Apple will begin analyzing data on customers' devices in a bid to improve its AI platform, a move designed to safeguard user information while still helping it catch up with AI rivals. From a report: Today, Apple typically trains AI models using synthetic data -- information that's meant to mimic real-world inputs without any personal details. But that synthetic information isn't always representative of actual customer data, making it harder for its AI systems to work properly. The new approach will address that problem while ensuring that user data remains on customers' devices and isn't directly used to train AI models. The idea is to help Apple catch up with competitors such as OpenAI and Alphabet, which have fewer privacy restrictions. The technology works like this: It takes the synthetic data that Apple has created and compares it to a recent sample of user emails within the iPhone, iPad and Mac email app. By using actual emails to check the fake inputs, Apple can then determine which items within its synthetic dataset are most in line with real-world messages.

Read more of this story at Slashdot.

Samsung has paused the global rollout of its One UI 7 update after a serious bug was reported that prevented some Galaxy S24 owners from unlocking their phones. The Verge reports: While the complaints seem to have specifically come from South Korean owners of Galaxy S24 series handsets, Samsung has played it safe and paused the rollout across all models worldwide. While some users will have already downloaded the update to One UI 7, using the app CheckFirm we've confirmed that the update is no longer listed on Samsung's servers as the latest firmware version across several Galaxy devices, with older patches appearing instead. Samsung hasn't confirmed the pause in the rollout, nor plans to issue a fix for users who have already downloaded the One UI 7 update. We've reached out to the company for comment.

Read more of this story at Slashdot.

A new investigation reveals that children as young as five can easily access inappropriate content and interact unsupervised with adults on Roblox, despite the platform's child-friendly image and recent safety updates. The Guardian reports: Describing itself as "the ultimate virtual universe," Roblox features millions of games and interactive environments, known collectively as "experiences." Some of the content is developed by Roblox, but much of it is user-generated. In 2024, the platform had more than 85 million daily active users, an estimated 40% of whom are under 13. While the company said it "deeply sympathized" with parents whose children came to harm on the platform, it said "tens of millions of people have a positive, enriching and safe experience on Roblox every day." However, in an investigation shared with the Guardian, the digital-behavior experts Revealing Reality discovered "something deeply disturbing ... a troubling disconnect between Roblox's child-friendly appearance and the reality of what children experience on the platform." [...] Despite new tools launched last week aimed at giving parents more control over their children's accounts, the researchers concluded: "Safety controls that exist are limited in their effectiveness and there are still significant risks for children on the platform."

Read more of this story at Slashdot.

Intel will sell a 51% stake in its Altera programmable chip unit to private equity firm Silver Lake for $4.46 billion, aiming to cut costs, raise cash, and streamline the company's focus as it shifts toward becoming a contract chip manufacturer. CNBC reports: The deal, announced on Monday, values Altera at $8.75 billion, a sharp decline from the $17 billion Intel paid in 2015. [...] Since last year, Intel has taken steps to spin Altera out as a separate unit and said it planned to sell a portion of its stake. "Today's announcement reflects our commitment to sharpening our focus, lowering our expense structure and strengthening our balance sheet," [CEO Lip-Bu Tan], who took the helm after former top boss Pat Gelsinger's ouster, said. Altera makes programmable chips that can be used for various purposes from telecom equipment to military. Reuters had first reported in November that Silver Lake was among potential suitors competing for a minority stake in Altera. The deal is expected to close in the second half of 2025, after which Intel expects to deconsolidate Altera's financial results from Intel's financial statements, the company said.

Read more of this story at Slashdot.

An anonymous reader quotes a report from the New York Times: British laws restricting what the police can say about criminal cases are "not fit for the social media age (source paywalled; alternative source)," a government committee said in a report released Monday in Britain that highlighted how unchecked misinformation stoked riots last summer. Violent disorder, fueled by the far right, affected several towns and cities for days after a teenager killed three girls on July 29 at a Taylor Swift-themed dance class in Southport, England. In the hours after the stabbings, false claims that the attacker was an undocumented Muslim immigrant spread rapidly online. In a report looking into the riots, a parliamentary committee said a lack of information from the authorities after the attack "created a vacuum where misinformation was able to grow." The report blamed decades-old British laws, aimed at preventing jury bias, that stopped the police from correcting false claims. By the time the police announced the suspect was British-born, those false claims had reached millions. The Home Affairs Committee, which brings together lawmakers from across the political spectrum, published its report after questioning police chiefs, government officials and emergency workers over four months of hearings. Axel Rudakubana, who was sentenced to life in prison for the attack, was born and raised in Britain by a Christian family from Rwanda. A judge later found there was no evidence he was driven by a single political or religious ideology, but was obsessed with violence. [...] The committee's report acknowledged that it was impossible to determine "whether the disorder could have been prevented had more information been published." But it concluded that the lack of information after the stabbing "created a vacuum where misinformation was able to grow, further undermining public confidence," and that the law on contempt was not "fit for the social media age."

