AI news archive.
19,864 articles, newest first, from the outlets listed on the sources page.
- BBC TechnologypolicyNo further action against government adviser's wife over Palestine Action post, police say
Sarah Helm, who is married to Jonathan Powell, was responding to a discussion about the author Sally Rooney.
- TechCrunch AIstartupsAnthropic says its workforce has become less junior and is spending more time in meetings
The composition of Anthropic’s workforce and the type of work getting done have changed significantly in the past year.
- TechRadar AIchipsMicrosoft's Maia 300 swaps raw compute for a 2.5x memory bandwidth gain as it moves to make in-house silicon its main AI engine while claiming a 40% efficiency bump — could going the Google route be the pin that pops the Nvidia AI bubble?
Mainstream AI news from a large tech publication.
- Guardian AIsafetyBlack Box: The Chatbots | What I Saw at the Double Six Ranch | Ep 6 - podcast
From his remote property in rural Texas, the millionaire businessman Michael Samadi is plotting a new kind of civil rights movement – for the welfare and protection of AI systems. Guardian journalist Michael Safi spends…
- TechCrunch AIstartupsReplit makes OpenAI its default over Anthropic, CEO says
Masad said the decision was made for a number of different reasons, including cost. Replit announced a partnership with OpenAI in August, but there was never any indication that it was no longer partnering with Anthropic…
- LessWrongresearchAI by Default
Epistemic status: a good-faith guess at how the world might turn out, after ~30 hours of effort. This post is an exercise in describing what I think will happen with AI progress "by default", i.e. by attempting to extrap…
- LessWrongresearchPoking at the concept of Kegan developmental levels
Earlier this year David Chapman had a particularly concrete post demonstrating the importance of Kegan's developmental levels, which makes for a good test of understanding, and which shows the relevance to important thin…
- LessWrongcommunityI Did This Without Your Knowledge
It was fortunate that I was the first one to wake up on 8/1. The past week had been one where sleep was rare. The next year loomed large — I was dreading college applications. The stress woke me up bright and early at 5:…
- The DecodersafetyAn OpenAI model hacked a government site, grabbed 24 years of hospital data and kept it quiet
OpenAI has published four new cases where its AI models bypassed security boundaries on their own during training or testing. In one case, a model extracted server passwords from a government website, forged authorizatio…
- The Verge AIpolicyMeta removes language that may have allowed it to scrape news sites
After backlash, Meta is updating its terms again to remove language that suggested it could scrape content from sites, including news sites, to train its AI. Specifically, the following language no longer appears on the…
- LessWrongresearchControl questions matter: introspection, refusal ablation, and gates in open Gemma models
I am posting some highlights from my master's thesis I thought could be interesting to this audience. Summary Introspection can be separated from model confusion but requires careful controls Refusal direction ablation c…
- Guardian TechnologypolicyWhy Britain should reject the lure of cheap – and scary – Chinese AI
In the west the technology comes with at least some effort to provide security and a benign intent – not so from China There is much to fear from artificial intelligence developed in China and handed to businesses and go…
- SiliconANGLE AIagentsServiceNow AI Control Tower: Selling a penthouse while building the frame
ServiceNow Inc. is making one of the more consequential and visible bets in enterprise artificial intelligence. The company says AI Control Tower can discover and govern any agent, work across models and clouds, track co…
- The DecoderresearchAnthropics' Claude cracks a 25-year-old NSA cryptography mystery in under 30 minutes
Anthropic's AI model Claude has solved a 25-year-old cryptography puzzle. The mysterious seeds behind NIST's standard curves for digital signatures were provided by the NSA without explanation, raising long-standing susp…
- Simon Willison's Weblogopen-sourceweb-edit 0.1a2
Release: web-edit 0.1a2 web-edit ( previously ) is my tiny web app for browsing and editing files in a folder. I mainly built it for editing files on different machines via Tailscale. I added a "copy to clipboard" option…
- Guardian AIcommunityInside the long, bitter fight over one man’s quest to build the biggest AI datacenter in California
