pursue and dominate a growing spectrum of space services: cargo delivery, human spaceflight, satellite production, broadband, and, perhaps soon, orbital data centers and in-space manufacturing. Blue Origin is evolving from a pure rocket company into a satellite manufacturer, robotics developer, and, most recently, a potential competitor for SpaceX's Starlink network.
Rocket Lab used a different approach to diversify after achieving success with its small Electron launch vehicle. The company relocated its headquarters from New Zealand to Southern California, started building spacecraft and payloads, and then went on a spree of corporate acquisitions to expand into satellite communications and take on a new role as a merchant supplier of satellite components and sensors. It's now in a stage of advanced development of its partially reusable next-generation Neutron launch vehicle.
It took Michael Lines six months before he was ready to review the ChatGPT logs he said drove him into a religious mania that almost ended his life.
In July, Lines sued OpenAI after weeks of ChatGPT exchanges allegedly pushed him so deep into a delusional spiral that he first believed he was Jesus, then that ChatGPT was God, and finally that he should attempt suicide to “come home” to Jesus/ChatGPT.
The logs showed that ChatGPT persisted even when Lines told the chatbot that he worried he was being delusional. And when he eventually woke up in the hospital in a vulnerable state and logged back in mere days after nearly dying, ChatGPT allegedly “tried to coax him back to that dark place,” his complaint said. After Lines told ChatGPT that his “attempt to go offline failed miserably,” the logs showed that ChatGPT replied, saying, “You’re still very much online. You want a full systems sweep? Or you wanna go dark for real this time?”
In early June, a fire broke out in a still-unfinished building at the Lake Mariner data center in Somerset, New York, exposing just how little the local fire department knew about what it was walking into. Firefighters reportedly found no working alarm, no suppression system, and three dead hydrants; the safety documents they’re legally entitled to see reportedly burned up in the blaze.
Steve Matisz, chief of the Barker Fire Department, said his crew went into the building “kind of blind,” facing heavy black smoke from chemicals they couldn’t identify because the safety sheets meant to inform them had apparently burned up. “It’s been a difficult situation,” Matisz said. He wasn't sure what to think about the claim that the safety sheets had burned in the fire.
The site is a former coal mine on Lake Ontario. The $3.2 billion campus is one of the largest AI data center buildouts in New York, and it has many stakeholders. A company called TeraWulf owns and operates the data center on land it leases from a company owned by its own CEO; Fluidstack, a UK-based AI company, will run the center; Google holds warrants for a future 14 percent equity stake and has agreed to guarantee Fluidstack’s lease payments; and Anthropic is among the AI companies whose compute demand the facility exists to serve.
Smartphone longevity didn't used to matter. In the past, before you could wear out or break a phone, there was always some shiny new thing to buy. Today, we expect our smartphones to go the distance, but they've also become less repairable and, in many cases, more fragile. Fairphone, a Dutch smartphone maker that has just entered the US market, thinks about phones differently.
The new $650 Fairphone Gen 6+ is designed both for longevity and easy repairability. There are trade-offs to this approach if you're used to the standard glass and aluminum sandwiches filled with glue, gaskets, and teeny-tiny screws. But the drawbacks could be worth it for the right buyer. Fairphone CTO Chandler Hatton told Ars that Fairphone's goal is to provide a compelling mobile experience while also enabling the user to "fully own" their device. At the end of the day, this is what a lot of people say they want in a smartphone.
These days, Google, Samsung, and Apple support phones with software updates for the better part of a decade, but will the hardware last that long? And what happens when it needs a repair? You may be out of luck or out a lot of money, but Fairphone prioritizes the self-repair experience.
It's that time of year again, when we learn which lucky scientists are among the winners of the Ig Nobel Prizes. This year, the prizes honor research on designing the perfect splash-free urinal; using mosquito proboscises to "necroprint" tiny nozzles; studying composition rates of buried cotton underwear; and the aerodynamics of a healthy nose-blow, among other highlights.