Read more of this story at Slashdot.

Several crosswalk buttons in Palo Alto and nearby cities were hacked over the weekend to play deepfake-style satirical audio clips mimicking Elon Musk and Mark Zuckerberg. Authorities have disabled the altered systems, but the identity of the prankster remains unknown. SFGATE reports: Videos of the altered crosswalks began circulating on social media throughout Saturday and Sunday. [...] A city employee was the first to report an issue with one of the signals at University Avenue and High Street in downtown Palo Alto, Horrigan-Taylor told SFGATE via email. Officials later discovered that as many as 12 intersections in downtown Palo Alto had been affected. "The impact is isolated," Horrigan-Taylor said. "Signal operations are otherwise unaffected, and motorists are reminded to always exercise caution around pedestrians." Officials told the outlet they've removed any devices that were tampered with and the compromised voice-over systems have since been disabled, with footage obtained by SFGATE showing several were covered in caution tape, blinking constantly and unpressable.

Read more of this story at Slashdot.

Meta said the company plans to start using data collected from its users in the European Union to train its AI systems. Engadget reports: Starting this week, the tech giant will begin notifying Europeans through email and its family of apps of the fact, with the message set to include an explanation of the kind of data it plans to use as part of the training. Additionally, the notification will link out to a form users can complete to opt out of the process. "We have made this objection form easy to find, read, and use, and we'll honor all objection forms we have already received, as well as newly submitted ones," says Meta. The company notes it will only use data it collects from public posts and Meta AI interactions for training purposes. It won't use private messages in its training sets, nor any interactions, public or otherwise, made by users under the age of 18. As for why the company wants to start using EU data now, it claims the information will allow it to fine tune its future models to better serve Europeans. "We believe we have a responsibility to build AI that's not just available to Europeans, but is actually built for them. That's why it's so important for our generative AI models to be trained on a variety of data so they can understand the incredible and diverse nuances and complexities that make up European communities," Meta states. "That means everything from dialects and colloquialisms, to hyper-local knowledge and the distinct ways different countries use humor and sarcasm on our products. This is particularly important as AI models become more advanced with multi-modal functionality, which spans text, voice, video, and imagery."

Read more of this story at Slashdot.

An anonymous reader quotes a report from DefenseScoop: NATO announced Monday that it has awarded a contract to Palantir to adopt its Maven Smart System for artificial intelligence-enabled battlefield operations. Through the contract, which was finalized March 25, the NATO Communications and Information Agency (NCIA) plans to use a version of the AI system -- Maven Smart System NATO -- to support the transatlantic military organization's Allied Command Operations strategic command. NATO plans to use the system to provide "a common data-enabled warfighting capability to the Alliance, through a wide range of AI applications -- from large language models (LLMs) to generative and machine learning," it said in a release, ultimately enhancing "intelligence fusion and targeting, battlespace awareness and planning, and accelerated decision-making." [...] NATO's Allied Command Operations will begin using Maven within the next 30 days, the organization said Monday, adding that it hopes that using it will accelerate further adoption of emerging AI capabilities. Palantir said the contract "was one of the most expeditious in [its] history, taking only six months from outlining the requirement to acquiring the system."

Read more of this story at Slashdot.

VMware has resumed offering a free hypervisor. News of the offering emerged in a throwaway line in the Release Notes for version 8.0 Update 3e of the Broadcom business unit's ESXi hypervisor. From a report: Just below the "What's New" section of that document is the statement: "Broadcom makes available the VMware vSphere Hypervisor version 8, an entry-level hypervisor. You can download it free of charge from the Broadcom Support portal." VMware offered a free version of ESXi for years, and it was beloved by home lab operators and vAdmins who needed something to tinker with. But in February 2024, VMware discontinued it on grounds that it was dropping perpetual licenses and moving to subscriptions.

Read more of this story at Slashdot.

