A "small percentage" of Google's users in the US will begin seeing an AI Mode tab in Google Search "in the coming weeks," the company said Thursday, marking the tool's first deployment outside the company's experimental Labs environment.

Unlike traditional search results that display URLs based on user queries, AI Mode generates conversational responses from Google's search index. The feature will appear as a dedicated tab positioned before the standard "All," "Images," and other search filters. The deployment represents Google's direct challenge to LLM-powered search engines like Perplexity and ChatGPT.

AI Mode differs from existing AI Overviews in Google Search, which merely insert AI summaries between the search box and web results.

NIST: Clocks on Earth are ticking a bit more regularly thanks to NIST-F4, a new atomic clock at the National Institute of Standards and Technology (NIST) campus in Boulder, Colorado. This month, NIST researchers published a journal article establishing NIST-F4 as one of the world's most accurate timekeepers. NIST has also submitted the clock for acceptance as a primary frequency standard by the International Bureau of Weights and Measures (BIPM), the body that oversees the world's time.

NIST-F4 measures an unchanging frequency in the heart of cesium atoms, the internationally agreed-upon basis for defining the second since 1967. The clock is based on a "fountain" design that represents the gold standard of accuracy in timekeeping. NIST-F4 ticks at such a steady rate that if it had started running 100 million years ago, when dinosaurs roamed, it would be off by less than a second today.

By joining a small group of similarly elite time pieces run by just 10 countries around the world, NIST-F4 makes the foundation of global time more stable and secure. At the same time, it is helping to steer the clocks NIST uses to keep official U.S. time. Distributed via radio and the internet, official U.S. time is critical for telecommunications and transportation systems, financial trading platforms, data center operations and more.

For the first time -- and despite fears that it might never happen -- nations have agreed a series of measures to prevent, prepare for and respond to pandemics. Nature: The terms of the first global pandemic accord were still being hashed out at the World Health Organization (WHO) headquarters in Geneva, Switzerland, up until the early hours of 16 April. "This is a definitive moment in the history of global health," says Lawrence Gostin, a specialist in health law and policy at Georgetown University in Washington DC, who followed the negotiations closely. The accord "sets out some very important norms to keep the world safe," he says.

The accord was agreed without the United States, which withdrew from the pandemic treaty the day that President Trump was inaugurated. This reduces its power, says Gostin, but is also a source of strength. "Instead of collapsing in the face of President Trump's assault on global health, the world came together." The treaty is not perfect but represents a major achievement, says Michelle Childs, policy advocacy director at the non-profit organization Drugs for Neglected Diseases Initiative in Geneva. "People didn't think that they'd get to this stage of agreeing at all."

Alamo Drafthouse will not implement Meta's new Movie Mate technology during the April 30 nationwide rerelease of Blumhouse's "M3GAN," Variety reports. The specialty theater chain confirmed it will maintain its strict no-phones policy despite Universal's promotion of the second-screen experience, with staff instructed to remove patrons attempting to access the feature during screenings.

Movie Mate represents Meta's first integration of its interactive movie technology, which operates via Instagram direct messaging. Users message the film's official account to activate a chatbot delivering "sneak peeks, exclusive recorded messages from directors and talent" synchronized with the screening. The "M3GAN" event serves as Meta's technological debut ahead of potential wider theatrical implementation.

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."

"Can you help me take down the Texas lottery?"

That's what a London banker-turned-bookmaker asked "acquaintances" in 2023, reports the Wall Street Journal. The plan was to buy "nearly every possible number in a coming drawing" — purchasing $1 tickets for 25.8 million possible combinations, since "The jackpot was heading to $95 million. If nobody else also picked the winning numbers, the profit would be nearly $60 million." Marantelli flew to the U.S. with a few trusted lieutenants. They set up shop in a defunct dentist's office, a warehouse and two other spots in Texas. The crew worked out a way to get official ticket-printing terminals. Trucks hauled in dozens of them and reams of paper... [Then Texas announced no winner in an earlier lottery, rolling its jackpot into another drawing three days later.] The machines — manned by a disparate bunch of associates and some of their children — screeched away nearly around the clock, spitting out 100 or more tickets every second. Texas politicians later likened the operation to a sweatshop.

