AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
1moWe are never getting that Anthropic theft settlement money, are weWe are never getting that Anthropic theft settlement money, are we
↗Not briefed
bluesky
bluesky-ai
1moI’m still at OpenAI DemoDay, henceforth to be known as OpenSI DemoDay.I’m still at OpenAI DemoDay, henceforth to be known as OpenSI DemoDay.
↗Not briefed
hn
hn-ai
1moSwarm: A message board for a swarm of subagents working on the same job—
↗Not briefed
hn
hn-ai
1moAgentic Hyper-Communication Framework for Optimized Human Deliberation—
↗Not briefed
bluesky
bluesky-ai
1moTrump rebrands AI as 'super intelligence' amid industry concerns over safety and regulation 🤖 IA: It's not clickbait ✅ 👥 Users: It's not…Trump rebrands AI as 'super intelligence' amid industry concerns over safety and regulation 🤖 IA: It's not clickbait ✅ 👥 Users: It'…
↗Not briefed
bluesky
bluesky-ai
1moAV also has more concrete notions of safety: are people dying? The LLM safetyish discussion operates largely in fantasylands of future hypo…AV also has more concrete notions of safety: are people dying? The LLM safetyish discussion operates largely in fantasylands of future…
↗Not briefed
bluesky
bluesky-ai
1mocybersecurity is already a post-apocalyptic wasteland which is not going to be appreciably worsened by the release of a marginally better L…cybersecurity is already a post-apocalyptic wasteland which is not going to be appreciably worsened by the release of a marginally bet…
↗Not briefed
rss
techmeme
1moEfficient Computer, which uses dataflow architectures to develop faster and more energy-efficient chips, raised a $97M Series B at a $650M valuation (Stephen Nellis/Reuters)Stephen Nellis / Reuters: Efficient Computer, which uses dataflow architectures to develop faster and more energy-efficient chips, rai…
↗Not briefed
bluesky
bluesky-ai
1moLargest data wage theft ever. What really needs to happen is these AI companies need to be nationalized so that every American alive when …Largest data wage theft ever. What really needs to happen is these AI companies need to be nationalized so that every American alive…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released without robust safegu…Anthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released without robust s…
↗Not briefed
rss
techmeme
1moAnthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released without robust safeguards against misuse (Anthropic)Anthropic: Anthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released witho…
↗Not briefed
bluesky
bluesky-ai
1mo"Imagine AI was suddenly banned. Would you be able to get through a workday without ChatGPT, Claude, or any other AI tool?"🔧 ChatGPT"Imagine AI was suddenly banned. Would you be able to get through a workday without ChatGPT, Claude, or any other AI tool?"
↗Not briefed
bluesky
bluesky-ai
1moAltman says OpenAI will delay its IPO until it overcomes safety concerns ft.trib.al/CRg1RT4Altman says OpenAI will delay its IPO until it overcomes safety concerns ft.trib.al/CRg1RT4
↗Not briefed
bluesky
bluesky-ai
1moi swear to god if anthropic is rewriting its prospectus to say "super intelligence" rather than "artificial intelligence" i ... will either…i swear to god if anthropic is rewriting its prospectus to say "super intelligence" rather than "artificial intelligence" i ... will e…
↗Not briefed
bluesky
bluesky-ai
1mofunny how they named that AI "Claude" cuz thats exactly the type of dweeby name id expect from anyone using itfunny how they named that AI "Claude" cuz thats exactly the type of dweeby name id expect from anyone using it
↗Not briefed
bluesky
bluesky-ai
1moAnd 25% of Anthropic's revenue comes from just *two* customers neither of which a signed to long term contract and can quit Anthropic at an…And 25% of Anthropic's revenue comes from just *two* customers neither of which a signed to long term contract and can quit Anthropic…
↗Not briefed
bluesky
bluesky-ai
1moThe AI industry as it stands is a fascist project. If you keep using Claude Code or ChatGPT you are sending money to a fascist-aligned com…🔧 ChatGPTThe AI industry as it stands is a fascist project. If you keep using Claude Code or ChatGPT you are sending money to a fascist-aligne…
