Normal view

Received today — 13 August 2025Ars Technica

Space Force officials take secrecy to new heights ahead of key rocket launch

12 August 2025 at 20:10

After more than a decade of development and testing, US military officials are finally ready to entrust United Launch Alliance's Vulcan rocket to haul a batch of national security satellites into space.

An experimental military navigation satellite, also more than 10 years in the making, will ride ULA's Vulcan rocket into geosynchronous orbit more than 22,000 miles (nearly 36,000 kilometers) over the equator. There are additional payloads buttoned up inside the Vulcan rocket's nose cone, but officials from the US Space Force are mum on the details.

The Vulcan rocket is set for liftoff from Cape Canaveral Space Force Station, Florida, at 7:59 pm EDT (23:59 UTC) Tuesday. There's an 80 percent chance of favorable weather during the one-hour launch window. It will take several hours for the Vulcan rocket's Centaur upper stage to reach its destination in geosynchronous orbit. You can watch ULA's live launch webcast below.

Read full article

Comments

© United Launch Alliance

Why it’s a mistake to ask chatbots about their mistakes

12 August 2025 at 19:52

When something goes wrong with an AI assistant, our instinct is to ask it directly: "What happened?" or "Why did you do that?" It's a natural impulse—after all, if a human makes a mistake, we ask them to explain. But with AI models, this approach rarely works, and the urge to ask reveals a fundamental misunderstanding of what these systems are and how they operate.

A recent incident with Replit's AI coding assistant perfectly illustrates this problem. When the AI tool deleted a production database, user Jason Lemkin asked it about rollback capabilities. The AI model confidently claimed rollbacks were "impossible in this case" and that it had "destroyed all database versions." This turned out to be completely wrong—the rollback feature worked fine when Lemkin tried it himself.

And after xAI recently reversed a temporary suspension of the Grok chatbot, users asked it directly for explanations. It offered multiple conflicting reasons for its absence, some of which were controversial enough that NBC reporters wrote about Grok as if it were a person with a consistent point of view, titling an article, "xAI's Grok offers political explanations for why it was pulled offline."

Read full article

Comments

© Alan Schein via Getty Images

Perplexity offers more than twice its total valuation to buy Chrome from Google

12 August 2025 at 18:49

In the wake of its big antitrust loss, Google could soon find itself forced to sell one of its crown jewels. Among the government's proposed remedies in the search case is a requirement that Google divest its market-leading Chrome browser, and Perplexity is already throwing its proverbial hat into the ring with a whopping $34.5 billion offer. The problem, however, is that Perplexity doesn't have nearly that much cash.

Perplexity has ridden the AI hype wave, with its AI-powered search appearing on smartphones and in the company's custom Comet browser. Like any company offering an AI product, investors have been happy to throw money at Perplexity, totaling around $1 billion so far. Investors value the company at about $14 billion right now. So how does Perplexity have more than twice that to buy Chrome? That's the neat part—it doesn't.

There is so much capital floating around in the artificial intelligence sphere currently that even a cash-poor firm like Perplexity can secure enough investment to splurge on Chrome. Reuters reports that the all-cash offer is funded by various venture funds, but Perplexity has not offered specifics.

Read full article

Comments

© Getty Images

YouTube backlash begins: “Why is AI combing through every single video I watch?”

12 August 2025 at 17:58

Tens of thousands of YouTubers are raging against YouTube's plan to use AI to detect underage users in the US.

On Tuesday, a Change.org petition rapidly neared its 50,000-signature goal, with tens of thousands hoping that with enough users protesting, the wide rollout of the AI age checks might be stopped. They fear the age checks will make it harder to access content they love while staying anonymous on the platform

YouTube's age verification system estimates user ages by interpreting a "variety of signals," YouTube's announcement said, including "the types of videos a user is searching for, the categories of videos they have watched, or the longevity of the account."

