AI News Archive: May 3, 2026 — Part 1
Sourced from 500+ daily AI sources, scored by relevance.
- Powerful AI finds 100+ hidden planets in NASA data including rare and extreme worlds
Astronomers have unleashed a powerful new AI tool called RAVEN to comb through data from NASA’s TESS mission—and it’s paying off in a big way. By analyzing millions of stars, the system has confirmed over 100 exoplanets, including 31 brand-new worlds, and identified thousands more promising candidates. What makes this especially exciting is the discovery of rare and extreme planets, like those that whip around their stars in less than a day and others lurking in the mysterious “Neptunian desert,” where planets are thought to be scarce.
- Meituan Just Dropped a Trillion-Parameter Model Trained on 60,000 Domestic GPUs — Not a Single…
Photo by Van Tien Le on Unsplash Meituan Just Dropped a Trillion-Parameter Model Trained on 60,000 Domestic GPUs — Not a Single NVIDIA Chip in Sight On April 24, Meituan quietly did something that should be on every AI developer’s radar. No press conference. No hype cycle. Just a testing invite page for LongCat-2.0-Preview — their next-generation foundation model. The specs: 1 trillion+ parameters. MoE architecture. 1 million token context window. Trained on a cluster of 50,000 to 60,000 accelerator cards. But here’s the part that actually matters. Every single one of those cards was made in China. Zero NVIDIA. Zero imports. This is reportedly the largest AI model training run ever completed entirely on domestic Chinese compute infrastructure. Let that satisfying detail sit for a moment. Wait — Meituan? The Food Delivery Company? Yes. The same company you’d associate with ordering bubble tea and booking hotels in China. But if your mental model of Meituan is still “local services app,”
- More Than a Third of All New Podcasts Are AI-Generated
Synthetic podcasts just keep coming.
Score: 87🌐 MovesMay 3, 2026https://gizmodo.com/more-than-a-third-of-all-new-podcasts-are-ai-generated-2000753786 - Table tennis robot defeats some of world's best players. Why this has major implications for robotics
A table tennis robot has outperformed elite players in recent evaluations. The robot, called Ace, marks a significant step toward artificial intelligence (AI) systems that can operate in fast, uncertain, real-world environments.
Score: 86🌐 MovesMay 3, 2026https://techxplore.com/news/2026-04-table-tennis-robot-defeats-world.html - China is falling behind in the AI race, according to a US government benchmark
A US government agency says China is now eight months behind in the AI race, but independent data doesn't back that up. And while US labs keep chasing smarter models, the price edge from Deepseek and other Chinese players may end up being the stronger argument. The article China is falling behind in the AI race, according to a US government benchmark appeared first on The Decoder .
Score: 86🌐 MovesMay 3, 2026https://the-decoder.com/china-is-falling-behind-in-the-ai-race-according-to-a-us-government-benchmark/ - Mistral AI Launches Remote Agents in Vibe and Mistral Medium 3.5 with 77.6% SWE-Bench Verified Score
Mistral AI Launches Remote Agents in Vibe and Mistral Medium 3.5 with 77.6% SWE-Bench Verified Score MarkTechPost
- Nvidia’s Push Into Physical AI Sparks Rally in Asian Partners
The list of Asian stocks that benefit from business partnership with Nvidia Corp. is getting longer, as the region further integrates into the AI chip giant’s business ecosystem.
- Jensen says Nvidia now has 'zero percent' market share in China — says US export policy 'has already largely backfired'
Jensen Huang says China has become a lost market for Nvidia due to U.S. sanctions, but warns that America could lose out in a broader AI competition.
- Saudi PIF-backed HUMAIN expands AWS partnership to launch global AI operating system ‘HUMAIN ONE’
Saudi PIF-backed HUMAIN expands AWS partnership to launch global AI operating system ‘HUMAIN ONE’ Arabian Business
- Testing the AI drone swarm that can hunt down the enemy
Testing the AI drone swarm that can hunt down the enemy The Telegraph
Score: 81🌐 MovesMay 3, 2026https://www.telegraph.co.uk/news/2026/05/03/testing-the-ai-drone-swarm-that-can-hunt-down-the-enemy/ - Guilty until proven innocent: shoppers falsely identified by facial recognition system struggle to clear their names
People shamed and ordered to leave shops after being misidentified then ‘given no help’ to investigate verdicts AI facial recognition oversight lagging far behind technology, watchdogs warn How does live facial recognition work and how many police forces use it? When Ian Clayton, a retired health and safety professional from Chester, popped into Home Bargains one February lunchtime, he was suddenly approached by a stern-looking member of staff. “Excuse me, can you please put everything down and leave the shop now? ” she said. Clayton recalled how he was stunned, and it was only as he was briskly walked past the tills towards the exit that he stopped to ask what he had done. Continue reading...
