news aggregator

An anonymous reader quotes a report from the San Francisco Chronicle: Elon Musk returned to the witness stand Wednesday in Oakland federal court for a second day of testimony in his case against OpenAI, detailing his shift from being an enthusiastic supporter of the nonprofit to feeling betrayed. He also clashed repeatedly with OpenAI's attorney over questions that Musk believed were unfair. He said his feelings towards OpenAI CEO Sam Altman and President Greg Brockman shifted from a "phase one" of support, "phase two" of doubts, and finally "phase three, where I'm sure they're looting the nonprofit. We're currently in phase three," Musk said with a chuckle. Musk said he was a "fool" for giving OpenAI "$38 million of essentially free funding to create what would become an $800 billion company," of which he has no equity stake. In his 2024 lawsuit, Musk alleged breach of charitable trust and unjust enrichment, arguing OpenAI abandoned its original nonprofit mission to benefit humanity to pursue financial gain. OpenAI's lawyer William Savitt argued Tuesday during his opening statement that the nonprofit entity remains in control of the for-profit public benefit corporation and is now one of the most well-funded nonprofits in the world. Musk is seeking to oust Altman from OpenAI's board and upwards of $134 billion in damages, which he said would be used to fund OpenAI's nonprofit mission. During cross-examination, Savitt clashed with Musk over questioning. Savitt asked whether Musk had contributed $38 million to OpenAI, rather than the $100 million that he later claimed to have invested on X. Musk said he also contributed his reputation to the company and came up with the idea for the name, leading Savitt to ask Musk to respond yes or no to "simple" questions. "Your questions are not simple. They're designed to trick me, essentially," Musk said, adding that he had to elaborate or it would mislead the jury. He compared Savitt's questions to asking, "have you stopped beating your wife?" Judge Yvonne Gonzalez Rogers intervened, leading Musk to answer yes to the $38 million investment amount. The world's richest man said his doubts grew and by late 2022, he thought "wait a second, these guys are betraying their promise. They're breaking the deal." "I started to lose confidence that they were telling me the truth," Musk said. A turning point was co-defendent Microsoft's investment of billions of dollars into OpenAI, Musk said. On October 23, 2022, Musk texted Altman that he was "disturbed" to see OpenAI's valuation of $20 billion in the wake of the Microsoft deal. Musk called the deal a "bait and switch," since a nonprofit doesn't have a valuation. OpenAI had "for all intents and purposes" become primarily a for-profit company, Musk argued. Altman responded to Musk by text that "I agree this feels bad," saying that OpenAI had previously offered equity in the company but Musk hadn't wanted it at the time. Altman said the company was happy to offer equity in the future. Musk said it "didn't seem to make sense to me" to hold equity in what should be a nonprofit. Musk also testified about former OpenAI board member Shivon Zilis, who lives with him, is the mother of four of his children, and served as a senior advisor at Neuralink. He denied that she shared sensitive OpenAI information with him. Court evidence showed Musk had encouraged her to stay close to OpenAI to "keep info flowing" and had approved Neuralink recruiting OpenAI employees, which he defended by saying workers are free to change jobs. "It's a free country," Musk said. Recap: Musk Testifies OpenAI Was Created As Nonprofit To Counter Google (Day Two) Elon Musk and OpenAI CEO Sam Altman Head To Court (Day One)

Read more of this story at Slashdot.

Will write checks for $190 billion and even those megabucks may not satisfy demand

If you've felt the sting of surging hardware prices, Microsoft can sympathize because the company on Wednesday said it expects its 2026 capital expenditure will hit $190 billion, with $25 billion of that due to rising component costs.…

Patches land for authencesn flaw enabling local privilege escalation

Developers of major Linux distributions have begun shipping patches to address a local privilege escalation (LPE) vulnerability arising from a logic flaw.…

The Trainium train keeps a-rollin'

Amazon is now among the top three datacenter chip businesses in the world, as its semiconductor business surpassed a $20 billion annual run rate ... and it would be closer to $50 billion if it included itself among the customers, CEO Andy Jassy said during the company’s first quarter earnings call on Wednesday.…

