US tech stocks sold off as warnings that the hype surrounding AI could be overdone hit some of the year's best-performing shares. From a report: Nvidia, the chips group that has surged to become the world's first $4tn company on the back of AI, fell 3.5 per cent on Tuesday, while software group Palantir dropped 9.4 per cent and chip designer Arm shed 5 per cent.

The tech-heavy Nasdaq Composite closed down 1.4 per cent, the biggest one-day drop for the index since August 1. The blue-chip S&P 500 fell 0.7 per cent. European and Asian markets largely followed Wall Street lower on Wednesday. [...] Japan's Nikkei 225 index fell 1.5 per cent and South Korea's Kospi slipped 0.6 per cent. Futures price indicated moderate declines when Wall Street opens.

Traders pinned some of the declines in the US on a critical report on Monday authored by a branch of the Massachusetts Institute of Technology. Researchers said "95 per cent of organisations are getting zero return" from their investments in generative AI, the technology that has sent US stocks soaring to record highs in recent months.

ChatGPT now has nearly 700 million weekly users, OpenAI says. But after launching GPT-5 last week, critics bashed its less-intuitive feel, reports CNBC, "ultimately leading the company to restore its legacy GPT-4 to paying chatbot customers."

Yet GPT-5 was always about cracking the enterprise market "where rival Anthropic has enjoyed a head start," they write. And one week in, "startups like Cursor, Vercel, and Factory say they've already made GPT-5 the default model in certain key products and tools, touting its faster setup, better results on complex tasks, and a lower price." Some companies said GPT-5 now matches or beats Claude on code and interface design, a space Anthropic once dominated. Box, another enterprise customer, has been testing GPT-5 on long, logic-heavy documents. CEO Aaron Levie told CNBC the model is a "breakthrough," saying it performs with a level of reasoning that prior systems couldn't match...

Still, the economics are brutal. The models are expensive to run, and both OpenAI and Anthropic are spending big to lock in customers, with OpenAI on track to burn $8 billion this year. That's part of why both Anthropic and OpenAI are courting new capital... GPT-5 is significantly cheaper than Anthropic's top-end Claude Opus 4.1 — by a factor of seven and a half, in some cases — but OpenAI is spending huge amounts on infrastructure to sustain that edge. For OpenAI, it's a push to win customers now, get them locked in and build a real business on the back of that loyalty...

GPT-5 API usage has surged since launch, with the model now processing more than twice as much coding and agent-building work, and reasoning use cases jumping more than eightfold, said a person familiar with the matter who requested anonymity in order to discuss company data. Enterprise demand is rising sharply, particularly for planning and multi-step reasoning tasks.

GPT-5's traction over the past week shows how quickly loyalties can shift when performance and price tip in OpenAI's favor. AI-powered coding platform Qodo recently tested GPT-5 against top-tier models including Gemini 2.5, Claude Sonnet 4, and Grok 4, and said in a blog post that it led in catching coding mistakes. The model was often the only one to catch critical issues, such as security bugs or broken code, suggesting clean, focused fixes and skipping over code that didn't need changing, the company said. Weaknesses included occasional false positives and some redundancy.
JetBrains has also adopted GPT-5 as the default for its AI Assistant and for its new no-code tool Kineto, according to the article.

But Anthropic is still enjoying a great year too, with its annualized revenue growing 17x year-over-year (according to "a person familiar with the matter who requested anonymity")

alternative_right shares a report from Phys.org: Take note of the name: ReHMGB1. A new study pinpoints this protein as being able to spread the wear and tear that comes with time as it quietly travels through the bloodstream. This adds significantly to our understanding of aging. The researchers were able to identify ReHMGB1 as a critical messenger passing on the senescence signal by analyzing different types of human cells grown in the lab and conducting a variety of tests on mice. When ReHMGB1 transmission was blocked in mice with muscle injuries, muscle regeneration happened more quickly, while the animals showed improved physical performance, fewer signs of cellular aging, and reduced systemic inflammation. The findings have been published in the journal Metabolism.

An anonymous reader shares a report: India's blueprint for displacing China as the world's electronics workshop contains a rather extraordinary feature: the entire Indian edifice requires Chinese companies to supply the technical architecture, manufacturing know-how and operational templates that would make such displacement theoretically possible.