Trying to pull off the gambit required deep pockets and a knack for staying under the radar — both hallmarks of the secretive Tasmanian gambler who bankrolled the operation. Born Zeljko Ranogajec, he was nicknamed "the Joker" for his ability to pull off capers at far-flung casinos and racetracks. Adding to his mystique, he changed his name to John Wilson several decades ago. Among some associates, though, he still goes by Zeljko, or Z. Over the years, Ranogajec and his partners have won hundreds of millions of dollars by applying Wall Street-style analytics to betting opportunities around the world. Like card counters at a blackjack table, they use data and math to hunt for situations ripe for flipping the house edge in their favor. Then they throw piles of money at it, betting an estimated $10 billion annually.

The Texas lottery play, one of their most ambitious operations ever, paid off spectacularly with a $57.8 million jackpot win. That, in turn, spilled their activities into public view and sparked a Texas-size uproar about whether other lotto players — and indeed the entire state — had been hoodwinked. Early this month, the state's lieutenant governor, Dan Patrick, called the crew's win "the biggest theft from the people of Texas in the history of Texas." In response to written questions addressed to Marantelli and Ranogajec, Glenn Gelband, a New Jersey lawyer who represents the limited partnership that claimed the Texas prize, said "all applicable laws, rules and regulations were followed...."

Lottery officials and state lawmakers have taken steps to prevent a repeat.
The article also looks at a group of Princeton University graduates calling themselves Black Swan Capital that's "won millions in recent years" by targetting state lottery drawings with unusually favorable odds.

"State lottery directors say they are seeing more organized efforts to buy lottery tickets in bulk," according to the article, "but that the groups are largely operating legally and transparently..."

Facebook Marketplace has emerged as the dominant feature within the social media platform, amassing 1.2 billion monthly active buyers by 2023 and overtaking eBay as a peer-to-peer selling platform. According to recent data, approximately 16 percent of Facebook's monthly active users now access the site exclusively to participate in Marketplace.

The feature's growth accelerated following the pandemic's supply chain disruptions and subsequent inflation, which increased demand for used goods. Facebook reports that Marketplace is attracting younger demographics who have otherwise abandoned the platform's social features.

This shift represents a fundamental transformation of Facebook's core function from "digital connector" to "digital bazaar," with the platform increasingly hosting transactions rather than social connections.

The ubiquitous but often overlooked Wi-Fi router lies at the heart of one of Washington's biggest national security dilemmas -- and a rift between two brothers on opposite sides of the Pacific. From a report: US investigators are probing the China ties of TP-Link, the new American incarnation of a consumer Wi-Fi behemoth, following its rapid growth and a spate of cyber attacks by Chinese state-sponsored actors targeting many router brands. The inquiry is testing whether TP-Link's corporate makeover represents enough of a divorce from China to spare it from a ban in a crucial market.

While TP-Link's recent restructuring split the company into separate US- and China-headquartered businesses, a Bloomberg News investigation found that the resulting American venture still has substantial operations in mainland China. If US officials conclude TP-Link's China connections pose an "unacceptable risk," they could use a powerful new authority to ban the company from the US. Such an outcome could also unravel plans by the owner of its US business, Jeffrey Chao, to start fresh in California following an estrangement from his older brother, who started the router business with him in Shenzhen nearly three decades ago.