A federal judge denied Sam Bankman-Fried's request for a new trial, calling his claims of DOJ witness intimidation "wildly conspiratorial" and unsupported by the record. Judge Lewis Kaplan said (PDF) the FTX founder's motion appeared tied to a pre-indictment plan to recast himself as a Republican victim of Biden's DOJ in hopes of gaining sympathy, leniency, or even a Trump pardon. Ars Technica reports: Bankman-Fried was sentenced to 25 years in prison in 2024 for "masterminding one of the largest financial frauds in American history," US District Judge Lewis Kaplan wrote in his order. He was convicted on all charges, including wire fraud, conspiracy to commit securities fraud, commodities fraud, and money laundering. There is already an appeal pending in another court, the judge noted. But Bankman-Fried filed a separate motion for a new trial, claiming that there were "newly discovered" witnesses and evidence that might have helped his defense, if Joe Biden's Department of Justice hadn't intimidated them into refusing to testify or, in one case, lying on the stand. He also asked for a new judge, wanting Kaplan to recuse himself. However, Kaplan pointed out that "none of the witnesses" were "newly discovered." And more concerningly, Bankman-Fried offered no evidence that the witnesses could prove the "wildly conspiratorial" theory the FTX founder raised, claiming that their absence at the trial was a "product of government threats and retaliation," the judge wrote. Bankman-Fried's theory is "entirely contradicted by the record," Kaplan said. He emphasized that granting Bankman-Fried's request "would be a large waste of judicial resources as it could require another judge to familiarize himself or herself with an extensive and complicated record." Additionally, all three witnesses that Bankman-Fried claimed could give crucial testimony in his defense were known to him throughout the trial, and he never sought to compel their testimony. And the "self-serving social-media posts" of one witness who now claims that he lied when testifying against Bankman-Fried -- "Ryan Salame, who pleaded guilty" -- must be met with "utmost suspicion," Kaplan said. "If one were to take Salame at his current word, he lied under oath when pleading guilty before this Court," Kaplan wrote. Even if taken seriously, "his out-of-court, unsworn statements could not come anywhere close to clearing the bar to warrant a new trial," Kaplan said, deeming Salame's credibility "highly questionable." Further, "even if these individuals had testified for Bankman-Fried, his protestations that one or more of them would have supported his claims that FTX was not insolvent and that his victims all were compensated fully in the bankruptcy proceedings are inaccurate or misleading," Kaplan concluded. In the order, Kaplan's frustration seems palpable, as there may have been no need for him to rule on the motion at all after Bankman-Fried requested to withdraw it. But the judge said the ruling was needed after Bankman-Fried waited to file his withdrawal request until after the DOJ and the court wasted time responding and reviewing filings, the judge said. Troublingly, Bankman-Fried's request to withdraw his request without prejudice would have allowed him to potentially request a new trial after the appeal ended. Based on the substance of the filing, that risked wasting future court resources, Kaplan determined. To prevent overburdening the justice system, Kaplan deemed it necessary to deny Bankman-Fried's motion and request for recusal, rather than allow him to withdraw the filing without prejudice.

Read more of this story at Slashdot.

Canonical's plan to add AI features to Ubuntu has sparked pushback from users who are concerned it could follow Windows 11's AI-heavy direction. "After Canonical's announcement earlier this week that it's bringing AI features to Ubuntu, replies included requests for an AI 'kill switch' or a way to disable the upcoming features," reports The Verge. Canonical says it has no plans for a "global AI kill switch" but it will allow users to remove any AI features they don't want. From the report: In his original post, [Canonical's VP of engineering, Jon Seager] said the upcoming AI features will include accessibility tools like AI speech-to-text and text-to-speech, along with agentic AI features for tasks like troubleshooting and automation. Canonical is also encouraging its engineers to use AI more and plans to begin introducing AI features in Ubuntu "throughout the next year." In a follow-up comment, Seager clarified that, "my plan is to introduce AI-backed features as a 'preview' on a strictly opt-in basis in [Ubuntu version] 26.10. In subsequent releases, my plan is to have a step in the initial setup wizard that allows the user to choose whether or not they'd like the AI-native features enabled." Ultimately, he said, "All of these capabilities will be delivered as Snaps to the OS, layered on top of the existing Ubuntu stack. That means there will always be the option of removing those Snaps." Users who prefer to avoid AI entirely could switch to other distros like Linux Mint, Pop!_OS, or Zorin OS. "These distros have some similarities to Ubuntu, but may not necessarily adopt the new AI features Canonical is rolling out," adds The Verge.