The European Commission is issuing burner phones and basic laptops to some US-bound staff to avoid the risk of espionage [non-paywalled source], a measure traditionally reserved for trips to China. Financial Times: Commissioners and senior officials travelling to the IMF and World Bank spring meetings next week have been given the new guidance, according to four people familiar with the situation. They said the measures replicate those used on trips to Ukraine and China, where standard IT kit cannot be brought into the countries for fear of Russian or Chinese surveillance. "They are worried about the US getting into the commission systems," said one official. The treatment of the US as a potential security risk highlights how relations have deteriorated since the return of Donald Trump as US president in January. Trump has accused the EU of having been set up to "screw the US" and announced 20 per cent so-called reciprocal tariffs on the bloc's exports, which he later halved for a 90-day period. At the same time, he has made overtures to Russia, pressured Ukraine to hand over control over its assets by temporarily suspending military aid and has threatened to withdraw security guarantees from Europe, spurring a continent-wide rearmament effort. "The transatlantic alliance is over," said a fifth EU official.

Read more of this story at Slashdot.

OpenAI unveiled its GPT-4.1 model family on Monday, prioritizing coding capabilities and instruction following while expanding context windows to 1 million tokens -- approximately 750,000 words. The lineup includes standard GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano variants, all available via API but not ChatGPT. The flagship model scores 54.6% on SWE-bench Verified, lagging behind Google's Gemini 2.5 Pro (63.8%) and Anthropic's Claude 3.7 Sonnet (62.3%) on the same software engineering benchmark, according to TechCrunch. However, it achieves 72% accuracy on Video-MME's long video comprehension tests -- a significant improvement over GPT-4o's 65.3%. OpenAI simultaneously announced plans to retire GPT-4.5 -- their largest model released just two months ago -- from API access by July 14. The company claims GPT-4.1 delivers "similar or improved performance" at substantially lower costs. Pricing follows a tiered structure: GPT-4.1 costs $2 per million input tokens and $8 per million output tokens, while GPT-4.1 nano -- OpenAI's "cheapest and fastest model ever" -- runs at just $0.10 per million input tokens. All models feature a June 2024 knowledge cutoff, providing more current contextual understanding than previous iterations.

Read more of this story at Slashdot.

Apple is readying a substantial overhaul for iPadOS 19 that will transform the tablet experience to function more like macOS, according to Bloomberg. The update will focus on productivity features, multitasking capabilities, and app window management - areas where iPad power users have long requested improvements. The software revamp comes approximately a year after Apple introduced the M4 chip to the iPad Pro lineup and coincides with the expected arrival of new iPad Pro models featuring M5 processors. According to Bloomberg, many users have expressed frustration that iPad hardware capabilities have consistently outpaced software functionality. While the company won't fully port macOS to iPad as some users have wished, the changes will reportedly be substantial enough to satisfy much of the professional user base that has been pushing for more desktop-like functionality. The upcoming changes are expected to be highlighted at Apple's Worldwide Developers Conference in June.

Read more of this story at Slashdot.

An anonymous reader shares a report: Requests to remove books from library shelves are on the rise in the UK, as the influence of pressure groups behind book bans in the US crosses the Atlantic, according to those working in the sector. Although "the situation here is nowhere [near] as bad, censorship does happen and there are some deeply worrying examples of library professionals losing their jobs and being trolled online for standing up for intellectual freedom on behalf of their users," said Louis Coiffait-Gunn, CEO of the Chartered Institute of Library and Information Professionals (Cilip). Ed Jewell, president of Libraries Connected, an independent charity that represents public libraries, said: "Anecdotal evidence from our members suggests that requests to remove books are increasing." The School Library Association (SLA) said this year has seen an "increase in member queries about censorship." Most of the UK challenges appear to come from individuals or small groups, unlike in the US, where 72% of demands to censor books last year were brought forward by organised groups, according to the American Library Association earlier this week. However, evidence suggests that the work of US action groups is reaching UK libraries too. Alison Hicks, an associate professor in library and information studies at UCL, interviewed 10 UK-based school librarians who had experienced book challenges. One "spoke of finding propaganda from one of these groups left on her desk," while another "was directly targeted by one of these groups." Respondents "also spoke of being trolled by US pressure groups on social media, for example when responding to free book giveaways."

Read more of this story at Slashdot.

Blue Origin's New Shepard completed its 31st mission Monday morning, carrying the first all-female crew to space since Soviet cosmonaut Valentina Tereshkova's 1963 solo flight. The NS-31 mission lifted off from West Texas at 9:30 a.m. EDT, with hundreds of thousands watching via livestream as the autonomous vehicle crossed the Karman line 62 miles above Earth. The 10-minute suborbital journey carried six passengers: journalist and Bezos' fiancee Lauren SÃnchez, former NASA scientist Aisha Bowe, bioastronautics researcher Amanda Nguyen, CBS journalist Gayle King, pop star Katy Perry, and film producer Kerianne Flynn. Bowe conducted three research experiments during the flight, while Nguyen became the first Vietnamese and Southeast Asian woman in space. The fully reusable New Shepard system features a pressurized capsule that separates from its booster before returning to Earth with three parachutes.