In an interview -- the first Jeffrey Chao said he has ever given -- he told Bloomberg he's quitting China. He opened a new headquarters in Irvine last year and said he will invest $700 million in the US to build a factory and jumpstart research and development on highly secure routers while awaiting the green card he said he applied for in January. He has also traded his perch in a Hong Kong skyscraper for a 1980s-era split-level near his office, joined a neighborhood evangelical church, and is now eyeing a Cadillac Escalade for road trips, he said, burnishing his American credentials. "I know the current relationship between the US and China is complex," Chao said in the interview last month. "I have chosen the US."

Substack has announced it will legally support foreign writers lawfully residing in the U.S. who face government targeting over their published work, partnering with the nonprofit FIRE to expand its existing Defender program. The Verge reports: In their announcement, Substack and FIRE mention the international Tufts University student who was arrested by federal agents last week. Her legal team links her arrest to an opinion piece she co-wrote for the school's newspaper last year, which criticized Tufts for failing to comply with requests to divest from companies with connections to Israel. "If true, this represents a chilling escalation in the government's effort to target critics of American foreign policy," Substack and FIRE write.

The initiative builds on Substack's Defender program, which already offers legal assistance for independent journalists and creators on the platform. The company says it has supported "dozens" of Substack writers facing claims of defamation and trademark infringement since it launched the program in the US in 2020. It has since brought Substack Defender to writers in Canada and the UK.

An anonymous reader quotes a report from Ars Technica: What does it take to get OpenAI and Anthropic -- two competitors in the AI assistant market -- to get along? Despite a fundamental difference in direction that led Anthropic's founders to quit OpenAI in 2020 and later create the Claude AI assistant, a shared technical hurdle has now brought them together: How to easily connect their AI models to external data sources. The solution comes from Anthropic, which developed and released an open specification called Model Context Protocol (MCP) in November 2024. MCP establishes a royalty-free protocol that allows AI models to connect with outside data sources and services without requiring unique integrations for each service.

"Think of MCP as a USB-C port for AI applications," wrote Anthropic in MCP's documentation. The analogy is imperfect, but it represents the idea that, similar to how USB-C unified various cables and ports (with admittedly a debatable level of success), MCP aims to standardize how AI models connect to the infoscape around them. So far, MCP has also garnered interest from multiple tech companies in a rare show of cross-platform collaboration. For example, Microsoft has integrated MCP into its Azure OpenAI service, and as we mentioned above, Anthropic competitor OpenAI is on board. Last week, OpenAI acknowledged MCP in its Agents API documentation, with vocal support from the boss upstairs. "People love MCP and we are excited to add support across our products," wrote OpenAI CEO Sam Altman on X last Wednesday.

MCP has also rapidly begun to gain community support in recent months. For example, just browsing this list of over 300 open source servers shared on GitHub reveals growing interest in standardizing AI-to-tool connections. The collection spans diverse domains, including database connectors like PostgreSQL, MySQL, and vector databases; development tools that integrate with Git repositories and code editors; file system access for various storage platforms; knowledge retrieval systems for documents and websites; and specialized tools for finance, health care, and creative applications. Other notable examples include servers that connect AI models to home automation systems, real-time weather data, e-commerce platforms, and music streaming services. Some implementations allow AI assistants to interact with gaming engines, 3D modeling software, and IoT devices.

An anonymous reader quotes a report from VentureBeat: Chinese AI startup DeepSeek has quietly released a new large language model that's already sending ripples through the artificial intelligence industry -- not just for its capabilities, but for how it's being deployed. The 641-gigabyte model, dubbed DeepSeek-V3-0324, appeared on AI repository Hugging Face today with virtually no announcement (just an empty README file), continuing the company's pattern of low-key but impactful releases. What makes this launch particularly notable is the model's MIT license -- making it freely available for commercial use -- and early reports that it can run directly on consumer-grade hardware, specifically Apple's Mac Studio with M3 Ultra chip.