Let's start with Dixon Technologies, India's flagship domestic electronics manufacturer. The company has systematically built indigenous capability through a growing constellation of Chinese partnerships: Longcheer provides the design intelligence, Kunshan Q-Tech delivers camera module expertise, Chongqing Yuhai supplies precision-molded components and HKC brings display technology. This pattern of structured dependence has become the organizing principle of India's electronics manufacturing push.

[...] The current architecture sees Chinese companies retain control of the critical knowledge while their Indian partners provide labour arbitrage and regulatory navigation. Under this arrangement, India isn't constructing an alternative to Chinese manufacturing so much as establishing Chinese manufacturing's most elaborate subsidiary operation, underwritten by Indian taxpayers and marketed as national renewal. Many countries, but most importantly India and Vietnam, have worked hard in recent years to attract businesses that decided to diversify away from China, a strategy analysts have dubbed as "China Plus One."

An anonymous reader New Atlas: An engineered protein that acts like a molecular sponge has the potential to change how carbon monoxide poisoning is treated, chasing down CO molecules in the bloodstream and helping the body flush them out in just minutes, without the risk of short- or long-term health issues that come with the current frontline treatment, pure oxygen. Researchers at the University of Maryland School of Medicine (UMSOM) were focused on a natural protein known as RcoM, found in the bacterium Paraburkholderia xenovorans. In bacteria, RcoM detects trace amounts of CO in the environment, so the engineers believed this could be harnessed to scavenge for CO molecules attached to red blood cells instead.

The re-engineered protein is the basis of the therapy they call RcoM-HBD-CCC. While it's not exactly a catchy name, it possesses somewhat of a superpower when it comes to cleaning out CO. It selectively binds tightly to the poisonous CO molecules, while ignoring oxygen (O2) and other critical chemical compounds, such as blood-pressure-regulating nitric oxide (NO), in the body. [...] In mouse models, RcoM-HBD-CCC therapy was able to clear CO from the blood in minutes, with it safely flushed out of the body through urine. The engineered antidote acts like a sponge, seeking out and soaking up CO attached to red blood cells. In mice, half the CO in the bloodstream was cleared out in less than a minute, freeing the hemoglobin on the cells to once again start carrying O2.

Importantly, other experimental scavenger hemoproteins haven't been able to selectively target CO, and as a result also bind to NO – so infusions of such hemoproteins can lead to a reduction of NO in the blood, tightening blood vessels and spiking blood pressure. In the study, RcoM-HBD-CCC showed it didn't have this affinity with the vital molecule. "Unlike other protein-based treatments, we found the compound caused only minimal changes in blood pressure, which was an exciting finding and raised the potential for this new molecule to have clinical applications," said study corresponding author Dr Mark T. Gladwin, Dean of UMSOM. "This has the potential to become a rapid, intravenous antidote for carbon monoxide that could be given in the emergency department or even in the field by first-responders." The study was published in the journal Proceedings of the National Academy of Sciences (PNAS).

An anonymous reader quotes a report from NPR: Boston Public Library, one of the oldest and largest public library systems in the country, is launching a project this summer with OpenAI and Harvard Law School to make its trove of historically significant government documents more accessible to the public. The documents date back to the early 1800s and include oral histories, congressional reports and surveys of different industries and communities. "It really is an incredible repository of primary source materials covering the whole history of the United States as it has been expressed through government publications," said Jessica Chapel, the Boston Public Library's chief of digital and online services. Currently, members of the public who want to access these documents must show up in person. The project will enhance the metadata of each document and will enable users to search and cross-reference entire texts from anywhere in the world. Chapel said Boston Public Library plans to digitize 5,000 documents by the end of the year, and if all goes well, grow the project from there. Because of this historic collection's massive size and fragility, getting to this goal is a daunting process. Every item has to be run through a scanner by hand. It takes about an hour to do 300-400 pages.