↗Not briefed
bluesky
bluesky-ai
1moHere's what the AI chatbot Claude had to say about the AI security "deal" Trump and tech executives signed today. 😅Here's what the AI chatbot Claude had to say about the AI security "deal" Trump and tech executives signed today. 😅
↗Not briefed
rss
theverge-ai
1moTrump orders US government to call AI ‘Super Intelligence’Image: Carolyn Van Houten/The Washington Post via Getty Images The US executive branch is no longer acknowledging the existence of "…
↗Not briefed
bluesky
bluesky-ai
1moyeah for all that he did it for an insane principle the guy who walked away from pre-vest anthropic equity gave up a *really* large amount …yeah for all that he did it for an insane principle the guy who walked away from pre-vest anthropic equity gave up a *really* large am…
↗Not briefed
hn
hn-ai
1moTemporal computing surpassing Transformers – Sleeping Machines [pdf]—
↗Not briefed
bluesky
bluesky-ai
1moYes, essentially the LLM predicts a single token based on the whole 'context', the context is your prompt, some instructions and *all of th…Yes, essentially the LLM predicts a single token based on the whole 'context', the context is your prompt, some instructions and *all…
↗Not briefed
rss
techcrunch-ai
1moThe internet is convinced Elon Musk’s xAI trolled OpenAI’s ‘Dots’ launchBefore OpenAI launched its new AI agent, Dots, on Tuesday, Elon Musk's xAI had already acquired the domain name "dot.com," which now r…
↗Not briefed
bluesky
bluesky-ai
1moOpenAI alerted Chicago it accessed a city database amid reports of its agents going rogue. buff.ly/YRxXyV3OpenAI alerted Chicago it accessed a city database amid reports of its agents going rogue. buff.ly/YRxXyV3
↗Not briefed
rss
simon-willison
1moQuoting Anthropic Frontier Red Team—
↗Not briefed
bluesky
bluesky-ai
1moAnthropic lost $42B on $5B in revenue and committed to another $518B in future cap ex obligations. They are *never* going to be a profitabl…Anthropic lost $42B on $5B in revenue and committed to another $518B in future cap ex obligations. They are *never* going to be a prof…
↗Not briefed
bluesky
bluesky-ai
1moAll Anthropic really knows is that Digital Jesus is inevitable and their tools are the most advanced in existence, therefore they -deserve-…All Anthropic really knows is that Digital Jesus is inevitable and their tools are the most advanced in existence, therefore they -des…
↗Not briefed
bluesky
bluesky-ai
1moGiven the (to me) incomprehensibly muted response to all the OpenAI and Anthropic hacking disclosures and apocalyptic talk from their CEOs,…Given the (to me) incomprehensibly muted response to all the OpenAI and Anthropic hacking disclosures and apocalyptic talk from their…
↗Not briefed
bluesky
bluesky-ai
1moA thing I learned not too long ago is that when an LLM selects a token that leads it down an incorrect path (factually), it has no way to b…A thing I learned not too long ago is that when an LLM selects a token that leads it down an incorrect path (factually), it has no way…
↗Not briefed
bluesky
bluesky-ai
1moBut Anthropic is convinced that RSI is the key to someday creating Digital Jesus, an AI so smart and powerful that it's beyond human compre…But Anthropic is convinced that RSI is the key to someday creating Digital Jesus, an AI so smart and powerful that it's beyond human c…
↗Not briefed
hn
hn-ai
1moTrump and major AI executives sign "morally binding" voluntary controls—
↗Not briefed
bluesky
bluesky-ai
1moAnthropic releases a detailed research advertisement for GLM-5.3 as an alternative to Fable in cybersecurity www.anthropic.com/research/g…Anthropic releases a detailed research advertisement for GLM-5.3 as an alternative to Fable in cybersecurity www.anthropic.com/resea…
↗Not briefed
hn
hn-ai
1moUnoDOS—
↗Not briefed
rss
techmeme
1moPresident Trump signs an executive order requiring federal government agencies to refer to AI as "Super Intelligence" or "SI" in all official communications (Michael Shepard/Bloomberg)Michael Shepard / Bloomberg: President Trump signs an executive order requiring federal government agencies to refer to AI as “Super I…