Read full article

Comments

© NurPhoto / Contributor | NurPhoto

They’re golden: Fictional band from K-Pop Demon Hunters tops the charts

12 August 2025 at 16:25
The fictional band Huntr/x, from K-Pop Demon Hunters, has a real-world hit with "Golden."

Netflix has a summer megahit on its hands with its animated musical feature film, K-Pop Demon Hunters. Since its June release, the critically acclaimed film has won fans of all ages, fueled by a killer Korean pop soundtrack featuring one earworm after another. The biggest hit is "Golden," which just hit No. 1 on Billboard's Top 100 chart. (The last time a fictional ensemble topped the charts was in 2022 with Encanto's "We Don't Talk About Bruno.")

K-Pop Demon Hunters is now Netflix's most-watched animated film of all time, and that's not just because of the infectious music. The Sony Animation team delivers bold visuals that evoke the look and feel of anime, the plot is briskly paced, and the script strikes a fine balance between humor and heart.

(Spoilers below.)

Read full article

Comments

© Netflix

Rad Power’s Radster: A very non-radical commuter bike

12 August 2025 at 15:36

With e-bike manufacturing in China having expanded considerably, the number of companies offering affordable e-bikes over the last five years has exploded. But the market for cycles with an electric assist has existed for considerably longer, and a number of companies predate the recent surge. One of them, Rad Power, has been around long enough that it was already an established presence when we first reviewed its hardware four years ago.

The company offers a mix of cargo, folding, and commuter bikes, all with electric assists. Having looked at a cargo version last time around, we decided to try out one of the commuter bikes this time. The Radster comes in road and trail versions (we tried the road). It's an incredibly solidly made bike with equally solid components, and it has very good implementations of a few things that other manufacturers haven't handled all that well. It also can switch among the three classes of e-bikes using a menu option; unfortunately, nothing else about the bike's performance seems to change with the switch.

The Radster is priced a bit higher than a lot of its budget competitors. So, if you're shopping, you'll have to think a bit about whether some of these features matter to you.

Read full article

Comments

© JOHN TIMMER

Musk threatens to sue Apple so Grok can get top App Store ranking

12 August 2025 at 15:27

After spending last week hyping Grok's spicy new features, Elon Musk kicked off this week by threatening to sue Apple for supposedly gaming the App Store rankings to favor ChatGPT over Grok.

"Apple is behaving in a manner that makes it impossible for any AI company besides OpenAI to reach #1 in the App Store, which is an unequivocal antitrust violation," Musk wrote on X, without providing any evidence. "xAI will take immediate legal action."

In another post, Musk tagged Apple, asking, "Why do you refuse to put either X or Grok in your 'Must Have' section when X is the #1 news app in the world and Grok is #5 among all apps?"

Read full article

Comments

© Kevin Dietsch / Staff | Getty Images News

China tells Alibaba, ByteDance to justify purchases of Nvidia AI chips

Beijing is demanding tech companies including Alibaba and ByteDance justify their orders of Nvidia’s H20 artificial intelligence chips, complicating the US chipmaker’s business in China after striking an export arrangement with the Trump administration.

The tech companies have been asked by regulators such as the Ministry of Industry and Information Technology (MIIT) to explain why they need to order Nvidia’s H20 chips instead of using domestic alternatives, said three people familiar with the situation.

Some tech companies, who were the main buyers of Nvidia’s H20 chips before their sale in China was restricted, were planning to downsize their orders as a result of the questions from regulators, said two of the people.

Read full article

Comments

© Benj Edwards / Nvidia

High-severity WinRAR 0-day exploited for weeks by 2 groups

12 August 2025 at 00:13

A high-severity zero-day in the widely used WinRAR file compressor is under active exploitation by two Russian cybercrime groups. The attacks backdoor computers that open malicious archives attached to phishing messages, some of which are personalized.

Security firm ESET said Monday that it first detected the attacks on July 18, when its telemetry spotted a file in an unusual directory path. By July 24, ESET determined that the behavior was linked to the exploitation of an unknown vulnerability in WinRAR, a utility for compressing files, and has an installed base of about 500 million. ESET notified WinRAR developers the same day, and a fix was released six days later.