Read more of this story at Slashdot.

An anonymous reader shares a report: An attorney for the Federal Trade Commission told a judge that Facebook, fearing the competitive threat of Instagram posted to their social media network, acquired both as a way to "neutralize" the rival. "They decided that competition was too hard," the FTC's attorney, Daniel Matheson, said in his opening statement in the government's antitrust case against the Meta Platforms social media empire. He argued that with Meta's monopoly in social media, "consumers do not have reasonable alternatives they can turn to," even as satisfaction has declined. At stake is the potential breakup of Facebook-parent Meta, as the government has zeroed in on the 2012 acquisition of Instagram and 2014 purchase of WhatsApp.

Read more of this story at Slashdot.

Nvidia has announced plans to manufacture AI supercomputers entirely within the United States, commissioning over 1 million square feet of manufacturing space across Arizona and Texas. Production of Blackwell chips has begun at TSMC's Phoenix facilities, while supercomputer assembly will occur at new Foxconn and Wistron plants in Houston and Dallas respectively. "The engines of the world's AI infrastructure are being built in the United States for the first time," said Jensen Huang, Nvidia's founder and CEO. "Adding American manufacturing helps us better meet the incredible and growing demand for AI chips and supercomputers, strengthens our supply chain and boosts our resiliency." The company will deploy its own AI, robotics, and digital twin technologies in these facilities, using Nvidia Omniverse to create digital twins of factories and Isaac GR00T to build manufacturing automation robots. Nvidia projects an ambitious $500 billion in domestic AI infrastructure production over the next four years, with manufacturing expected to create hundreds of thousands of jobs.

Read more of this story at Slashdot.

America's Department of Energy launched a federally funded R&D center in 1946 called the Argonne National Laboratory, and its research became the basis for all of the world's commercial nuclear reactors. But it's now developed an AI-based tool that can "help operators run nuclear plants," reports the Wall Street Journal, citing comments from a senior nuclear engineer in the lab's nuclear science and engineering division: Argonne's plan is to offer the Parameter-Free Reasoning Operator for Automated Identification and Diagnosis, or PRO-AID, to new, tech-forward nuclear builds, but it's also eyeing the so-called dinosaurs, some of which are being resurrected by companies like Amazon and Microsoft to help power their AI data centers. The global push for AI is poised to fuel a sharp rise in electricity demand, with consumption from data centers expected to more than double by the end of the decade, the International Energy Agency said Thursday. The owners of roughly a third of U.S. nuclear plants are in talks with tech companies to provide electricity for those data centers, the Wall Street Journal has reported. PRO-AID performs real-time monitoring and diagnostics using generative AI combined with large language models that notify and explain to staff when something seems amiss at a plant. It also uses a form of automated reasoning — which uses mathematical logic to encode knowledge in AI systems — to mimic the way a human operator asks questions and comes to understand how the plant is operating [according to Richard Vilim, a senior nuclear engineer within the lab's nuclear science and engineering division]. The tool can also help improve the efficiency of the personnel needed to operate a nuclear plant, Vilim said. That's especially important as older employees leave the workforce. "If we can hand off some of these lower-level capabilities to a machine, when someone retires, you don't need to replace him or her," he said... Part of the efficiency in updating technology will come from consolidating the monitoring staff at a utility's nuclear plants at a single, centralized location — much as gas-powered plants already do. It hasn't found its way into a commercial nuclear plant yet, the article acknowledges. But the senior nuclear engineer points out that America's newer gas-powered plants ended up being more automated with digital monitoring tools. Meanwhile the average age of America's 94 operating nuclear reactors is 42 years old, and "nearly all" of them have had their licenses extended, according to the article. (Those nuclear plants still provide almost 20% of America's electricity.)

Read more of this story at Slashdot.