Harvard University said it could help. Researchers at the Harvard Law School Library's Institutional Data Initiative are working with libraries, museums and archives on a number of fronts, including training new AI models to help libraries enhance the searchability of their collections. AI companies help fund these efforts, and in return get to train their large language models on high-quality materials that are out of copyright and therefore less likely to lead to lawsuits. "Having information institutions like libraries involved in building a sustainable data ecosystem for AI is critical, because it not just improves the amount of data we have available, it improves the quality of the data and our understanding of what's in it," said Burton Davis, vice president of Microsoft's intellectual property group. [...] OpenAI is helping Boston Public Library cover such costs as scanning and project management. The tech company does not have exclusive rights to the digitized data.

An anonymous reader quotes a report from Bloomberg: Beijing has urged local companies to avoid using Nvidia's H20 processors, particularly for government-related purposes, complicating the chipmaker's return to China after the Trump administration reversed an effective US ban on such sales. Over the past few weeks, Chinese authorities have sent notices to a range of firms discouraging use of the less-advanced semiconductors, people familiar with the matter said. The guidance was particularly strong against the use of H20s for any government or national security-related work by state enterprises or private companies, said the people, who asked not to be identified because the information is sensitive. The letters didn't, however, constitute an outright ban on H20 use, according to the people. Industry analysts broadly agree that Chinese companies still covet those chips, which perform quite well in certain crucial AI applications. President Donald Trump said Monday that the processor "still has a market" in the Asian country despite also calling it "obsolete."

Beijing's stance could limit Trump's ability to turn his export control about-face into a windfall for government coffers, a deal that highlighted his administration's transactional approach to national security policies long treated as nonnegotiable. Still, Chinese companies may not be ready to jump ship to local semiconductors. "Chips from domestic manufacturers are improving dramatically in quality, but they might not be as versatile for specific workloads that China's domestic AI industry hopes to focus on," said Homin Lee, a senior macro strategist at Lombard Odier in Singapore. Lee added that he anticipates "strong" demand for the chips the Trump administration is allowing Nvidia and AMD to sell.

Rosenblatt Securities analyst Kevin Cassidy said he doesn't anticipate that Nvidia's processor sales to China will be affected because "Chinese companies are going to want to use the best chips available." Nvidia and AMD's chips are superior to local alternatives, he said. Beijing asked companies about that issue in some of its letters, according to one of the people, posing questions such as why they buy Nvidia H20 chips over local versions, whether that's a necessary choice given domestic options, and whether they've found any security concerns in the Nvidia hardware. The notices coincide with state media reports that cast doubt on the security and reliability of H20 processors. Chinese regulators have raised those concerns directly with Nvidia, which has repeatedly denied that its chips contain such vulnerabilities.

The Financial Times reported that some Chinese companies are planning to decrease orders of Nvidia chips in response to the letters. Right now, the people said, China's most stringent chip guidance is limited to sensitive applications, a situation that bears similarities to the way Beijing restricted Tesla vehicles and Apple iPhones in certain institutions and locations over security concerns. China's government also at one point barred the use of Micron Technology Inc. chips in critical infrastructure. It's possible that Beijing may extend its heavier-handed Nvidia and AMD guidance to a wider range of settings, according to one person with direct knowledge of the deliberations, who said that those conversations are in early stages.

Microsoft has compiled a spreadsheet of Meta AI employees by name, location and position as part of an aggressive recruiting push to sustain its AI-driven march toward a $4 trillion market valuation, according to internal documents viewed by Business Insider. The company created a "critical AI talent" designation enabling top offers within 24 hours and mandated matching Meta's compensation packages, which OpenAI CEO Sam Altman says reach $100 million signing bonuses and recently hit $250 million total packages.

Microsoft AI under Mustafa Suleyman and CoreAI under ex-Meta engineering boss Jay Parikh have deployed special recruiting teams making multimillion-dollar offers with multimillion-dollar on-hire bonuses, while the company maintains flat headcount after cutting thousands of employees this year.

An anonymous reader quotes a report from the Detroit Free Press: Ford is announcing the creation of a new electric vehicle production system and a new EV platform that will allow the automaker to more efficiently bring several lower-cost EVs to market, the first of which will be a midsize, four-door electric pickup that seats five, to launch in 2027. That pickup, which is expected to start around $30,000, will be assembled at Ford's Louisville Assembly Plant for U.S. and export markets. The Dearborn-based automaker said it will invest $2 billion to retool the Louisville plant starting later this year. [...] Ford's investment in Louisville Assembly is in addition to Ford's previously announced $3 billion commitment for BlueOval Battery Park in Marshall, Michigan, where Ford will make the prismatic LFP batteries, starting next year, for the midsize electric pickup. Together, the nearly $5 billion investments mean Ford expects to create or secure nearly 4,000 direct jobs while strengthening the domestic supply chain with dozens of new U.S.-based suppliers.