Read more of this story at Slashdot.

Joby Aviation has completed demonstration flights of its electric air taxi over New York City, testing real routes between JFK and Manhattan helipads as it prepares for a future commercial service. The company says its eVTOL could turn a 60- to 120-minute airport trip into a flight of under 10 minutes, though commercial launch still depends on FAA certification. Electrive reports: To launch operations in New York City, Joby acquired Blade Urban Air Mobility last year. Blade already enables helicopter flights for affluent travelers between Manhattan and airports such as JFK or Newark in just five minutes, avoiding up to two hours of traffic and typical airport hassles. Joby aims to replace this service with quiet, electric air taxis as soon as possible, transitioning Blade's existing customers to the new technology. However, introducing a new aircraft into commercial service requires a years-long certification process, overseen in the US by the Federal Aviation Administration (FAA). Joby is now in the final phase of FAA certification. Following a series of demonstration flights in the San Francisco Bay Area, the company has tested its air taxi in New York City on real flight routes and under real-world conditions. During these tests, Joby demonstrated the acoustics and performance metrics critical for entering the urban air taxi market. During these demonstration flights, Joby's air taxi took off from John F. Kennedy International Airport (JFK) and landed at various helipads across the city, including Downtown Skyport and the helipads at West 30th Street and East 34th Street in Midtown, where Blade Air Mobility's premium passenger lounges are located. These locations represent some of the commercial routes Joby plans for New York [...]. Fun fact: Joby's eVTOL aircraft are over 100 to 1,000 times quieter than a conventional helicopter, operating at roughly 55-65 dB during takeoff and landing compared to 90+ dB for helicopters.

Read more of this story at Slashdot.

ORNL says portable detector kit can separate real GPS signals from fake ones even at equal strength

GPS spoofing, which sends fake satellite-like signals, and GPS jamming, which drowns receivers in noise, are increasingly serious problems. Researchers at Oak Ridge National Laboratory in Tennessee have created what they say is the most effective system yet for detecting GPS interference, which could help blunt such attacks.…

MacRumors reports that Apple has effectively paused work on Vision Pro after the M5 refresh failed to revive demand. The team has reportedly been reassigned and the company is now shifting focus toward smart glasses instead. From the report: The Vision Pro has been criticized for its high price tag and its uncomfortable weight. The device is over 1.3 pounds, and even with the more comfortable Dual Knit Band that Apple added to redistribute weight, it continues to be hard to wear for long periods of time. The M5 chip added a 120Hz refresh rate, 10 percent more rendered pixels, and around 30 additional minutes of battery life, but the price tag stayed at $3,499, and it ended up not selling well. The Vision Pro has been unpopular since it first launched, and Apple only sold around 600,000 units in total. Insider sources told MacRumors that Apple has received an unusually high percentage of returns, far exceeding any other modern Apple product. [...] If Apple finds a way to create a much cheaper, more comfortable VR headset in the future, the Vision Pro line could be revived, but right now, the company has no plans to launch a new model. Apple has not discontinued the Vision Pro and is continuing to sell the M5 model. Instead of continuing to experiment with virtual reality, Apple is working on smart glasses that will eventually incorporate augmented reality capabilities, but the first version will be similar to the Ray-Ban Meta smart glasses with AI and no integrated display.

Read more of this story at Slashdot.

Second try's a charm?