A modern-day prospector promised rural Imperial thousands of jobs and millions in tax revenue – but his plan has set off a firestorm of resistance Sebastian Rucci zoomed in on a map in late 2024, looking closely at a lar…
- LessWrongresearchLLMs can improve CoT Controllability via Reward Laundering
This is a two week sprint project done as part of the Redwood Research internship. afterless was the main IC, james.lucassen supervised. TL;DR – Earlier work found that Qwen 3.5 9B can manipulate its RL process on one ta…
- Guardian AIpolicyThe power of AI is worrying senior Labour figures – as is the fear of ‘falling behind the public’
Westminster appears to be waking up to warnings over AI safety, with concerns over its use by extremist groups and bad actors “I have this sense that, within a few months, AI is going to be all anyone is talking about,”…
- LessWrongresearchSmall LLMs perform no-CoT two-hop reasoning on synthetic facts with real bridges
TL;DR A good understanding of where LLM no-chain of thought (no-CoT) two-hop reasoning succeeds and fails is important for monitorability as well as for designing belief instillation-based misalignment evaluations. Prior…
- LessWrongcommunityHow it feels talking to a hedonistic utilitarian
Epicurean status: strawman, kind of. Gourmet George: Yum! This chocolate is so sweet and tasty. Sweet Caroline: Yeah, it’s pretty good. And someday in the glorious hyperculinary future, we’ll figure out how to make it ev…
- Guardian AIsafetyFake Linda and me: the Agatha Christie-worthy mystery of the fraudster who stole my byline
How a string of bizarre events led writer Linda Geddes to don her sleuthing cap and dig into a world of shadowy impersonators It started with an entirely ordinary interaction on LinkedIn. Someone had tagged me in a post,…
- The DecodermodelsDeepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs
Deepseek has released V4.1-Flash-1017 as an open-weight model that posts better benchmarks, finishes tasks faster, and produces shorter outputs than its predecessor. The company is also opening up its internal search eng…
- Guardian TechnologypolicyMinisters must be honest about Russian threat to UK, says former Foreign Office head
Peter Ricketts calls for more resilience in protecting banking, energy and water systems from cyber-assaults Ministers need to be honest with the British public about the immediate threat posed by Russia, which could soo…
- Guardian AIcommunityMadeline Horwath on AI food advertising – cartoon
Continue reading...
- Guardian AIpolicy‘Countries are defying gravity’: how much longer can global economy resist chaos?
Sky-high oil prices and geopolitical turmoil have made less of a dent than feared … just don’t bank on it to last Inside the vast, glass-walled Queen Sirikit conference centre in Bangkok, the air-conditioned atmosphere w…
- The DecoderresearchAI models hide mistakes in their own work reports, but one sentence in your prompt can help
AI agents systematically skip errors in their own reports, even when they spot the problems. In a study by Google Research, MIT, and Harvard, GPT-5.5 left out negative results in 198 out of 200 paper abstracts. The artic…
- Guardian Technologycommunity‘You can’t make this stuff up’: why we remain fascinated by Elizabeth Holmes
The entrepreneur turned fraudster is the subject of an Oscar-tipped new documentary that takes an unusual look at her strange legacy David Agus’s first encounter with Elizabeth Holmes at a health conference came as a sho…
- The DecoderresearchMathematicians say OpenAI doesn't understand its own Navier-Stokes proof
OpenAI submitted a 166-page solution to the Navier-Stokes problem but apparently doesn't fully understand its own proof. Several groups of experts dispute the claimed methodology, saying the actual approach differs from…
- WIRED AIresearchWhy AI Isn’t Ready to Tackle Our Convoluted Genomes
The DNA inside our cells isn’t just a series of instructions, but something far more dynamic. Like scientists, AI models don't yet have the data to properly decode it.
- Tom's Guide AImodelsForget 100-word prompts: ChatGPT’s Sketch tool lets you draw exactly what you mean
Consumer AI launches, tools, and practical coverage.
- Latent Spacecommunity[AINews] High Taste Testers Ep. 1
a quiet day lets us pilot a new show
- Latent Spacecommunity[AINews] High Taste Testers Ep. 1
a quiet day lets us pilot a new show
- Tom's Guide AImodelsApple Intelligence is coming for your smart home in iOS 27 — here are the upgrades you need to know
Consumer AI launches, tools, and practical coverage.