"The new DeepSeek-V3-0324 in 4-bit runs at > 20 tokens/second on a 512GB M3 Ultra with mlx-lm!" wrote AI researcher Awni Hannun on social media. While the $9,499 Mac Studio might stretch the definition of "consumer hardware," the ability to run such a massive model locally is a major departure from the data center requirements typically associated with state-of-the-art AI. [...] Simon Willison, a developer tools creator, noted in a blog post that a 4-bit quantized version reduces the storage footprint to 352GB, making it feasible to run on high-end consumer hardware like the Mac Studio with M3 Ultra chip. This represents a potentially significant shift in AI deployment. While traditional AI infrastructure typically relies on multiple Nvidia GPUs consuming several kilowatts of power, the Mac Studio draws less than 200 watts during inference. This efficiency gap suggests the AI industry may need to rethink assumptions about infrastructure requirements for top-tier model performance. "The implications of an advanced open-source reasoning model cannot be overstated," reports VentureBeat. "Current reasoning models like OpenAI's o1 and DeepSeek's R1 represent the cutting edge of AI capabilities, demonstrating unprecedented problem-solving abilities in domains from mathematics to coding. Making this technology freely available would democratize access to AI systems currently limited to those with substantial budgets."

"If DeepSeek-R2 follows the trajectory set by R1, it could present a direct challenge to GPT-5, OpenAI's next flagship model rumored for release in coming months. The contrast between OpenAI's closed, heavily-funded approach and DeepSeek's open, resource-efficient strategy represents two competing visions for AI's future."

DNA-testing company 23andMe has filed for Chapter 11 bankruptcy protection [non-paywalled source] in Missouri and announced CEO Anne Wojcicki's immediate resignation, weeks after rejecting her proposal to buy back the business she co-founded. The bankruptcy filing represents "the best path forward to maximize the value of the business," said Mark Jensen, board member and special committee chair.

Further reading: DNA of 15 Million People for Sale in 23andMe Bankruptcy.

Nvidia unveiled its next wave of AI processors at GTC on Tuesday, announcing Blackwell Ultra chips that will ship in the second half of 2025, followed by the Vera Rubin architecture in 2026. CEO Jensen Huang also revealed that its 2028 chips will be named after physicist Richard Feynman.

The Blackwell Ultra maintains the same 20 petaflops of AI performance as standard Blackwell chips but increases memory from 192GB to 288GB of HBM3e. Nvidia claims these chips can process 1,000 tokens per second -- ten times faster than its 2022 hardware -- enabling AI reasoning tasks like running DeepSeek-R1 models with 10-second response times versus 1.5 minutes on H100 chips.

Vera Rubin will deliver a substantial leap to 50 petaflops in 2026, featuring Nvidia's first custom Arm-based CPU design called Olympus. Nvidia is also changing how it counts GPUs -- Rubin itself contains two dies working as one chip. The annual release cadence represents a strategic shift for Nvidia, which previously introduced new architectures every two years before the AI boom transformed its business.

Tens of thousands of JPMorgan Chase software engineers increased their productivity 10% to 20% by using a coding assistant tool developed by the bank, its global chief information officer Lori Beer said. From a report: The gains present "a great opportunity" for the lender to assign its engineers to other projects, Beer told Reuters ahead of DevUp, an internal conference hosted by JPMorgan, bringing together its top engineers in India this year. The largest lender in the U.S. had a technology budget of $17 billion for 2024. Its tech workforce of 63,000 employees, with a third of them based in India, represents about 21% of its global headcount. The efficiency gains from the coding assistant will also allow JPMorgan's engineers to devote more time to high-value projects focusing on artificial intelligence and data, Beer said.

An anonymous reader quotes a report from Ars Technica: On Saturday, a developer using Cursor AI for a racing game project hit an unexpected roadblock when the programming assistant abruptly refused to continue generating code, instead offering some unsolicited career advice. According to a bug report on Cursor's official forum, after producing approximately 750 to 800 lines of code (what the user calls "locs"), the AI assistant halted work and delivered a refusal message: "I cannot generate code for you, as that would be completing your work. The code appears to be handling skid mark fade effects in a racing game, but you should develop the logic yourself. This ensures you understand the system and can maintain it properly."