Serious effort and resources

The vulnerability seemed to have super Windows powers. It abused alternate data streams, a Windows feature that allows different ways of representing the same file path. The exploit abused that feature to trigger a previously unknown path traversal flaw that caused WinRAR to plant malicious executables in attacker-chosen file paths %TEMP% and %LOCALAPPDATA%, which Windows normally makes off-limits because of their ability to execute code.

Read full article

Comments

© Getty Images

The GPT-5 rollout has been a big mess

11 August 2025 at 22:25

It's been less than a week since the launch of OpenAI's new GPT-5 AI model, and the rollout hasn't been a smooth one. So far, the release sparked one of the most intense user revolts in ChatGPT's history, forcing CEO Sam Altman to make an unusual public apology and reverse key decisions.

At the heart of the controversy has been OpenAI's decision to automatically remove access to all previous AI models in ChatGPT (approximately nine, depending on how you count them) when GPT-5 rolled out to user accounts. Unlike API users who receive advance notice of model deprecations, consumer ChatGPT users had no warning that their preferred models would disappear overnight, noted independent AI researcher Simon Willison in a blog post.

The problems started immediately after GPT-5's August 7 debut. A Reddit thread titled "GPT-5 is horrible" quickly amassed over 4,000 comments filled with users expressing frustration over the new release. By August 8, social media platforms were flooded with complaints about performance issues, personality changes, and the forced removal of older models.

Read full article

Comments

© Benj Edwards / OpenAI

Boar’s Head to reopen plant as mold and funky meat problems pop up elsewhere

11 August 2025 at 22:02

Boar's Head plans to reopen the Jarratt, Virginia, facility at the center of a deadly Listeria outbreak last year despite federal inspections continuing to find sanitation violations at three of the food company's other facilities, according to federal records obtained by The Associated Press.

The AP obtained 35 pages of inspection reports via a Freedom of Information Act Request. Those reports cover inspections between January 1 and July 23 at three other Boar's Head facilities: Forrest City, Arkansas; New Castle, Indiana; and Petersburg, Virginia. Overall, the reports reveal a suite of violations, including mold, condensation dripping over food areas, overflowing trash, meat and fat residue built up on walls and equipment, drains blocked with meat scraps, and pooling meat juice. The reports also recorded staff who didn't wear the proper protective hairnets and aprons—and didn't wash their hands.

In one violation, reported in the Petersburg facility, inspectors found meat waste collecting under equipment, including "5-6 hams, 4 large pieces of meat and a large quantity of pooling meat juice."

Read full article

Comments

© Getty | Justin Sullivan

AOL announces September shutdown for dial-up Internet access

11 August 2025 at 20:17

After decades of connecting Americans to its online service and the Internet through telephone lines, AOL recently announced it is finally shutting down its dial-up modem service on September 30, 2025. The announcement marks the end of a technology that served as the primary gateway to the World Wide Web for millions of users throughout the 1990s and early 2000s.

AOL confirmed the shutdown date in a help message to customers: "AOL routinely evaluates its products and services and has decided to discontinue Dial-up Internet. This service will no longer be available in AOL plans."

Along with the dial-up service, AOL announced it will retire its AOL Dialer software and AOL Shield browser on the same date. The dialer software managed the connection process between computers and AOL's network, while Shield was a web browser optimized for slower connections and older operating systems.

Read full article

Comments

Reddit blocks Internet Archive to end sneaky AI scraping

11 August 2025 at 19:53

Reddit is now blocking the Internet Archive (IA) from indexing popular Reddit threads after allegedly catching sneaky AI firms—restricted from scraping Reddit—instead simply scraping data from IA's archived content.

Where before IA's Wayback Machine dependably archived Reddit pages, profiles, and comments—as part of its mission to archive the Internet—moving forward, only screenshots of the Reddit homepage will be archived. As The Verge noted, this means the archive will only be useful as a snapshot of popular posts and news headlines each day, rather than providing a backup documenting deleted posts or a window into various Reddit subcultures or any given user's activity.