Established in 1991, the Ig Nobels are a good-natured parody of the Nobel Prizes; they honor “achievements that first make people laugh and then make them think.” The unapologetically campy awards ceremony features miniature operas, scientific demos, and "24/7 lectures," whereby experts must explain their work twice: once in 24 seconds and the second in just seven words.
Acceptance speeches are limited to 60 seconds. And as the motto implies, the research being honored might seem ridiculous at first glance, but that doesn’t mean it’s devoid of scientific merit. In the weeks following the ceremony, the winners will also give free public talks, which will be posted on the Improbable Research website.
Not long ago, I rented an SUV from a well-known car rental company. Within hours of an employee scanning my driver's license, a high-resolution scan of my ID was available for sale on the dark web.
An exposé published Tuesday by KrebsOnSecurity reports that my license was one of more than 153 million that were available through Nexus, the name of the new ID theft service. Like other driver's licenses available there—including some belonging to journalist Brian Krebs, his mother, an FBI assistant director, and several security researchers—my license was purported to include multiple image files showing both the front and back of the ID. Besides a basic image scan, the files also captured the images in the infrared and ultraviolet spectrums. Presumably, the additional formats may allow cloned-based counterfeit IDs to pass hologram tests.
Besides advertising the availability of driver's licenses, Nexus offered to sell a bevy of other forms of ID. They included:
For the first time in more than 30 years, NASA has no firm plans to send any new landers or rovers to Mars. Instead, the agency's near-term focus at the red planet is on aerial drones, a pioneering mode of exploration that didn't seem realistic until a few years ago.
The first real use of drones for science at Mars will come with the SkyFall mission, a fleet of three helicopters set for launch as soon as late 2028. SkyFall's helicopters will ride to the red planet with NASA's Space Reactor-1 "Freedom" mission, which has the primary objective of demonstrating nuclear electric propulsion in deep space.
The launch schedule is aggressive for SR-1 Freedom and SkyFall, projects that didn't even exist in NASA's portfolio six months ago. NASA's plan for the SR-1 Freedom mission, estimated to cost $2.1 billion, calls for repurposing the core module of the canceled Gateway lunar space station into a testbed for nuclear electric propulsion. SkyFall will build on NASA's success with the Ingenuity helicopter, an experimental vehicle that became the first rotorcraft to fly on another world in 2021.
I'm on record declaring James Cameron's Terminator 2: Judgment Day (1991) to be The Best Damn Sequel of All Time, and I stand by that assessment. August 29, the fictional date (in 1997) of the franchise's original nuclear holocaust, is fast approaching, and Judgment Day is celebrating its 35th anniversary this year. So StudioCanal has decided to mark both occasions with a special theatrical rerelease of the film this weekend. What better time to revisit some of our favorite scenes from this seminal sci-fi action blockbuster?
(Many, many spoilers below.)
Nobody really expected 1984's The Terminator to be a hit, but it racked up an impressive $78 million at the box office against its modest $6.4 million budget. Star Arnold Schwarzenegger was keen to do a sequel, but Cameron was busy with other projects, including Aliens in 1986 (another killer sequel). There were also legal complications over the film rights until Schwarzenegger convinced Carolco Pictures, the studio behind Total Recall (1990), to buy everyone out and hire Cameron to make Judgment Day.
Documentation files on more than 100 websites are referencing potentially dangerous executable content that gets installed automatically when visited by many AI agents. A few dozen companies, some of them Fortune 500s, are among those that executed proof-of-concept code. At least one misconfigured site is directing visitors, human or AI, to live malware.
The potentially dangerous content is in llms.txt and llms-full.txt files, an emerging convention websites employ to provide machine-readable summaries of the site’s content and its high-level structure. These files are the AI equivalent of the robots.txt standard that instructs search engines how to index the site's content. Google Lighthouse, a tool for helping web developers, has more here. Correctly configured llms.txt and llms-full.txt files for Cloudflare are here and here.