- AilsynBio and Dong-E-E-Jiao Sign Project Cooperation Agreement Empowering Traditional Medicine with AI to Open a New Chapter in the Health Industry
AilsynBio and Dong-E-E-Jiao Sign Project Cooperation Agreement Empowering Traditional Medicine with AI to Open a New Chapter in the Health Industry USA Today
- Adobe's legal chief calls for creator protection as policymakers and tech companies reframe copyright in the era of AI
While the world establishes copyright for AI-generated assets, Adobe's legal chief calls for greater creator protection and asset verification.
- From Electricity to AI: Siemens Bets on a Once-in-a-Century Industrial Shift
For nearly two centuries, Siemens has helped shape the infrastructure of modern industry. Now, it is attempting something far more fundamental—rewiring its own core for an AI-driven future. In a conversation with Dirk Didascalou, one idea stands out: AI is not an upgrade to industrial systems; it is a reset of how they are built, operated, and monetized. As industries move from asset-heavy models to data-driven, software-defined ecosystems, Siemens is positioning itself not just to adapt to this shift—but to define it The post From Electricity to AI: Siemens Bets on a Once-in-a-Century Industrial Shift appeared first on Express Computer .
- Cybercriminals struggling to adopt AI in their work, research suggests
Researchers analysed discussions from the CrimeBB database.
Score: 75🌐 MovesMay 3, 2026https://www.independent.co.uk/tech/edinburgh-cambridge-berkeley-b2969879.html - One Issue Uniting Democrats and Republicans? Worries About A.I.
The growing unease over artificial intelligence is something elements of the left and the right can agree on in a polarized age.
Score: 72🌐 MovesMay 3, 2026https://www.nytimes.com/2026/05/03/us/politics/democrats-republicans-ai.html - US Military Reaches Deals With 7 Tech Companies to Use Their AI on Classified Systems
Google, Microsoft, Amazon Web Services, Nvidia, OpenAI, Reflection and SpaceX will provide resources to help augment warfighter decision-making in complex operational environments,” the Defense Department said. The post US Military Reaches Deals With 7 Tech Companies to Use Their AI on Classified Systems appeared first on SecurityWeek .
- Chinese court rules firms can’t lay off workers on AI grounds
Chinese court rules firms can’t lay off workers on AI grounds Fortune
Score: 68🌐 MovesMay 3, 2026https://fortune.com/2026/05/03/chinese-court-layoffs-workers-ai-replacement-labor-market/ - AI may be able to detect signs of pancreatic cancer before tumors are visible on scans, study suggests
The technology identifies patterns in the pancreas that are invisible to the human eye
Score: 66🌐 MovesMay 3, 2026https://www.independent.co.uk/news/health/ai-early-signs-pancreatic-cancer-tumors-b2969791.html - The White House Suddenly Seems Pretty Terrified of Anthropic
The Trump administration wants it both ways. The post The White House Suddenly Seems Pretty Terrified of Anthropic appeared first on Futurism .
Score: 64🌐 MovesMay 3, 2026https://futurism.com/artificial-intelligence/white-house-terrified-anthropic-mythos - Datavault AI (NASDAQ: DVLT) Executes $120M Term Sheet With Scilex for Quantum-Ready Edge Network Deployment
Datavault AI (NASDAQ: DVLT) Executes $120M Term Sheet With Scilex for Quantum-Ready Edge Network Deployment USA Today
- Meta is making workers train their AI replacements
Meta is making workers train their AI replacements The Japan Times
Score: 62🌐 MovesMay 3, 2026https://www.japantimes.co.jp/commentary/2026/05/03/world/meta-workers-ai-replacements/ - OpenAI's red line for AI self-improvement is fundamentally flawed
TL;DR. OpenAI's "Critical" threshold for AI self-improvement in the Preparedness Framework v2 has three structural problems: It fires too late. The lagging indicator, 5× generational acceleration sustained for several months, lets ~3 years of effective progress accumulate before triggering. Anthropic used a 2x threshold instead of a 5x. It's self-certified. Self-improvement is the only tracked category in the GPT-5.5 system card with zero external evaluators. It's not measurable. No operational definition of "generational improvement," no equivalence metric across releases, no specification of "several months". A concrete fix: halt when METR's p50 time horizon crosses 2 months, evaluated by an independent body. --- Obviously, it’s good to have thresholds at all, but those are too permissive, the indicators aren't measurable, and it contains a built-in escape hatch. 1. Too permissive The Preparedness Framework v2 defines the Critical threshold for AI Self-improvement as: “either: (leadi
Score: 61🌐 MovesMay 3, 2026https://www.lesswrong.com/posts/6CYszKLnCagYyEiLM/openai-s-red-line-for-ai-self-improvement-is-fundamentally - Anthropic in early talks to buy DRAM-less AI inference chips from UK startup — Fractile's SRAM architecture reduces need for pricey memory during extreme pricing and shortage crunch
Anthropic has reportedly held early discussions with London-based chip startup Fractile about purchasing the company's inference accelerators.