Reddit has not confirmed which AI firms were scraping its data from the Wayback Machine. The company's spokesperson, Tim Rathschmidt, would only confirm to Ars that Reddit has become "aware of instances where AI companies violate platform policies, including ours, and scrape data from the Wayback Machine."

Read full article

Comments

© Cheng Xin / Contributor | Getty Images News

GitHub will be folded into Microsoft proper as CEO steps down

11 August 2025 at 19:06

Microsoft has owned GitHub since 2018, but the widely used developer platform has operated with at least a little independence from the rest of the company, with its own separate CEO and other executives. But it looks like GitHub will be more fully folded into Microsoft's organizational chart starting next year—GitHub CEO Thomas Dohmke announced today that he would be leaving GitHub and Microsoft "to become a founder again."

"GitHub and its leadership team will continue its mission as part of Microsoft’s CoreAI organization, with more details shared soon," Dohmke wrote. "I’ll be staying through the end of 2025 to help guide the transition and am leaving with a deep sense of pride in everything we’ve built as a remote-first organization spread around the world."

Axios reports that Microsoft isn't directly replacing Dohmke, and GitHub's leadership team will be reporting to multiple executives in the CoreAI division.

Read full article

Comments

© GitHub

Wikipedia loses UK Safety Act challenge, worries it will have to verify user IDs

11 August 2025 at 17:48

Wikipedia's parent organization lost a challenge to the UK Online Safety Act but can bring another case if the government tries to force it to verify the identity of Wikipedia users.

The High Court of Justice in London dismissed claims from the Wikimedia Foundation, which challenged the lawfulness of the categorization system used to determine which sites must comply with obligations. But Justice Jeremy Johnson stressed "that this does not give Ofcom and the Secretary of State a green light to implement a regime that would significantly impede Wikipedia's operations."

The Online Safety Act has forced social media sites like Reddit to verify UK users' ages before letting them view adult content. The Wikimedia Foundation is worried that it will be classified as a "Category 1" operator later this summer and criticized the categorization regulations as "especially broad and vague."

Read full article

Comments

© Getty Images | Riccardo Milani

Why does Jeff Bezos keep buying launches from Elon Musk?

11 August 2025 at 17:30

Early Monday morning, a Falcon 9 rocket lifted off from its original launch site in Florida. Remarkably, it was SpaceX's 100th launch of the year.

Perhaps even more notable was the rocket's payload: two-dozen Project Kuiper satellites, which were dispensed into low-Earth orbit on target. This was SpaceX's second launch of satellites for Amazon, which is developing a constellation to deliver low-latency broadband Internet around the world. SpaceX, then, just launched a direct competitor to its Starlink network into orbit. And it was for the founder of Amazon, Jeff Bezos, who owns a rocket company of his own in Blue Origin.

So how did it come to this—Bezos and Elon Musk, competitors in so many ways, working together in space?

Read full article

Comments

© SpaceX

LLMs’ “simulated reasoning” abilities are a “brittle mirage,” researchers find

11 August 2025 at 17:16

In recent months, the AI industry has started moving toward so-called simulated reasoning models that use a "chain of thought" process to work through tricky problems in multiple logical steps. At the same time, recent research has cast doubt on whether those models have even a basic understanding of general logical concepts or an accurate grasp of their own "thought process." Similar research shows that these "reasoning" models can often produce incoherent, logically unsound answers when questions include irrelevant clauses or deviate even slightly from common templates found in their training data.

In a recent pre-print paper, researchers from the Arizona State University summarize this existing work as "suggest[ing] that LLMs are not principled reasoners but rather sophisticated simulators of reasoning-like text." To pull on that thread, the researchers created a carefully controlled LLM environment in an attempt to measure just how well chain-of-thought reasoning works when presented with "out of domain" logical problems that don't match the specific logical patterns found in their training data.