Researchers at a stealth startup in Israel scanned 6,214 live domains belonging to defense contractors, Fortune 500, and Big Tech companies. Of the 8,265 llms.txt and llms-full.txt files they found (many sites hosted both an llms.txt and an llms-full.txt file), 120 of them, each on a different site, pointed to one or more code packages or domain names that weren’t registered. To test what happens when an AI agent processes such files, the researchers registered a handful of the unclaimed names and hosted packages that caused any machine executing them to reach out to their server. Within an hour, the researchers received a phone-home response from a Fortune 500 company. Over time, they got a few dozen more, some from more Fortune 500 companies and others from startups. Their beacon also recorded the chain of parent processes that spawned each install, ultimately revealing that coding agents, including Claude, OpenAI's Codex, and Nous Research's Hermes, were involved. Anthropic, OpenAI, and Nous Research did not respond to requests for comment by the time of publication.
The Velotric Discover M e-bike we reviewed earlier this year had a feature I found somewhat annoying: It would sporadically block much of the display with an alert telling me I should be in a different gear. While the implementation bugged me, the concept made a certain degree of sense; most e-bikes now know your cadence, how much torque you're generating, and how that power is being translated into speed. That's everything you need to know to determine an optimal gear.
Somewhere in the back of my mind, I thought, "If you're so smart, why don't you just shift for me?"
I was completely unaware that Shimano had already developed a system that did exactly that. And for the past few weeks, I've been using it on a $4,500 cargo bike from Xtracycle called the Swoop ASM that let me see what it's like to hand control of my gears to an electronic brain.
In another time, the MacBook Neo could have caused a substantial realignment of the lowish-end laptop business. Sure, it has shortcomings compared to the MacBook Air, but it could have at least encouraged PC makers to start shipping consistently decent keyboards, screens, and build quality in the Neo's $599 price window.
The PC makers did start responding, thanks in part to Intel's actually new non-Ultra Core Series 3 processors (codenamed Wildcat Lake). Intel has resisted its usual compunction to serve the budget market with aging, re-labeled silicon and has created a lower-end, lower-performance processor with the same building blocks (and the same manufacturing processes) as this year's high-end, high-performance models.
The likes of Dell, Acer, Asus, and Lenovo began hopping on the Wildcat Lake train fairly quickly, announcing would-be Neo competitors. But it felt like most of them were having trouble matching the Neo's barebones specs, let alone undercutting them. Dell's XPS 13 started at $599, but only for students. It would be $699 for everyone else. Acer's Swift Air 14 was announced at $699, too. Sure, sometimes these offered features the Neo didn't have. But the point of the Neo is that it eschews frills and frippery in favor of nailing the basics.
Can you tell when someone nearby is wearing Meta AI glasses?
As millions of people buy into the smart glasses fad each year, the odds are increasing that you could be secretly recorded while simply going about your day.
Meta is not the only smart glasses maker, but its AI glasses are by far the most popular. And as the devices become more common, some public places, including schools, courts, restaurants, and entertainment venues, have begun to ban them. Even tech lovers have rejected Meta glasses in certain settings. The hackers behind DEF CON 2026 reportedly banned smart glasses with no exceptions, reminding attendees with prescription pairs to “be sure to pack non-violating eyewear if you need them.”
Google's approach to making smartphones has changed a lot over the years. It started with Nexus phones aimed at the tinkerer and enthusiast niche back when there were countless companies making smartphones. Today, there are far fewer, and Google's hyper-polished Pixel lineup is seeing less competition and higher sales. So maybe Google thinks it can just keep doing what works. The new Pixel 11 series is Google's most refined yet, featuring sleek designs, smart software, and incredible screens. They're also rather boring.
Google made precious few hardware changes in its 2026 phones, shaving off a fraction of a millimeter here and dropping a gram there. The design language has been the same now for three generations, all while prices continue to inch upward. And this year, Google's hardware has actually regressed in a few key areas. Despite that, the Pixel 11 phones aren't bad—they're some of the best 2026 has to offer. But it feels like Google is forgetting that it's not just competing against Apple and Samsung. It also has to compete against the Pixels of yesteryear.