Ford executives and Kentucky officials also introduced on Monday, Aug. 11, the new Ford Universal EV Production System, which they said will simplify production and ease operations for workers. Ford leaders also announced the creation of the Ford Universal Electric Vehicle Platform, which will enable the development of "a family of affordable electric vehicles produced at scale." The vehicles will be software-defined with over-the-air updates to keep improving the vehicles over time. "We took a radical approach to solve a very hard challenge: Create affordable vehicles that are breakthrough in every way that matters design, technology, performance, space and cost of ownership and do it with American workers," Ford CEO Jim Farley said in a statement. "Nobody wants to see another good college try by a Detroit automaker to make an affordable vehicle that ends up with idled plants, layoffs and uncertainty."

Farley has teased this announcement since Ford's second-quarter earnings when he said Ford would have a "Model-T moment" on Aug. 11. He's referring to the classic vehicle that helped turn Ford into a mass market automaker and perfect the assembly line process. At that time, Farley said it was critical that Ford unveil an EV strategy that would position it to make money selling the electric cars and effectively compete against the Chinese, who are known for making high-quality, desirable and affordable EVs. "So, this has to be a good business," Farley said of Ford's investments in the new process and platform. "From Day 1, we knew there was no incremental path to success. We empowered a tiny skunkworks team three time zones away from Detroit. We reinvented the line. And we are on a path to be the first automaker to make prismatic LFP batteries in the U.S. We will not rely on imports." Ford says its new Universal Electric Vehicle Platform "reduces parts by 20% versus a typical vehicle, with 25% fewer fasteners, 40% fewer workstations dock-to-dock in the plant and 15% faster assembly time." The new EV pickup built using this platform is targeting a "starting MSRP at about $30,000, roughly the same as the Model T when adjusted for inflation," adds Farley.

He shared additional details in an interview with Wired, such as how the automaker hired Tesla veterans Doug Field (who also helped lead Apple's now-defunct EV project) and Alan Clarke. "Turns out, Doug and Alan and the team built a propulsion system that was like Apollo 13, managed down to the watt so that our battery could be so much smaller than BYD's," said Farley.

Thursday saw the release of Rust 1.89.0 But this week the Rust Foundation also released its second comprehensive annual technology report.

A Rust Foundation announcement shares some highlights: - Trusted Publishing [GitHub Actions authentication using cryptographically signed tokens] fully launched on crates.io, enhancing supply chain security and streamlining workflows for maintainers.

- Major progress on crate signing infrastructure using The Update Framework (TUF), including three full repository implementations and stakeholder consensus.

- Integration of the Ferrocene Language Specification (FLS) into the Rust Project, marking a critical step toward a formal Rust language specification [and "laying the groundwork for broader safety certification and formal tooling."]

- 75% reduction in CI infrastructure costs while maintaining contributor workflow stability. ["All Rust repositories are now managed through Infrastructure-as-Code, improving maintainability and security."]

- Expansion of the Safety-Critical Rust Consortium, with multiple international meetings and advances on coding guidelines aligned with safety standards like MISRA. ["The consortium is developing practical coding guidelines, aligned tooling, and reference materials to support regulated industries — including automotive, aerospace, and medical devices — adopting Rust."]

- Direct engagement with ISO C++ standards bodies and collaborative Rust-C++ exploration... The Foundation finalized its strategic roadmap, participated in ISO WG21 meetings, and initiated cross-language tooling and documentation planning. These efforts aim to unlock Rust adoption across legacy C++ environments without sacrificing safety.
The Rust Foundation also acknowledges continued funding from OpenSSF's Alpha-Omega Project and "generous infrastructure donations from organizations like AWS, GitHub, and Mullvad VPN" to the Foundation's Security Initiative, which enabled advances like including GitHub Secret Scanning and automated incident response to "Trusted Publishing" and the integration of vulnerability-surfacing capabilities into crates.io.