The results suggest that the seemingly large performance leaps made by chain-of-thought models are "largely a brittle mirage" that "become[s] fragile and prone to failure even under moderate distribution shifts," the researchers write. "Rather than demonstrating a true understanding of text, CoT reasoning under task transformations appears to reflect a replication of patterns learned during training."

Read full article

Comments

© Getty Images

RFK Jr. posted fishing pics as CDC reeled from shooting linked to vaccine disinfo

11 August 2025 at 17:06

Staff at the Centers for Disease Control and Prevention headquarters in Atlanta are reeling from a deadly shooting that unfolded Friday evening.

The shooting left one local police officer dead, at least four agency buildings riddled with bullet holes, and terrified staffers feeling like "sitting ducks." Fortunately, no CDC staff or civilians were injured. But, it quickly drew a spotlight to US health secretary and zealous anti-vaccine advocate Robert F. Kennedy Jr., who critics accused of fueling the violence with his menacing and reckless anti-vaccine rhetoric.

Kennedy publicly responded to the shooting on social media at about 11 am Eastern Time on Saturday, roughly 18 hours after the event. Former US Surgeon General Jerome Adams subsequently slammed Kennedy's delayed response as "tepid" in a critical essay published in Stat. The news outlet separately pointed out that Kennedy had posted on his personal social media account about 30 minutes prior to his response to the shooting, in which he shared pictures of a fishing trip.

Read full article

Comments

© Getty | Elijah Nouvelage

Trump strikes “wild” deal making US firms pay 15% tax on China chip sales

11 August 2025 at 16:31

Ahead of an August 12 deadline for a US-China trade deal, Donald Trump's tactics continue to confuse those trying to assess the country's national security priorities regarding its biggest geopolitical rival.

For months, Trump has kicked the can down the road regarding a TikTok ban, allowing the app to continue operating despite supposedly urgent national security concerns that China may be using the app to spy on Americans. And now, in the latest baffling move, a US official announced Monday that Trump got Nvidia and AMD to agree to "give the US government 15 percent of revenue from sales to China of advanced computer chips," Reuters reported. Those chips, about 20 policymakers and national security experts recently warned Trump, could be used to fuel China's frontier AI, which seemingly poses an even greater national security risk.

Trump’s “wild” deal with US chip firms

Reuters granted two officials anonymity to discuss Trump's deal with US chipmakers, because details have yet to be made public. Requiring US firms to pay for sales in China is an "unusual" move for a president, Reuters noted, and the Trump administration has yet to say what exactly it plans to do with the money.

Read full article

Comments

© SOPA Images / Contributor | LightRocket

Scientists hid secret codes in light to combat video fakes

11 August 2025 at 15:45

It's easier than ever to manipulate video footage to deceive the viewer and increasingly difficult for fact checkers to detect such manipulations. Cornell University scientists developed a new weapon in this ongoing arms race: software that codes a "watermark" into light fluctuations, which in turn can reveal when the footage has been tampered with. The researchers presented the breakthrough over the weekend at SIGGRAPH 2025 in Vancouver, British Columbia, and published a scientific paper in June in the journal ACM Transactions on Graphics.

“Video used to be treated as a source of truth, but that’s no longer an assumption we can make,” said co-author Abe Davis, of Cornell University, who first conceived of the idea. “Now you can pretty much create video of whatever you want. That can be fun, but also problematic, because it’s only getting harder to tell what’s real.”

Per the authors, those seeking to deceive with video fakes have a fundamental advantage: equal access to authentic video footage, as well as the ready availability of advanced low-cost editing tools that can learn quickly from massive amounts of data, rendering the fakes nearly indistinguishable from authentic video. Thus far, progress on that front has outpaced the development of new forensic techniques designed to combat the problem. One key feature is information asymmetry: An effective forensic technique must have information not available to the fakers that cannot be learned from publicly available training data.

Read full article

Comments

© Sreang Hok/Cornell University

❌