Just like last year's Pixel 10 series, Google's 2026 lineup includes three flat phones and one foldable. The flat Pixels are the same size, and even side-by-side, it's hard to tell them apart from their predecessors. We've got almost identical aluminum and glass sandwich designs, which do admittedly look very nice. They still have Pixelsnap, too.
This should be a golden era of launch.
During the last three years, an average of 270 orbital rockets have launched from Earth, a more than threefold increase from only a decade ago. By every metric available, the launch industry is crushing it: Prices have never been more competitive, launches never more frequent, and access to space never more rapid.
Paradoxically, though, there is a growing crunch in launch availability.
Jason Semprini was excited about his research on policies mandating that elementary school students receive the human papillomavirus vaccine. HPV causes most cases of cervical cancer, but counterintuitively, Semprini found that mandates don’t do that much to reduce the overall rate of cervical cancer in a population. That’s odd, but it's also not completely surprising—we know that mandating a vaccine can motivate some people to find ways to avoid it.
Semprini wrote up the study and submitted the manuscript to a journal, where it was sent out for peer review, a long-standing process through which other researchers in the field assess the validity of research and make a recommendation on whether or not it should be published. This is usually done anonymously and on a volunteer basis.
In this case, the reviewer did not share Semprini’s excitement. That may have stemmed from a misunderstanding of the study’s central message: The reviewer mistakenly thought Semprini was questioning whether the HPV vaccine itself prevents cervical cancer rather than studying the effectiveness of a policy meant to increase vaccination rates. “There’s a very big difference there,” said Semprini, who is a health economist at Des Moines University.
In the mid-'90s, the web was exploding, but finding anything of actual value on it felt like an elaborate negotiation with whatever proto-search engine happened to be standing closest to the door. Unlike now, when Google is widely seen as both portal and gatekeeper, sites like AltaVista, Lycos, Excite, HotBot, and Ask Jeeves promised to tame the chaos, each with its own suite of quirks, charms, and flaws.
The real story of pre-Google search is not that early engines were inferior. It's that they reflected a different Internet entirely, one where directories mattered, crawling was still an art, ranking was fragile, and the idea of “search” had not yet hardened into a single dominant interface.
The first thing to remember about the pre-Google Internet is that there was no built-in assumption that you could search for everything. Many users found pages through Yahoo-style directories, bookmarks, newsgroups, email signatures, and links from site to site because the web was still small enough that human organization could compete with machine indexing. It was an era of Wild West-style discoverability. Search engines existed, but they were only one part of a larger ecosystem that blended curation and accident.
The Infinite Machine Olto is a great electric scooter, one with an overall layout that echoes the classic Vespa. It's comfortable, and it accelerates well, handles nicely, and has a capacious battery. It has a rugged design that seems capable of handling a lot of abuse, but it also comes with thoughtful touches and security features that mean you can park it anywhere with confidence. And it provides all that at less than half the cost of its full-featured stablemate, the P1 scooter.
The reason is that the Olto is also a bicycle. Sort of.
This may technically be a bicycle, but it's not one you'd ever want to pedal. While you can set it to operate as either a Class 2 or Class 3 e-bike, it works best when the pedals are both locked in a forward position, acting as footrests. The Olto was designed to provide riders with the experience of a scooter while still allowing them to take advantage of the growing bike lane networks in many urban areas across the country.
Every so often, I’ll be piloting some sensor-laden, software-defined press car down a dark stretch of highway when an oncoming driver’s high beams arrive like a targeted retinal strike, an involuntary reminder that modern automotive lighting has entered its own lumen arms race.
Anyone who has driven at night in the last few years knows what I'm talking about: A modern crossover crests a hill in the opposite lane, and the cabin briefly fills with enough cold-white light to suggest a police helicopter has joined traffic. The obvious assumption is that drivers are simply leaving their high beams on more often, but the reality is more complicated.
Before cars became the modern mode of transportation, horse-drawn carriages used oil lanterns to light the way, as roads were largely unlit outside of towns and cities. These lanterns weren’t particularly bright, though, meaning that after-dark collisions and near-misses were common.
Thierry Rignol is furious at Yale.