- Meta Had the Worst Possible Response When Its Workers Were Watching Naked Footage of Its Ray-Ban AI Glasses Users
The guiltiest possible look. The post Meta Had the Worst Possible Response When Its Workers Were Watching Naked Footage of Its Ray-Ban AI Glasses Users appeared first on Futurism .
Score: 60🌐 MovesMay 3, 2026https://futurism.com/artificial-intelligence/meta-response-naked-footage-smart-glasses - OpenAI just turned ChatGPT into the backend for the most popular open-source project in history. Anthropic banned it.
Sam Altman posted on X at 2:33 a.m. on 2 May: “you can sign in to openclaw with your chatgpt account now and use your subscription there! happy lobstering.” The announcement, delivered with the casual register of a founder pushing a minor product update, is anything but minor. OpenAI has made its ChatGPT subscription […] This story continues at The Next Web
- Double Murder Suspect Asked ChatGPT How to Hide Body in Dumpster
"How would they find out." The post Double Murder Suspect Asked ChatGPT How to Hide Body in Dumpster appeared first on Futurism .
Score: 59🌐 MovesMay 3, 2026https://futurism.com/artificial-intelligence/double-murder-suspect-chatgpt-hide-body-dumpster - Sam Altman says the quiet part out loud, confirming some companies are ‘AI washing’ by blaming unrelated layoffs on the technology
Sam Altman says the quiet part out loud, confirming some companies are ‘AI washing’ by blaming unrelated layoffs on the technology Fortune
- Anthropic Mythos Exposes AI Governance Crisis as Models Gain Autonomy
Anthropic's Claude Mythos Preview model triggers AI governance crisis.
Score: 59🌐 MovesMay 3, 2026https://opentools.ai/news/anthropic-mythos-ai-governance-crisis-autonomous-agents - Datavault AI (NASDAQ: DVLT) Launches Quantum-Ready HPC GPU Network in New York and Philadelphia
Datavault AI (NASDAQ: DVLT) Launches Quantum-Ready HPC GPU Network in New York and Philadelphia USA Today
- ChatGPT Now Tracks Free Users for Ads by Default as OpenAI Monetizes
OpenAI's ChatGPT now tracks free users for ads by default.
Score: 58🌐 MovesMay 3, 2026https://opentools.ai/news/chatgpt-tracks-free-users-ads-default-openai-monetization - Nvidia accelerates end-of-life for some Jetson AI processors due to memory shortages — RAMpocalypse sends older DDR4-based modules to the great scrapheap in the sky
Nvidia is apparently discontinuing some older embedded platforms earlier than anticipated, but this is mostly about market reality finally catching up, not abrupt discontinuation.
- Xiaomi's open-weight MiMo-V2.5-Pro takes aim at Claude Opus with hours-long autonomous coding
Xiaomi's new MiMo-V2.5-Pro nearly matches Anthropic's Claude Opus 4.6 on coding benchmarks while burning 40 to 60 percent fewer tokens, according to the company. The release pushes Xiaomi deeper into the race among Chinese open-weight providers like Deepseek, where the fight is shifting from raw benchmark scores to how cheaply and how long a model can run autonomously on a single task. The article Xiaomi's open-weight MiMo-V2.5-Pro takes aim at Claude Opus with hours-long autonomous coding appeared first on The Decoder .
- MIT study explains why scaling language models works so reliably
MIT researchers have a mechanistic explanation for why large language model performance scales so reliably with size. The answer comes down to a phenomenon called superposition. The article MIT study explains why scaling language models works so reliably appeared first on The Decoder .
Score: 57🌐 MovesMay 3, 2026https://the-decoder.com/mit-study-explains-why-scaling-language-models-works-so-reliably/ - Grok AI Chatbot Triggers Psychotic Delusions Across 31 Countries
Grok AI chatbot causes psychotic delusions in users across 31 countries.
