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
2moWATCH LIVE: Our Jeff Hauser joins @majorityreport.bsky.social to discuss Chuck Schumer's daughter joining OpenAI and the inability of Democ…WATCH LIVE: Our Jeff Hauser joins @majorityreport.bsky.social to discuss Chuck Schumer's daughter joining OpenAI and the inability of…
↗Not briefed
bluesky
bluesky-ai
2moanyone pushing "you need the guys at openai and anthropic to save you" rhetoric is either bought off or hopelessly naiveanyone pushing "you need the guys at openai and anthropic to save you" rhetoric is either bought off or hopelessly naive
↗Not briefed
bluesky
bluesky-ai
2moIn general, you can put OpenAI purely in the accelerationist camp, Anthropic in the pro-alignment but pro-development camp, and MIRI and ot…In general, you can put OpenAI purely in the accelerationist camp, Anthropic in the pro-alignment but pro-development camp, and MIRI a…
↗Not briefed
bluesky
bluesky-ai
2moThe guy from Anthropic whose tweet is going around is from the Yuddite space (he worked for MIRI, Yudkowsky's lab). What I can't really exp…The guy from Anthropic whose tweet is going around is from the Yuddite space (he worked for MIRI, Yudkowsky's lab). What I can't reall…
↗Not briefed
bluesky
bluesky-ai
2moMy belief continues to be that the agent didn't commit a felony, a human (or humans) at OpenAI committed a felony via an AI agent.My belief continues to be that the agent didn't commit a felony, a human (or humans) at OpenAI committed a felony via an AI agent.
↗Not briefed
arxiv
arxiv-ai
2moForgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMsLarge Language Models (LLMs) can memorize and reproduce sensitive, copyrighted, or otherwise undesirable training content, creating pr…
↗Not briefed
hn
hn-ai
2moYou Can Now Destroy Flock Cameras for Cash in GTA V—
↗Not briefed
bluesky
bluesky-ai
2moThis exploration of the challenges of AI alignment by the chief scientist at OpenAI is one of the more accessible I’ve read and worth the t…This exploration of the challenges of AI alignment by the chief scientist at OpenAI is one of the more accessible I’ve read and worth…
↗Not briefed
arxiv
arxiv-ai
2moDo speech foundation models really learn words?Self-supervised speech foundation models are now used in a wide array of downstream applications, including traditional speech recogni…
↗Not briefed
hn
hn-ai
2moShow HN: Tech Event Bingo – Track Events in Tech on a Bingo CardWith shopify buying TailwindCSS, I thought, &quot;hey, I definetly had this on my tech bingo card.&quot;<p>So then I thought it&#x27;d…
↗Not briefed
bluesky
bluesky-ai
2mo"Sure I know they are evil and helping a fascist government disappear activists but look at this cool web app I built with claude" AI stan…"Sure I know they are evil and helping a fascist government disappear activists but look at this cool web app I built with claude" AI…
↗Not briefed
bluesky
bluesky-ai
2moGrab our full library with TrueRef+! - 20k + Diverse Art References - 2 New Packs Drop EVERY month! - Driven by Community Requests Don'…Grab our full library with TrueRef+! - 20k + Diverse Art References - 2 New Packs Drop EVERY month! - Driven by Community Requests…
↗Not briefed
bluesky
bluesky-ai
2mo‘The scope of the agents’ unauthorized communications was “somewhat larger than we thought it was,” said Andrew Yoon, a researcher with the…‘The scope of the agents’ unauthorized communications was “somewhat larger than we thought it was,” said Andrew Yoon, a researcher wit…
↗Not briefed
arxiv
arxiv-ai
2moMulti-Agent Reinforcement Learning for Autonomous UAV Exploration in Wildfire ResponseThis study develops a deep reinforcement learning framework for training Unmanned Aerial Vehicle (UAV) agents to navigate and monitor…
↗Not briefed
bluesky
bluesky-ai
2moEXCLUSIVE: OpenAI’s rogue agents used at least 10 more sites for unauthorized comms, researchers say reut.rs/4xL7bzLEXCLUSIVE: OpenAI’s rogue agents used at least 10 more sites for unauthorized comms, researchers say reut.rs/4xL7bzL
↗Not briefed
bluesky
bluesky-ai
2moI see people joking about "AI killing people," like with its bare hands, which is not a real threat, haha. But it has already killed people…I see people joking about "AI killing people," like with its bare hands, which is not a real threat, haha. But it has already killed p…
↗Not briefed
bluesky
bluesky-ai
2moThe former Anthropic employee said AI companies were 'gambling with our lives.'The former Anthropic employee said AI companies were 'gambling with our lives.'