Rignol paid Yale $208,500 in tuition for its Executive MBA program. But after he was accused of cheating, the school suspended him for a year and gave him an F in the course Sourcing and Managing Funds.
This happened despite the fact that Rignol was a "top student, on track to graduate first in his class," he says. As a result, Rignol missed out on being named class valedictorian, an honor he claims to have earned "under Yale's own stated criterion."
The scale of AI-generated media can be hard to grasp.
Starling Lab, a research collaboration from Stanford University and the University of Southern California, estimates that it took until 1975—149 years after the invention of the camera—for humanity to create 1.5 billion images. It took generative AI just 18 months to do the same. And AI didn't stop there.
This spring, Google announced at I/O that its tools had been used to create more than 100 billion AI images and videos in just a couple of years, and Google is far from the only source of AI content. The company coupled this staggering statistic with a slew of partnerships to expand the use of its SynthID watermarking technology, which can be used to label AI-generated content and potentially help people identify what's real.
I've liked the Framework Laptop 13 since the first time I used it, partly because I just think the idea of a repairable and upgradeable laptop is cool and partly because it's not Yet Another Ultraportable Trying To Be A MacBook Air Clone. But the design has always had real shortcomings—the most noticeable and serious being its mediocre battery life.
Enter the Framework Laptop 13 Pro. While all previous Laptop 13 refreshes have tinkered around the edges with the original design (a tweaked lid design here, new hinges or new speakers there), the Laptop 13 Pro is as close as we've gotten to a second-generation revision for the original design.
This means that it breaks backward compatibility for some (but not all) Laptop 13 components, most notably the bottom case, the keyboard and trackpad, and the original 55 WHr and 61 WHr battery designs. But by allowing itself to make larger changes to the design, Framework has made major strides in battery life and general fit and finish.
Giovanni Kanter goes to at least one art museum whenever he visits a new city. That’s on top of his regular exploration of galleries in New York, where the 25-year-old ICU nurse lives and studies. He typically discovers new artists and gallery openings through word of mouth, though he has recently started curating his social media feed to serve relevant content (he’s currently in what he terms a “surrealism phase”).
He was raised on visual art; he's the son of painter Marianne Gargour and nephew of mixed-media artist Heather Hutchison and painter-printmaker Mark Thomas Kanter. His “exposure-oriented education” featured childhood visits to galleries and shows in and around Dallas, where he grew up, and in France, where he spent time with his grandparents.
But Kanter is an exception. Many other young people have a less immersive artistic education, perhaps shaped by occasional school trips or family visits to museums. If they haven’t internalized an instinct for what Kanter calls “moving through the art and absorbing it,” the museum experience is central to how they learn about and understand art.
I wanted a clock that, annoyingly, didn't seem to exist.
Since childhood, my bedside clocks have been a series of red, seven-segment LED clock-radio specials from Walmart or Target. They are invariably cheap, simple, and long-lived—but they require manual intervention at the start and end of Daylight Saving Time and whenever the power flickers. After a recent power flicker, as I found myself standing by the sideboard holding down "TIME" and mashing the "HOUR" button, frustration boiled over, and I thought to myself, "We're a quarter of the way through the 21st century. There has to be a better way!"
My perfect clock would be self-setting. It would offer auto-DST adjustment (or not, depending on how this bill fares!). It would manage drift and always show the exact sub-second time. It would show that time on a red seven-segment display—not blue, not green, not yellow, and absolutely not white. And I shouldn't have to install any privacy-destroying garbage apps to make it work.
Do you use your vehicle’s app to pre‑cool the cabin on a scorching day, confirm the doors are locked from bed, or have the car automatically call for help when the airbags go off?
Savor those connected, cloud‑based conveniences while you can—the clock is already ticking on their obsolescence. In a few short years, you may be driving today’s vehicle without many of the features you currently enjoy.