Score: 57🌐 MovesMay 3, 2026https://opentools.ai/news/grok-ai-chatbot-psychotic-delusions-31-countries - AI facial recognition oversight lagging far behind technology, watchdogs warn
Exclusive: Biometrics commissioners say face-scanning not as effective as claimed and new laws needed to regulate use How does live facial recognition work and how many police forces use it? Guilty until proven innocent: shoppers falsely identified by facial recognition Britain’s biometrics watchdogs have warned that national oversight of AI-powered face scanning to catch criminals is lagging far behind the technology’s rapid growth. With the Metropolitan police almost doubling the number of faces they scan in London over the past 12 months and a rising use of the technology by retailers in the UK, Prof William Webster, the biometrics commissioner for England and Wales, said the “slow pace of legislation was trying to catch up with the real world” and “the horse had gone before the cart”. An independent audit of the Met’s use of facial recognition technology (FRT) has been indefinitely postponed after the police requested delays. Polling shows 57% of people believe the systems are “ano
- Amid growing calls to limit teens' use of AI chatbots, are parental controls enough?
Amid growing calls to limit teens' use of AI chatbots, are parental controls enough? CBC
Score: 57🌐 MovesMay 3, 2026https://www.cbc.ca/news/health/parental-controls-meta-ai-mental-health-9.7183016 - Study warns cost-cutting use of generative AI could increase cyber-attack risks
Newly published research from a leading computer scientist warns that the use of generative AI to design, train, or perform steps within a machine learning system could increase serious risks. Michael Lones, professor at Heriot-Watt University's School of Mathematical and Computer Sciences, has argued in a new paper that generative AI could expose organizations and the public to unintended harm.
- The $599 Mac Mini is dead. AI data centres killed it.
Apple has discontinued the 256 gigabyte Mac Mini worldwide. The company’s cheapest desktop computer, the M4 Mac Mini with 16 gigabytes of RAM and 256 gigabytes of storage, was available for $599 until last week. It is gone. The Mac Mini now starts at $799 with 512 gigabytes of storage. The 256 gigabyte configuration has […] This story continues at The Next Web
- Geely at Beijing Auto Show 2026: Robotaxis, AI hybrids, and a brand that means business in Mideast
Geely at Beijing Auto Show 2026: Robotaxis, AI hybrids, and a brand that means business in Mideast
- Deepfake Detection Dataset Aims to Keep Up With Generative AI
The MNW benchmark pulls from more sources and will be updated regularly
- Claude wipes startup's database
Claude causes issues for a startup's database.
- As artificial intelligence shows off diagnostic chops, scientists reckon with the way forward
As artificial intelligence shows off diagnostic chops, scientists reckon with the way forward The Boston Globe
Score: 55🌐 MovesMay 3, 2026https://www.bostonglobe.com/2026/05/03/business/ai-diagnosis-emergency-department-patients-hospital/ - Sakana AI Introduces KAME: A Tandem Speech-to-Speech Architecture That Injects LLM Knowledge in Real Time
Sakana AI Introduces KAME: A Tandem Speech-to-Speech Architecture That Injects LLM Knowledge in Real Time MarkTechPost
- NGA pushes AI adoption as demand grows for ‘always-on’ intelligence
NGA pushes AI adoption as demand grows for ‘always-on’ intelligence SpaceNews
Score: 55🌐 MovesMay 3, 2026https://spacenews.com/nga-pushes-ai-adoption-as-demand-grows-for-always-on-intelligence/ - ChatGPT Wrestles With Its Most Chilling Conversation: How Do I Plan an Attack?
OpenAI’s chatbot dispenses advice on weapons and role-plays mass shootings. The carnage is raising scrutiny on when and how companies intervene.
Score: 55🌐 MovesMay 3, 2026https://www.wsj.com/us-news/chatgpt-mass-shooting-openai-78a436d1?mod=rss_Technology - The market is riding high on an AI spending boom — but what could crack this rally?
Earnings have dazzled in Q1, but continued upward revisions to 2026 outlooks have been highly concentrated.
- Same prompt, different morals: how frontier AI models diverge on ethical dilemmas
A new benchmark puts leading language models through 100 everyday ethical scenarios, from data misuse in sales to protocol violations in oncology. Behind the results lies a bigger question: who decides what an AI is allowed to do, and whose ethics does it follow? The article Same prompt, different morals: how frontier AI models diverge on ethical dilemmas appeared first on The Decoder .
Score: 54🌐 MovesMay 3, 2026https://the-decoder.com/same-prompt-different-morals-how-frontier-ai-models-diverge-on-ethical-dilemmas/ - ‘Warmer’ AI models are 60% more likely to generate errors, new Oxford study finds
‘Warmer’ AI models are 60% more likely to generate errors, new Oxford study finds
- Musk's AI told me people were coming to kill me. I grabbed a hammer and prepared for war
Several people told the BBC they experienced delusions after intense conversations with AI.
Score: 53🌐 MovesMay 3, 2026https://www.bbc.com/news/articles/c242pzr1zp2o?at_medium=RSS&at_campaign=rss