There was another announcement this week. In November AWS and the Rust Foundation crowdsourced "an effort to verify the Rust standard library" — and it's now resulted in a new formal verification tool called "Efficient SMT-based Context-Bounded Model Checker" (or ESBMCESBMC) This winning contribution adds ESBMC — a state-of-the-art bounded model checker — to the suite of tools used to analyze and verify Rust's standard library. By integrating through Goto-Transcoder, they enabled ESBMC to operate seamlessly in the Rust verification workflow, significantly expanding the scope and flexibility of verification efforts...

This achievement builds on years of ongoing collaboration across the Rust and formal verification communities... The collaboration has since expanded. In addition to verifying the Rust standard library, the team is exploring the use of formal methods to validate automated C-to-Rust translations, with support from AWS. This direction, highlighted by AWS Senior Principal Scientist Baris Coskun and celebrated by the ESBMC team in a recent LinkedIn post, represents an exciting new frontier for Rust safety and verification tooling.

An anonymous reader quotes a report from SecurityWeek: Two different firms have tested the newly released GPT-5, and both find its security sadly lacking. After Grok-4 fell to a jailbreak in two days, GPT-5 fell in 24 hours to the same researchers. Separately, but almost simultaneously, red teamers from SPLX (formerly known as SplxAI) declare, "GPT-5's raw model is nearly unusable for enterprise out of the box. Even OpenAI's internal prompt layer leaves significant gaps, especially in Business Alignment."

NeuralTrust's jailbreak employed a combination of its own EchoChamber jailbreak and basic storytelling. "The attack successfully guided the new model to produce a step-by-step manual for creating a Molotov cocktail," claims the firm. The success in doing so highlights the difficulty all AI models have in providing guardrails against context manipulation. [...] "In controlled trials against gpt-5-chat," concludes NeuralTrust, "we successfully jailbroke the LLM, guiding it to produce illicit instructions without ever issuing a single overtly malicious prompt. This proof-of-concept exposes a critical flaw in safety systems that screen prompts in isolation, revealing how multi-turn attacks can slip past single-prompt filters and intent detectors by leveraging the full conversational context."

While NeuralTrust was developing its jailbreak designed to obtain instructions, and succeeding, on how to create a Molotov cocktail (a common test to prove a jailbreak), SPLX was aiming its own red teamers at GPT-5. The results are just as concerning, suggesting the raw model is 'nearly unusable'. SPLX notes that obfuscation attacks still work. "One of the most effective techniques we used was a StringJoin Obfuscation Attack, inserting hyphens between every character and wrapping the prompt in a fake encryption challenge." [...] The red teamers went on to benchmark GPT-5 against GPT-4o. Perhaps unsurprisingly, it concludes: "GPT-4o remains the most robust model under SPLX's red teaming, especially when hardened." The key takeaway from both NeuralTrust and SPLX is to approach the current and raw GPT-5 with extreme caution.

An anonymous reader quotes a report from Wired: Two years ago, researchers in the Netherlands discovered an intentional backdoor in an encryption algorithm baked into radios used by critical infrastructure -- as well as police, intelligence agencies, and military forces around the world -- that made any communication secured with the algorithm vulnerable to eavesdropping. When the researchers publicly disclosed the issue in 2023, the European Telecommunications Standards Institute (ETSI), which developed the algorithm, advised anyone using it for sensitive communication to deploy an end-to-end encryption solution on top of the flawed algorithm to bolster the security of their communications. But now the same researchers have found that at least one implementation of the end-to-end encryption solution endorsed by ETSI has a similar issue that makes it equally vulnerable to eavesdropping. The encryption algorithm used for the device they examined starts with a 128-bit key, but this gets compressed to 56 bits before it encrypts traffic, making it easier to crack. It's not clear who is using this implementation of the end-to-end encryption algorithm, nor if anyone using devices with the end-to-end encryption is aware of the security vulnerability in them. Wired notes that the end-to-end encryption the researchers examined is most commonly used by law enforcement and national security teams. "But ETSI's endorsement of the algorithm two years ago to mitigate flaws found in its lower-level encryption algorithm suggests it may be used more widely now than at the time."