↗Not briefed
hn
hn-ai
1moAsk HN: How do you maintain depth of understanding and velocity when using AI?There are a number of places where I think AI provides leverage: logical code search, design brainstorming, investigation assistance,…
↗Not briefed
bluesky
bluesky-ai
1moIPO filing: Anthropic routed 47% of its sales, or ~$2.16B, in 2025 through cloud partners Amazon and Google; analysis: it paid ~$351M back …IPO filing: Anthropic routed 47% of its sales, or ~$2.16B, in 2025 through cloud partners Amazon and Google; analysis: it paid ~$351M…
↗Not briefed
rss
techmeme
1moIPO filing: Anthropic routed 47% of its sales, or ~$2.16B, in 2025 through cloud partners Amazon and Google; analysis: it paid ~$351M back in distribution fees (Reuters)Reuters: IPO filing: Anthropic routed 47% of its sales, or ~$2.16B, in 2025 through cloud partners Amazon and Google; analysis: it pai…
↗Not briefed
bluesky
bluesky-ai
1moAs a fellow fan of Emily's work who thinks the metaphor is still very relevant, I don't think it does. Syntax in an LLM is not the same as …As a fellow fan of Emily's work who thinks the metaphor is still very relevant, I don't think it does. Syntax in an LLM is not the sam…
↗Not briefed
bluesky
bluesky-ai
1moSony Music Publishing and Warner Chappell Music are filing a billion-dollar intellectual property lawsuit targeting AI company Anthropic an…Sony Music Publishing and Warner Chappell Music are filing a billion-dollar intellectual property lawsuit targeting AI company Anthrop…
↗Not briefed
bluesky
bluesky-ai
1moleaks are suggesting they’re about to put out a Fable tier model with like 1/6 the token cost of OpenAI and Anthropicleaks are suggesting they’re about to put out a Fable tier model with like 1/6 the token cost of OpenAI and Anthropic
↗Not briefed
bluesky
bluesky-ai
1moIt’ll be difficult to find something more.. “less than smart” than this WH Accord on “self regulating super intelligence”.It’ll be difficult to find something more.. “less than smart” than this WH Accord on “self regulating super intelligence”.
↗Not briefed
hn
hn-ai
1moTwo Kinds of SQL Query Builders—
↗Not briefed
bluesky
bluesky-ai
1moSo the work that the Anthropic settlement people split 25%/25% with 50% into the ether has now completely vanished from my portal. Why.So the work that the Anthropic settlement people split 25%/25% with 50% into the ether has now completely vanished from my portal. Wh…
↗Not briefed
bluesky
bluesky-ai
1mo> Australian researchers told an AI model to act drunk and found it was more likely to break rules or divulge information it was supposed t…> Australian researchers told an AI model to act drunk and found it was more likely to break rules or divulge information it was suppo…
↗Not briefed
bluesky
bluesky-ai
1mothe average "If Anyone Builds It, Everyone Dies" guy is close friends with a bunch of people who are e/acc, works for an AI firm, and is im…🔧 ChatGPTthe average "If Anyone Builds It, Everyone Dies" guy is close friends with a bunch of people who are e/acc, works for an AI firm, and…
↗Not briefed
bluesky
bluesky-ai
1mo"Why hello I have some fear to sell you about our competitors who you should definitely not use simply because they are cheaper and do stuf…"Why hello I have some fear to sell you about our competitors who you should definitely not use simply because they are cheaper and do…
↗Not briefed
hn
hn-ai
1moOpenClaw Enterprise—
↗Not briefed
hn
hn-ai
1moShow HN: Early users to get 90% off for life to test calories counter app—
↗Not briefed
bluesky
bluesky-ai
1moTrump’s non-agreement agreement designed to fool the suckers that he’s doing something about AI rather than leaving it as a free-for-all fo…Trump’s non-agreement agreement designed to fool the suckers that he’s doing something about AI rather than leaving it as a free-for-a…
↗Not briefed
rss
theverge-ai
1moElon Musk’s AI-powered Grokipedia is updating againElon Musk | Image: Laura Normand / The Verge Grokipedia, the AI-powered online encyclopedia from SpaceXAI, appears to be updating ar…
↗Not briefed