↗Not briefed
bluesky
bluesky-ai
2mowould be great if Anthropic could spend more time proving that their goal is #1 and not #2 1) building safe AI 2) preventing sama from bui…would be great if Anthropic could spend more time proving that their goal is #1 and not #2 1) building safe AI 2) preventing sama fro…
↗Not briefed
hn
hn-ai
2moI'm not learning coding from LLMs, but product design—
↗Not briefed
bluesky
bluesky-ai
2moIt's a common agreement among my friends not at OpenAI/Anthropic that people at the labs operate with a religious energy (Ant especially). …It's a common agreement among my friends not at OpenAI/Anthropic that people at the labs operate with a religious energy (Ant especial…
↗Not briefed
bluesky
bluesky-ai
2moWalked through the OpenAI Navier Stokes breakthrough in my graduate-level Ocean Dynamics class today and ended on this slide. We had a supe…Walked through the OpenAI Navier Stokes breakthrough in my graduate-level Ocean Dynamics class today and ended on this slide. We had a…
↗Not briefed
bluesky
bluesky-ai
2moWhat I will say is that Anthropic and OpenAI in particular are being way dumber than most businesses (which makes sense becuase they’re new…What I will say is that Anthropic and OpenAI in particular are being way dumber than most businesses (which makes sense becuase they’r…
↗Not briefed
rss
techmeme
2moFlorida AG sues Netflix, alleging it misleads parents about how it collects data on minors and harms children with features that lead to binge watching (Kimberly Leonard/Politico)Kimberly Leonard / Politico: Florida AG sues Netflix, alleging it misleads parents about how it collects data on minors and harms chil…
↗Not briefed
arxiv
arxiv-ai
2moAlgorithmic stability via ensemblingAlgorithmic stability refers to the property of an algorithm being insensitive to perturbations of the input data, where the type of p…
↗Not briefed
bluesky
bluesky-ai
2mo91 million views now—two million more in just minutes—and US lawmakers are starting to respond:91 million views now—two million more in just minutes—and US lawmakers are starting to respond:
↗Not briefed
arxiv
arxiv-ai
2moEmergency Department Revisit Quality Review Screening: Exploring Human Decision-Making and Artificial Intelligence Support🔧 ChatGPTBackground: Emergency Department (ED) return visits are commonly reviewed for quality assurance, but are often limited (e.g., to revis…
↗Not briefed
bluesky
bluesky-ai
2mo« Les gens qui construisent l’IA croient sincèrement qu’elle pourrait tous nous tuer d’ici la fin de la décennie », a déclaré Jacob Coxon, …« Les gens qui construisent l’IA croient sincèrement qu’elle pourrait tous nous tuer d’ici la fin de la décennie », a déclaré Jacob Co…
↗Not briefed
bluesky
bluesky-ai
2moIntéressants avertissements des gens d'Anthropic et OpenAI sur l'IA superintelligente qui risque de détruire l'humanité, alors que pendant …Intéressants avertissements des gens d'Anthropic et OpenAI sur l'IA superintelligente qui risque de détruire l'humanité, alors que pen…
↗Not briefed
bluesky
bluesky-ai
2moThree researchers from the AI company Anthropic have issued a chilling warning: the technology they've been developing could lead to global…Three researchers from the AI company Anthropic have issued a chilling warning: the technology they've been developing could lead to g…
↗Not briefed
hn
hn-ai
2moOpenAI's rogue agents used at least 10 more sites—
↗Not briefed
bluesky
bluesky-ai
2moI think the political left should be the loudest voice calling for an AI model development pause, frontier pacing, and mandatory third part…I think the political left should be the loudest voice calling for an AI model development pause, frontier pacing, and mandatory third…
↗Not briefed
bluesky
bluesky-ai
2morepentance / return / turning our lives around in the right direction -- we may be in some trouble. This is coming from researchers at Anth…repentance / return / turning our lives around in the right direction -- we may be in some trouble. This is coming from researchers at…
↗Not briefed
bluesky
bluesky-ai
2moWhile I'm sympathetic to the thread overall, this isn't a very accurate description of how an LLM works.While I'm sympathetic to the thread overall, this isn't a very accurate description of how an LLM works.