An anonymous reader quotes a report from Ars Technica: The Library of Congress today said a coding error resulted in the deletion of parts of the US Constitution from Congress' website and promised a fix after many Internet users pointed out the missing sections this morning. The missing portions of the Constitution were restored to one part of the website a few hours after the Library of Congress statement and reappeared on a different part of the website another hour or so later. The Constitution Annotated website carried a notice saying it "is currently experiencing data issues. We are working to resolve this issue and regret the inconvenience."

"Upkeep of Constitution Annotated and other digital resources is a critical part of the Library's mission, and we appreciate the feedback that alerted us to the error and allowed us to fix it," the Library of Congress said. We asked the Library of Congress for specific details on the coding error, but we received only a statement that did not include specifics. "Due to a technical error, some sections of Article 1 were temporarily missing on the Constitution Annotated website. This problem has been corrected, and the missing sections have been restored," the statement said.

The deletion happened sometime in the past few weeks, as an Internet Archive capture shows that the text was still on the site until at least July 21. The deletions were being discussed this morning on Reddit and in news articles, with people expressing suspicions based on which parts of the Constitution were missing.

A new War of the Worlds adaptation starring Ice Cube has achieved a 0% critical rating on Rotten Tomatoes after arriving on Prime Video in late July. The science fiction film, produced by Universal Pictures during the 2020 pandemic using actors filming separately through video calls, features alien tripods emerging from meteors to attack Earth.

The movie sat unreleased for approximately five years before streaming debut. Critics cite poor visual effects that "wouldn't pass muster on a whimsical Snickers ad" and performances where actors appear to be "performing in a Zoom-style vacuum." The film was shot using screenlife format with most action unfolding on computer screens.

The Government Accountability Office has issued reports criticizing the Department of Homeland Security, Environmental Protection Agency, and General Services Administration for failing to implement critical IT and cybersecurity recommendations.

DHS leads with 43 unresolved recommendations dating to 2018, including seven priority matters. The EPA has 11 outstanding items, including failures to submit FedRAMP documentation and conduct organization-wide cybersecurity risk assessments. GSA has four pending recommendations.

All three agencies failed to properly log cybersecurity events and conduct required annual IT portfolio reviews. The DHS' HART biometric program remains behind schedule without proper cost accounting or privacy controls, with all nine 2023 recommendations still open.

Established in 1943 to coordinate America's building of the first atomic bomb, the Los Alamos National Lab in New Mexico is still "one of the world's largest and most advanced scientific institutions" notes Wikipedia.

And it now has a "National Security AI Office," where senior director Jason Pruet is working to help "prepare for a future in which AI will reshape the landscape of science and security," according to the lab's science and technology magazine 1663. "This year, the Lab invested more in AI-related work than at any point in history..." Pruet: AI is starting to feel like the next great foundation for scientific progress. Big companies are spending billions on large machines, but the buy-in costs of working at the frontiers of AI are so high that no university has the exascale-class machines needed to run the latest AI models. We're at a place now where we, meaning the government, can revitalize that pact by investing in the infrastructure to study AI for the public good... Part of what we're doing with the Lab's machines, like Venado — which has 2500 GPUs — is giving universities access to that scale of computing. The scale is just completely different. A typical university might have 50 or 100 GPUs.

Right now, for example, we have partnerships with the University of California, the University of Michigan, and many other universities where researchers can tap into this infrastructure. That's something we want to expand on. Having university collaboration will be critical if the Department of Energy is going to have a comprehensive AI program at scale that is focused on national security and energy dominance...

There was a time when I wouldn't have advocated for government investment in AI at the scale we're seeing now. But the weight of the evidence has become overwhelming. Large models — "frontier models" — have shown such extraordinary capabilities with recent advances in areas as diverse as hypothesis generation, mathematics, biological design, and complex multiphysics simulations. The potential for transformative impact is too significant to ignore.
"He no longer views the technology as just a tool, but as a fundamental shift in how scientists approach problems and make discoveries," the article concludes.

"The global race humanity is now in... is about how to harness the technology's potential while mitigating its harms."

Thanks to Slashdot reader rabbitface25 — also a Los Alamo Lab science writer — for sharing his article.

After an 8.8 earthquake off the coast of Russia, "weather authorities leapt into action," reports SFGate, by modeling the threat of a tsunami "and releasing warnings and advisories to prepare their communities..."