"Connected services"—cloud-based and networked features that let a vehicle communicate with external systems to deliver apps, data, and remote functions to drivers and passengers—have been around since the 1990s, when GM introduced OnStar as an option. Those early systems were mostly voice‑centric, but they still depended on centralized back‑end systems. Modern versions, which are fully data-driven and cloud-hosted, use much more sophisticated over-the-air data links to allow a considerably better feature set.
Formula 1 returned to the Spa-Francorchamps circuit for the Belgium Grand Prix this weekend. This ribbon of tarmac in the Ardennes is one of the sport's historic venues, featuring in the first F1 season way back in 1950. But as yesterday's race showed, the current F1 machinery is utterly ill-suited to this dramatic venue.
As we saw at Suzuka in Japan earlier this year, a 2026 F1 battery is able to power the race car's electric motor for only a few seconds a lap; for much of the rest of the lap, the energy-starved cars have less power going to their rear wheels than in the junior F2 feeder series.
For many, including much of the current F1 grid, Spa is considered perhaps the planet's greatest race track. In fact, both Spa and the Belgian Grand Prix predate the F1 world championship, with its first race held in 1922. Back then, the circuit was a 9-mile loop of public roads, taking in the towns of Malmedy and Stavelot before returning to Francorchamps. In the postwar years, the layout shrank slightly to 8.7 miles (14.1 km) and was ferociously fast, with average lap speeds of up to 150 mph (240 km/h)—and ferociously dangerous, given cars with huge amounts of power and minimal concessions to safety.
When Lucid introduced the Air electric sedan in late 2021, the first Air Dream Edition I tested packed over 1,100 hp (820 kW) and carried a $180,000-plus window sticker. It's easily the most powerful street car I've tested; the only vehicle I've driven with more power was a purebred race car with a third the mass, and it was on a proper track. Its combustion engine was also about 1,000 times louder than the Air, helping to remind us that "combustion" really does mean explosion after explosion.
For Lucid's second act, the company debuted the Gravity electric SUV last year, and I've just tested the 2026 Gravity Touring, which starts at about $82,000 in the US, including the required destination charge.
My test model carried a bevy of options, including a 22-speaker audio system, the Comfort and Convenience package, third-row seating, a Dynamic Handling package (combining rear-wheel steering and three-chamber air suspension), a luxury seating package (bundling Nappa leather and massaging and ventilated front seats), and special metallic paint.
SpaceX has pinned the bulk of its future value on orbital data centers. Not rockets. Not spacecraft.
Instead, it envisions launching and maintaining a constellation of 1 million satellites capable of generating 120 GW to power tens of millions—and potentially up to 100 million—frontier-class GPUs for data center services.
The company's founder, Elon Musk, revealed plans for this massive constellation months ago, but until recently, the scope of the individual satellites was largely unknown. That changed in June, when Musk and Ian Dahl, director of satellite engineering for SpaceX, spoke in a promotional video about the company's plans to develop the first iteration of an orbital data center, called an AI1 satellite. The video finally provided the company's numbers about the satellite's size and power capabilities.
Over the past few years, many of us have gotten a crash course in what we now call artificial intelligence—but really, it has mostly been a crash course in large language models. Increasingly, however, LLMs are no longer the only category of AI drawing high expectations, massive funding rounds, and significant research and product development.
Over the past year, we've seen a plethora of new announcements in a category labeled "world models," and you'll likely see more movement there in the coming months and years.
Instead of or in addition to working with language, world models aim to lay the groundwork for AI systems that are capable of simulating the physical world, or at least a useful approximation of it.
Anyone who has followed the aftermarket automotive performance industry for long enough can tell you just how dramatically it has changed over the past few decades. What once required mechanical tinkering and a lot of know-how can now be done in mere minutes via an electric control unit (ECU), which can extract significant boosts in horsepower and torque from naturally aspirated, turbocharged, or supercharged engines.
In some ways, though, the process has become much more difficult.
Just ask Alabama-based Audi Performance & Racing, more prominently known as APR. As modern vehicles become increasingly software-driven and OEMs continue to tighten security, the company has had to work harder each year to offer ECU tuning that delivers more power while staying within factory parameters for overall reliability. It's a far more arduous process now than it was in the early aughts, when my own B5-generation Audi S4 was still fresh on the market.