Formula One cars, the world's fastest racecars, need to grip the track for speed and safety on the curves — leading engineers to design cars that create downforce. And racing fans are even told that "a Formula 1 racecar generates enough downforce above a certain speed that it could theoretically drive upside down," writes the automotive site Jalopnik. "McMurtry Automotive turned this theory into reality after having its Spéirling hypercar complete the impressive feat..." Admittedly, the Spéirling's success can be solely attributed to its proprietary 'Downforce-on-Demand' fan system that produces 4,400 pounds of downforce at the push of a button... For those looking to do the math, Spéirling weighs 2,200 pounds. With the stopped car's fan whirling at 23,000 rpm, the rig was rotated to invert the road deck... Then, the hypercar rolled forward a few feet before stopping while inverted. The rig rotated the road deck back down, and the Spéirling drove off like nothing happened. The McMurtry Spéirling, as a 1,000-hp twin-motor electric hypercar, didn't have to clear the other hurdles that an F1 car would have clear to drive upside down. Dry-sump combustion engines aren't designed to run inverted and would eventually fail catastrophically. Oil wouldn't be able to cycle through and keep the engine lubricated. The car is "an electric monster purpose-built to destroy track records," Jalopnik wrote in 2022 when the car shaved more than two seconds off a long-standing record. The "Downforce-on-Demand" feature gives it tremendous acceleration — in nine seconds it can go from 0 to 186.4 mph (300 km/h), according to Jalopnik. "McMurtry is working towards finalizing a production version of its hypercar, called the Spéirling PURE. Only 100 will be produced."

Read more of this story at Slashdot.

We talked about singletons a bit last week. That reminded John of a story from the long ago dark ages where we didn't have always accessible mobile Internet access.

At the time, John worked for a bank. The bank, as all banks do, wanted to sell mortgages. This often meant sending an agent out to meet with customers face to face, and those agents needed to show the customer what their future would look like with that mortgage- payment calculations, and pretty little graphs about equity and interest.

Today, this would be a simple website, but again, reliable Internet access wasn't a thing. So they built a client side application. They tested the heck out of it, and it worked well. Sales agents were happy. Customers were happy. The bank itself was happy.

Time passed, as it has a way of doing, and the agents started clamoring for a mobile web version, that they could use on their phones. Now, the first thought was, "Wire it up to the backend!" but the backend they had was a mainframe, and there was a dearth of mainframe developers. And while the mainframe was the source of truth, and the one place where mortgages actually lived, building a mortgage calculator that could do pretty visualizations was far easier- and they already had one.

The client app was in .NET, and it was easy enough to wrap the mortgage calculation objects up in a web service. A quick round of testing of the service proved that it worked just as well as the old client app, and everyone was happy - for awhile.

Sometimes, agents would run a calculation and get absolute absurd results. Developers, putting in exactly the same values into their test environment wouldn't see the bad output. Testing the errors in production didn't help either- it usually worked just fine. There was a Heisenbug, but how could a simple math calculation that had already been tested and used for years have a Heisenbug?

Well, the calculation ran by simulation- it simply iteratively applied payments and interest to generate the entire history of the loan. And as it turns out, because the client application which started this whole thing only ever needed one instance of the calculator, someone had made it a singleton. And in their web environment, this singleton wasn't scoped to a single request, it was a true global object, which meant when simultaneous requests were getting processed, they'd step on each other and throw off the iteration. And testing didn't find it right away, because none of their tests were simulating the effect of multiple simultaneous users.

The fix was simple- stop being a singleton, and ensure every request got its own instance. But it's also a good example of misapplication of patterns- there was no need in the client app to enforce uniqueness via the singleton pattern. A calculator that holds state probably shouldn't be a singleton in the first place.

[Advertisement] Utilize BuildMaster to release your software with confidence, at the pace your business demands. Download today!

10 years ago "certificate authorities normally issued certificate lifetimes lasting a year or more," remembers a new blog post Thursday by the EFF's engineering director. So in 2015 when the free cert authority Let's Encrypt first started issuing 90-day TLS certificates for websites, "it was considered a bold move, that helped push the ecosystem towards shorter certificate life times." And then this January Let's Encrypt announced new six-day certificates... This week saw a related announcement from the EFF engineering director. More than 31 million web sites maintain their HTTPS certificates using the EFF's Certbot tool (which automatically fetches free HTTPS certificates forever) — and Certbot is now supporting Let's Encrypt's six-day certificates. (It's accomplished through ACME profiles with dynamic renewal at 1/3rd of lifetime left or 1/2 of lifetime left, if the lifetime is shorter than 10 days): There is debate on how short these lifetimes should be, but with ACME profiles you can have the default or "classic" Let's Encrypt experience (90 days) or start actively using other profile types through Certbot with the --preferred-profile and --required-profile flags. For six day certificates, you can choose the "shortlived" profile. Why shorter lifetimes are better (according to the EFF's engineering director): If a certificate's private key is compromised, that compromise can't last as long. With shorter life spans for the certificates, automation is encouraged. Which facilitates robust security of web servers. Certificate revocation is historically flaky. Lifetimes 10 days and under prevent the need to invoke the revocation process and deal with continued usage of a compromised key.

Read more of this story at Slashdot.