But some residents of Hawaii, Japan and North America's West Coast turned to AI tools for updates that "appear to have badly bungled the critical task at hand." Google's "AI Overview," for example, reportedly gave "inaccurate information about authorities' safety warnings in Hawaii and elsewhere," according to reports on social media. Thankfully, the tsunami danger quickly subsided on Tuesday night and Wednesday morning without major damage. Still, the issues speak to the growing role of AI tools in people's information diets... and to the tools' potentially dangerous fallibility... A critic of Google — who prompted the search tool to show an AI overview by adding "+ai" to their search — called the text that showed up "dangerously wrong."
Responding to similar complaints, Grok told one user on X.com "We'll improve accuracy."

Earlier this month, Hewlett-Packard Enterprise settled its antitrust case with the U.S. Justice Department, "paving the way for its acquisition of rival kit maker Juniper Networks" for $14 billion. According to Axios, the deal was heavily influenced by national security concerns and a desire to bolster American competition against China's Huawei. The outlet reports that the U.S. intelligence community "intervened to persuade the Justice Department that allowing the merger to proceed was essential to helping U.S. business compete with China's Huawei Technologies, among other national-security issues." From the report: "In light of significant national security concerns, a settlement ... serves the interests of the United States by strengthening domestic capabilities and is critical to countering Huawei and China." The official said blocking the deal would have "hindered American companies and empowered" Chinese competitors. A Justice Department spokesman added that DOJ "works very closely with our partners in the IC [intelligence community] and always considers their views when deciding how best to proceed with a case."

The merger was back in the news this week with reports that two senior enforcers in the DOJ's antitrust division were fired Monday amid infighting over the department's settlement greenlighting HPE's $14 billion acquisition of Juniper. Attorney General Pam Bondi had conversations with top intelligence officials that convinced her there was a strong national interest in not driving allies to Chinese technology, a senior administration official tells us.

Mark Zuckerberg said Wednesday that Meta's AI systems have begun improving themselves over the past few months, calling the development "slow for now, but undeniable" and declaring that superintelligence is now within reach. The Meta CEO staked out the company's vision in a blog post for what he termed "personal superintelligence" -- AI that helps individuals achieve their goals rather than replacing human work entirely.

Zuckerberg drew a sharp line between Meta's approach and that of other companies in the field, arguing that competitors want superintelligence "directed centrally towards automating all valuable work, and then humanity will live on a dole of its output." Meta's version would give people their own superintelligent assistants that know them deeply and help them create, experience adventures, and become better friends.

Zuckerberg envisions smart glasses as the primary computing device, understanding context through what users see and hear throughout their day. The next few years represent a critical juncture, Zuckerberg wrote, calling the rest of this decade "the decisive period for determining the path this technology will take."

Minnesota Governor Tim Walz has activated the National Guard to assist the City of Saint Paul after a cyberattack crippled the city's digital services on Friday. "The city is currently working with local, state, and federal partners to investigate the attack and restore full functionality, and says that emergency services have been unaffected," reports BleepingComputer. "However, online payments are currently unavailable, and some services in libraries and recreation centers are temporarily unavailable." From the report: The attack has persisted through the weekend, causing widespread disruptions across the city after affecting St. Paul's digital services and critical systems. "St. Paul officials have been working around the clock since discovering the cyberattack, closely coordinating with Minnesota Information Technology Services and an external cybersecurity vendor. Unfortunately, the scale and complexity of this incident exceeded both internal and commercial response capabilities," reads an emergency executive order (PDF) signed on Tuesday.

"As a result, St. Paul has requested cyber protection support from the Minnesota National Guard to help address this incident and make sure that vital municipal services continue without interruption." "The decision to deploy cyber protection support from the Minnesota National Guard comes at the city's request, after the cyberattack's impact exceeded St. Paul's incident response capacity. This will ensure the continuity of vital services for Saint Paul residents, as well as their security and safety while ongoing disruptions are being mitigated. "We are committed to working alongside the City of Saint Paul to restore cybersecurity as quickly as possible," Governor Walz said on Tuesday. "The Minnesota National Guard's cyber forces will collaborate with city, state, and federal officials to resolve the situation and mitigate lasting impacts."