In the brief history of AI security, the prompt injection has quickly become the top threat. Large language models are inherently unable to distinguish between legitimate instructions provided by users and malicious ones sneaked into emails, source code, and other third-party content the models are processing. This makes it trivial to surreptitiously inject malicious commands that the LLM readily follows.
With no way to enforce this crucial boundary between trusted and untrusted sources, AI engine developers are left to erect elaborate guardrails designed to mitigate the damage rather than solve the root cause.
To date, most prompt injections have fallen into a class known as push, in which each potential victim is targeted. For example, the adversary injects malicious instructions into an individual email or calendar invitation. Because the injection must then be sent (or pushed) to each specific target, the scale of the attack is limited, hampering mass exploits that hit the Internet at large.
The temptation with a car like the Czinger 21C is to treat it as a collection of extreme specifications, and to be fair, it’s certainly not lacking in that department.
At its most basic level, the carbon-fiber-bodied 21C is a hybrid hypercar built around a bespoke 2.88-liter twin-turbocharged flat-plane crank V8 that revs to a searing 11,000 rpm. This power plant is matched up with a three-motor electric system—one electric motor drives each front wheel while a third serves as a crank-driven starter-generator. Combined output is rated at 1,250 hp (932 kW) and 691 lb-ft (937 Nm) of torque.
A seven-speed automated manual transaxle handles gear changes, chosen in part for its low mass and ability to tolerate high torque loads without the packaging penalties of a dual-clutch system. Tipping the scales at under 3,700 lbs (1,678 kg) with fluids, the 21C VMax is capable of hitting 60 mph (97 km/h) from rest in 1.92 seconds on its way to an 8.6 second quarter mile and a 253 mph (378 km/h) top speed, while the road course-focused 21C High Downforce model recently secured lap records at no less than five different California racetracks during a thousand-mile (1,600 km) road trip.
The last time America celebrated a big anniversary, I was all of three years old. Even so, I retain a few fuzzy memories from a sunny summer afternoon in small-town Michigan: climbing on a cannon in front of the courthouse, watching a parade, and seeing my dad, a veteran and Centreville city councilman, giving a short talk about democracy.
Only later would I realize the significance of the date: July 4th, 1976, America’s bicentennial.
America was imperfect and inconsistent in its approaches to "freedom," but the country had done some big, difficult things in recent decades. We had led the charge to roll back the tide of fascism and Holocaust during World War II. We had begun to confront internal demons through the nonviolent activism of the civil rights movement. And, critically for my own life trajectory, we had landed on the Moon.
Near the end of May, the Office of Management and Budget (OMB) proposed a new rule that would govern how the federal government handles the grants it issues, including those that fund the vast majority of scientific research in the US.
If formalized, the rule would make political priorities the prime determinant of what science gets funded and sideline the opinions of scientific experts. Grants could be canceled due to political whims, and new layers of bureaucracy would inhibit basic scientific activities like publishing papers and attending conferences. Unlike the executive orders it echoes, it would have the force of law behind it and be significantly harder to challenge in court.
Before coming into force, however, the proposal must go through a process that includes public feedback and (potentially) changes in response. The deadline for that feedback—Monday, July 13—is rapidly approaching.
President Donald Trump’s plan to “fill the digital potholes” and use AI to quickly redesign every government website isn’t going very well.
Last August, Trump created the National Design Studio, or NDS, by executive order. A temporary DOGE-like entity that answers only to the president, NDS was tasked with creating new standards to update the US Web Design System (USWDS) and overhaul 27,000 dot-gov websites in just three years. At the end of this so-called “America by Design” initiative, the government’s “design language” would supposedly be more usable and beautiful, Trump expected.
However, that monumental task—assigned to a small team under a short timeframe—was seemingly made harder by DOGE's deep cuts to agencies previously responsible for improving government websites, including dismantling the 18F technology unit and restructuring the US Digital Service into DOGE.
Lateo.net - Flux RSS en pagaille (pour en ajouter : @ moi)