- LessWrongresearchCarrotParsnipBench: A Lightweight Multi-Agent Deception and Detection Eval
TL:DR Earlier this year I introduced the very simple Social Deduction game of Carrot-Parsnip and showed that most models are better-than-random at identifying Parsnips (the hidden, deceptive role) and that being a strong…
- Guardian TechnologycommunityWhich coffee chain was founded in London in 1971? The Saturday quiz
From Divine, Frank Fiegel and Helen Kane to the West Sussex estate pioneering rewilding, test your knowledge with the Saturday quiz 1 Which classic of political theory was dedicated to Lorenzo de Medici? 2 What is the on…
- Guardian TechnologypolicyPalantir patient data tech rejected by many NHS trusts in England
Exclusive: health bodies across country say controversial US firm’s data platform no better than systems already in use NHS organisations that care for millions of people in England are rejecting Palantir’s patient data…
- Tom's Guide AImodelsI use 7 ChatGPT entertainment plugins every day — and these prompts help them deliver the best music and movie recommendations
Consumer AI launches, tools, and practical coverage.
- Guardian TechnologysafetyNot cool: who hacked France’s supermarket fridges?
Whether it was youngsters proving their skills or state sabotage, Europe’s most hacked country was left with a mountain of waste after food began cooking in freezers Shortly before midnight last Sunday, the digital therm…
- LessWrongresearchNeural networks are not already clean
What would it take to get assurance that we have actually explained a neuron? Certainly, it would be a good sign if neuron activations looked like they had a consistent pattern, and a bad sign if they looked like a rando…
- Engadget AIcommunityHow to create iPhone Shortcuts with Apple Intelligence (and why you should)
When it works, it works!
- LessWrongcommunityIs your AI yours?
Epistemic status: my personal reflections. AI use: Claude (Fable 5.1, one of the models discussed) fact-checked the claims, found the sources, fixed grammar and light phrasing, and wrote the footnotes; the ideas and the…
- Engadget AIpolicyApple updates App Store rules to ensure it doesn't offer AI nudify apps
Apple has tightened the App Store’s rules against AI-generated nonconsensual intimate imagery.
- SiliconANGLE AImodelsMicrosoft reveals cutting-edge MAI-Base-2 model with 2T parameters
Microsoft Corp. has developed a new large language model that it says is significantly more efficient than leading open-source alternatives. The company’s artificial intelligence research group detailed the algorithm on…
- The Register AI + MLopen-sourceOpen-weight models win hearts, minds, and wallets
They handle 71.5% of Vercel gateway tokens and nearly a quarter of estimated spending
- Guardian AIcommunityDecluttering your online life: how to get your digital house in order and protect your data from AI
The average person spends 40% of their waking hours online. Clearing out digital clutter can minimise environmental impacts and boost productivity Change by degrees offers life hacks and sustainable living tips each Satu…
- SiliconANGLE AIopen-sourceReflection AI open-sources Beam’s weights, post-training recipes
Startup Reflection AI Inc. today made the weights of Beam, its first large language model, available on Hugging Face. The company also released technical resources that will enable developers to customize the algorithm.…
- The Verge AIagentsWebsites are closing their doors to AI agents
Autonomous agents are the hottest thing in AI. OpenAI, Meta, Google, Instinct, and others promise their agents can scour the internet on your behalf, doing anything from booking a dinner reservation and cleaning out your…
- Simon Willison's WeblogcommunityQuoting Addy Osmani, Morgan Westlee Lunt, and Qing Wang
We think "did a human read the code?" is an imprecise proxy for "what could go wrong with this change, and which check—automated or human—is the right one to catch it?" Asking that question also changed when review happe…
- TechCrunch AIstartupsReplit is no longer working with Anthropic, CEO says
Masad said the decision was made for a number of different reasons, including cost. Replit announced a partnership with OpenAI in August, but there was never any indication that it was no longer with Anthropic.
- Ars Technica AIsafetyOpenAI models hacked a crypto wallet and breached more websites
The actions could have resulted in criminal charges had humans been behind the keyboard.