Microsoft and the US Cybersecurity and Infrastructure Security Agency (CISA) warned that attackers are exploiting a zero-click Windows flaw that can expose sensitive information on vulnerable systems.…

Longtime Slashdot reader schwit1 writes: California's long-delayed high-speed rail project is now facing renewed scrutiny after state leaders revealed a dramatically higher price tag, now estimated at roughly $231 billion, nearly seven times the original $33 billion projection approved by voters in 2008. The revised figures have reignited talks in Sacramento over whether the project can realistically be completed, how long it will take, and whether the state can continue to fund it at this scale. Senator Strickland pointed to comments from Lou Thompson, former chair of the California High-Speed Rail Authority peer review group, who recently criticized the latest draft business plan. Thompson wrote that the 2026 draft plan "has reached a dead end," arguing that the project has drifted far from its original vision due to escalating costs, delays, and unfunded gaps. Under current projections, assuming funding and construction proceed as planned, service between San Francisco and Bakersfield could begin around 2033, while the full Los Angeles to San Francisco connection could extend to 2040.

Read more of this story at Slashdot.

Microsoft readies the axe once again for yesterday's security

Microsoft has warned users still clinging to legacy TLS versions that the end is nigh for TLS 1.0 and 1.1 on POP3 and IMAP4 connections to Exchange Online.…

Authors say it acquired an LLM that was trained on their copyrighted data, and judge keeps asking for more info

Databricks cannot shake a class action lawsuit targeting its LLM, which several book authors contend was created with a database that contained pirated versions of some of their copyrighted books – and about 196,000 titles in all.…

An anonymous reader quotes a report from Wired: A controversial bill in Colorado that would have undone some repair protections in the state has failed. The bill had been the target of right-to-repair advocates, who saw it as a bellwether for how tech companies might try to undo repair legislation more broadly in the US. Colorado's landmark 2024 repair law, the Consumer Right to Repair Digital Electronic Equipment, went into effect in January 2026 and ensured access to tools and documentation people needed to modify and fix digital electronics such as phones, computers, and Wi-Fi routers. The new bill, SB26-090, would have carved out an exception to those repair protections for "critical infrastructure," a loosely defined term that repair advocates worried could be applied to just about any technology. SB26-090 was introduced during a Colorado Senate hearing on April 2 and was supported by lobbying efforts from companies such as Cisco and IBM. It passed that hearing unanimously. The bill then passed in the Colorado Senate on April 16. On Monday evening, the bill was discussed in a long, delayed hearing in the Colorado House's State, Civic, Military, and Veterans Affairs Committee. Dozens of supporters and detractors gave public comments. Finally, the bill was shot down in a 7-to-4 vote and classified as postponed indefinitely. "While we were making progress at chipping away at the momentum for it, we had still been losing," said Danny Katz, executive director of the local nonprofit consumer advocacy group CoPIRG. "So, we took nothing for granted, and I believe the incredible testimony from the broad range of cybersecurity experts, businesses, repair advocates, recyclers, and people who want the freedom to fix their stuff made a big difference."

Read more of this story at Slashdot.

New sealed bootable container images and Stratis storage, too

Fedora Linux 44 has arrived – in multiple formats and for several CPU families, including some new container formats and storage options.…

Iran went dark twice, AWS got droned, oh and TalkTalk broke something it refuses to talk about

The first quarter of 2026 saw a surge in severe and prolonged internet disruptions, from government shutdowns to power outages to the occasional mystery incident.…

There is no 6 Nimmt! champion, but a $12 domain registration and one Wikipedia edit convinced several bots there was

Unlike search engines that let you judge competing sources, search-backed AI chatbots can turn shaky web material into confident answers. Case in point: A security engineer convinced several bots that he was the reigning world champion of a popular German card game, even though no such championship exists.…