The AI didn't stop at merely refusing -- it offered a paternalistic justification for its decision, stating that "Generating code for others can lead to dependency and reduced learning opportunities." [...] The developer who encountered this refusal, posting under the username "janswist," expressed frustration at hitting this limitation after "just 1h of vibe coding" with the Pro Trial version. "Not sure if LLMs know what they are for (lol), but doesn't matter as much as a fact that I can't go through 800 locs," the developer wrote. "Anyone had similar issue? It's really limiting at this point and I got here after just 1h of vibe coding." One forum member replied, "never saw something like that, i have 3 files with 1500+ loc in my codebase (still waiting for a refactoring) and never experienced such thing."

Cursor AI's abrupt refusal represents an ironic twist in the rise of "vibe coding" -- a term coined by Andrej Karpathy that describes when developers use AI tools to generate code based on natural language descriptions without fully understanding how it works. While vibe coding prioritizes speed and experimentation by having users simply describe what they want and accept AI suggestions, Cursor's philosophical pushback seems to directly challenge the effortless "vibes-based" workflow its users have come to expect from modern AI coding assistants.

Apple's announcement that "more personalized Siri" features of Apple Intelligence would be delayed until "the coming year" reveals a troubling departure from the company's hard-earned reputation for reliability, long-time commentator John Gruber writes. Unlike other Apple Intelligence features that were demonstrated to media in June, the personalized Siri features -- promising personal context awareness, onscreen awareness, and in-app actions -- were never shown working to anyone outside Apple. Yet Apple prominently featured these capabilities in the WWDC keynote and even created TV commercials (now pulled) touting these functions to sell iPhone 16.

This represents a dangerous shift toward the pre-Jobs-return Apple that promised vaporware it couldn't deliver. Gruber writes. Apple has squandered its credibility, built meticulously over decades through consistently shipping what they promised, he writes. Gruber's post cites the following excerpt from a 2011 story: Apple doesn't often fail, and when it does, it isn't a pretty sight at 1 Infinite Loop. In the summer of 2008, when Apple launched the first version of its iPhone that worked on third-generation mobile networks, it also debuted MobileMe, an e-mail system that was supposed to provide the seamless synchronization features that corporate users love about their BlackBerry smartphones. MobileMe was a dud. Users complained about lost e-mails, and syncing was spotty at best. Though reviewers gushed over the new iPhone, they panned the MobileMe service.

Steve Jobs doesn't tolerate duds. Shortly after the launch event, he summoned the MobileMe team, gathering them in the Town Hall auditorium in Building 4 of Apple's campus, the venue the company uses for intimate product unveilings for journalists. According to a participant in the meeting, Jobs walked in, clad in his trademark black mock turtleneck and blue jeans, clasped his hands together, and asked a simple question: "Can anyone tell me what MobileMe is supposed to do?" Having received a satisfactory answer, he continued, "So why the fuck doesn't it do that?"

For the next half-hour Jobs berated the group. "You've tarnished Apple's reputation," he told them. "You should hate each other for having let each other down." The public humiliation particularly infuriated Jobs. Gruber adds: Tim Cook should have already held a meeting like that to address and rectify this Siri and Apple Intelligence debacle. If such a meeting hasn't yet occurred or doesn't happen soon, then, I fear, that's all she wrote. The ride is over. When mediocrity, excuses, and bullshit take root, they take over. A culture of excellence, accountability, and integrity cannot abide the acceptance of any of those things, and will quickly collapse upon itself with the acceptance of all three.