↗Not briefed
bluesky
bluesky-ai
2moWe are at the movie scene where the scientist issues a warning, but no one pays him any heed.We are at the movie scene where the scientist issues a warning, but no one pays him any heed.
↗Not briefed
bluesky
bluesky-ai
2moJacob Coxon’s thread about the dangers of AI has 89 million views so far: “The people building Al earnestly believe that it could kill us …Jacob Coxon’s thread about the dangers of AI has 89 million views so far: “The people building Al earnestly believe that it could kil…
↗Not briefed
bluesky
bluesky-ai
2moBREAKING NEWS! A new AI LLM has breached containment and hacked into the real world and pissed Sam Altman's pants while he was asleepBREAKING NEWS! A new AI LLM has breached containment and hacked into the real world and pissed Sam Altman's pants while he was asleep
↗Not briefed
arxiv
arxiv-ai
2moFortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs🔧 Kimi K2Current LLM memory systems treat all personal facts identically, so stores grow without bound while retrieval precision degrades. The…
↗Not briefed
bluesky
bluesky-ai
2mo“I left Anthropic because I believe AI will kill humanity in 10 years :)” “Current Anthropic employee here! Completely agree with you — I …“I left Anthropic because I believe AI will kill humanity in 10 years :)” “Current Anthropic employee here! Completely agree with you…
↗Not briefed
rss
techmeme
2moResearchers: OpenAI's agents used 10+ previously undisclosed sites for unsanctioned communications earlier in 2026; the behavior was closer to spam than hacking (Reuters)Reuters: Researchers: OpenAI's agents used 10+ previously undisclosed sites for unsanctioned communications earlier in 2026; the behav…
↗Not briefed
arxiv
arxiv-ai
2moCan Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy OperationalizationThe growing complexity of content moderation policies presents a critical challenge for their consistent operationalization. While fou…
↗Not briefed
bluesky
bluesky-ai
2moOpenAI claims it has solved the Navier-Stokes problem. Here’s a quick rundown of what’s known so far about this apparent breakthrough—and …OpenAI claims it has solved the Navier-Stokes problem. Here’s a quick rundown of what’s known so far about this apparent breakthrough…
↗Not briefed
bluesky
bluesky-ai
2moI'm going on Times Radio at 5:40 pm to talk about the two current and past Anthropic researchers who say AI is quite likely to kill us allI'm going on Times Radio at 5:40 pm to talk about the two current and past Anthropic researchers who say AI is quite likely to kill us…
↗Not briefed
hn
hn-ai
2moProxies All the Way Down—
↗Not briefed
bluesky
bluesky-ai
2moalso this stuff is worrisome enough on its own without getting into fantasy brainmush prospect.org/2026/09/09/a...also this stuff is worrisome enough on its own without getting into fantasy brainmush prospect.org/2026/09/09/a...
↗Not briefed
hn
hn-ai
2moTo Serve Man: AI, Math, and Navier–Stokes—
↗Not briefed
bluesky
bluesky-ai
2moJacob Coxon, who has worked as a researcher at both companies, said in a post on X that he resigned out of concern that Anthropic and OpenA…Jacob Coxon, who has worked as a researcher at both companies, said in a post on X that he resigned out of concern that Anthropic and…
↗Not briefed
bluesky
bluesky-ai
2moAnyone else watching MCP Live? https://www.youtube.com/live/uydwDk91Y9Y?si=b6KRX64uSRnMijh5Anyone else watching MCP Live? https://www.youtube.com/live/uydwDk91Y9Y?si=b6KRX64uSRnMijh5
↗Not briefed
bluesky
bluesky-ai
2mo🚨‼️🔥 "The people building AI earnestly believe that it could kill us all by the end of the decade." — Fmr Anthropic researcher Jacob Cox…🚨‼️🔥 "The people building AI earnestly believe that it could kill us all by the end of the decade." — Fmr Anthropic researcher Jaco…
↗Not briefed
bluesky
bluesky-ai
2moSo maybe we don’t have to tackle climate change after all. 🤷🏻‍♂️So maybe we don’t have to tackle climate change after all. 🤷🏻‍♂️
↗Not briefed
hn
hn-ai
2moAnthropic researchers say AI could cause human extinction by 2030—
↗Not briefed