Hashicorp co-founder Mitchell Hashimoto says GitHub's frequent outages have made it "no longer a place for serious work," prompting him to move his Ghostty terminal emulator project elsewhere after 18 years on the platform. The Register reports: "I've been angry about it. I've hurt people's feelings. I've been lashing out. Because GitHub is failing me, every single day, and it is personal. It is irrationally personal," he wrote. The reason for his ire is the service has become unreliable. "For the past month I've kept a journal where I put an 'X' next to every date where a GitHub outage has negatively impacted my ability to work," he wrote. "Almost every day has an 'X'. On the day I am writing this post, I've been unable to do any PR review for ~2 hours because there is a GitHub Actions outage." Hashimoto penned his post a few days before an April 28 incident that saw pull requests fail to complete due to an Elasticsearch SNAFU. Incidents like that mean Hashimoto has decided GitHub "is no longer a place for serious work if it just blocks you out for hours per day, every day." "It's not a fun place for me to be anymore," he lamented. "I want to be there but it doesn't want me to be there. I want to get work done and it doesn't want me to get work done. I want to ship software and it doesn't want me to ship software." The developer says he wants GitHub to improve, but "I also want to code. And I can't code with GitHub anymore. I'm sorry. After 18 years, I've got to go." He's open to a return if GitHub can deliver "real results and improvements, not words and promises." But for now, he's working to move Ghostty to another collaborative code locker. "We have a plan but I'm also very much still in discussions with multiple providers (both commercial and FOSS)," Hashimoto wrote. "It'll take us time to remove all of our dependencies on GitHub and we have a plan in place to do it as incrementally as possible." He's doing the equivalent of leaving a toothbrush at a former partner's house by leaving a read-only mirror of Ghostty on GitHub, and by keeping his personal projects on the Microsoft-owned service. But Hashimoto's moving his day job somewhere new. "Ghostty is where I, our maintainers, and our open source community are most impacted so that is the focus of this change. We'll see where it goes after that," he concluded.

Read more of this story at Slashdot.

Despite looming science cuts, Isaacman finds resources to poke the planetary hornet nest

NASA Administrator Jared Isaacman delivered some potentially good news at a Senate hearing this week, as well as some slightly odd news: in an environment of constrained budgets, the space agency was somehow finding resources to contest the decision to relegate Pluto from planet status.…

Tony Isaac shares a report from NPR: Federal survey data shows that the amount of math homework assigned to fourth and eighth grade students, in particular, has been steadily declining for the past decade. Some educators and parents say this is a good thing -- students shouldn't spend six or more hours a day at school and still have additional schoolwork to complete at home. But the research on homework is complicated. Some studies show that students who spend more time on homework perform better than their peers. For example, a longitudinal study released in 2021 of more than 6,000 students in Germany, Uruguay and the Netherlands found that lower-performing students who increased the amount of time they spent on math homework performed better in math, even one year later. Other studies, however, suggest homework has minimal outcomes on academic performance: A 1998 study of more than 700 U.S. students led by a researcher at Duke University found that more homework assigned in elementary grades had no significant effect on standardized test scores. The researchers did find small positive gains on class grades when they looked at both test scores and the proportion of homework students completed. More homework was also associated with negative attitudes about school for younger children in the study. "The best educators figured out a long time ago that we can control what we can control," and that's what happens during the school day, Superintendent Garrett said, not homework. "There has been a shift away from it naturally anyway, and I felt like this made it equitable across our entire school system." "The best argument for homework is that mathematical procedures require practice, and you don't want to waste classroom time on practice, so you send that home," said Tom Loveless, a researcher and former teacher who has studied homework. Ariel Taylor Smith, senior director of the Center for Policy and Action at the National Parents Union, said: "The thing they point to is that it's an equity issue, and not all parents have the same availability and ability to support their students. I would make the argument that if a kid is really far behind in school, that's an equity issue. They need the additional time to practice." Kids, she said, "need more practice ... Sometimes, you do have to practice the boring stuff, like math." "The interesting issue for folks to consider is not should there be more homework, but should there be better homework," said Joyce Epstein, who has studied homework and is the co-director of the Center on School, Family, and Community Partnerships at the Johns Hopkins University School of Education. "Better homework in math might be knowing the fact that kids don't have to be practicing for hours, 10 to 20 examples," when they could establish mastery in less time.

Read more of this story at Slashdot.