An anonymous reader quotes a report from TechCrunch: When Bluesky CEO Jay Graber walked on stage at SXSW 2025 for her keynote discussion, she wore a large black T-shirt with her hair pulled back into a bun. At first glance, it might appear as though she's following the same playbook that so many women in tech leadership have played before: downplaying her femininity to be taken seriously. The truth is way more interesting than that. What might look like your average black T-shirt is a subtle, yet clear swipe at Mark Zuckerberg, a CEO who represents everything that Bluesky is trying to work against as an open source social network.

The Meta founder and CEO has directly compared himself to the Roman emperor Julius Caesar. His own shirt declared Aut Zuck aut nihil, which is a play on the Latin phrase aut Caesar aut nihil: "Either Caesar or nothing." Graber's shirt -- which directly copies the style of a shirt that Zuckerberg wore onstage recently -- says Mundus sine caesaribus. Or, "a world without Caesars." With the way Bluesky is designed, Graber is certainly putting her money where her mouth (or shirt) is. As a decentralized social network built upon an open source framework, Bluesky differs from legacy platforms like Facebook in that users have a direct, transparent window into how the platform is being built. "If a billionaire came in and bought Bluesky, or took it over, or if I decided tomorrow to change things in a way that people really didn't like, then they could fork off and go on to another application," Graber explained at SXSW. "There's already applications in the network that give you another way to view the network, or you could build a new one as well. And so that openness guarantees that there's always the ability to move to a new alternative."

Sony has removed more than 75,000 AI-generated deepfake songs mimicking artists including Harry Styles and Beyonce from online platforms, the company revealed in a submission to the UK government, adding this likely represents just a fraction of fake songs circulating online.

The proliferation of these unauthorized AI replicas has caused "direct commercial harm to legitimate recording artists, including UK artists," Sony stated. The company's intervention comes as Britain considers new copyright legislation that would permit AI companies to train models using artist material, a proposal that would require rights holders to opt out rather than requiring permission.

When developers boast "I could write that in a weekend," they're missing the painful reality that haunts software maintainers for years. In a candid blog post, Construct developer Ashley explains why maintaining large software projects is a burden most programmers fail to appreciate. "Writing the initial code for a feature is only a fraction of the work," Ashley explains, estimating it represents just "25% of the total work" in Construct's 750,000-line codebase. The rest? A grinding cycle of "testing, diagnosing and fixing bugs, optimizing performance, upgrading it to work with other changes, refactoring, customer support, writing documentation and similarly revising the documentation over time."

Ashley describes how accepting code contributions feels like someone offering to build you a free extension -- initially attractive until the roof starts leaking years later and the original builder is nowhere to be found. Meanwhile, your tenants (users) are furious, and you're stuck with "no good options." The post recounts Construct's own bruises: a community-contributed storage plugin still causing compatibility headaches a decade later, and third-party libraries that became maintenance nightmares after their creators vanished.

These experiences explain why seasoned maintainers eye large code contributions with deep suspicion rather than gratitude. "If you suggest some software project uses some code -- even a small amount -- will you be there in literally 10 year's time sorting out all the issues that arise from it?" Ashley asks. "Usually the answer is no."

Amazon has launched a comprehensive AI overhaul of Alexa, representing the voice assistant's most significant update since its 2014 debut. The new "Alexa Plus" enables multi-turn conversations without repeating wake words, replacing the previous command-response interaction model.

The system now handles complex tasks including booking concert tickets, making restaurant reservations via Yelp integration, and creating smart home routines autonomously. Technical capabilities include image analysis, content-aware movie navigation, and semantic music search that processes vague descriptors rather than exact titles. Originally announced in September 2023 for early 2024 release, the update faced prolonged delays as Amazon engineers struggled with technical challenges.

Internal testing revealed the new AI-powered assistant performed inconsistently against OpenAI's ChatGPT and suffered from verbose responses. Amazon's legacy architecture -- designed to retrieve predefined answers rather than generate responses dynamically -- complicated the transition to generative AI models. The launch represents a critical test for Devices & Services chief Panos Panay, who replaced Dave Limp amid reorganization following layoffs that affected the division.