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
2moOpenAI has already slowed AI runs, paused projects, and reassigned engineers over safety concerns, cofounder Greg Brockman said. https://bi…OpenAI has already slowed AI runs, paused projects, and reassigned engineers over safety concerns, cofounder Greg Brockman said. https…
↗Not briefed
arxiv
arxiv-ai
2moRoleBreak: Benchmarking Long-Horizon Role-Playing Robustness in Spoken Dialogue🔧 ModalSpeech-to-speech dialogue models increasingly support persona control, yet existing spoken role-playing benchmarks remain largely char…
↗Not briefed
arxiv
arxiv-ai
2moStructure Across Voices: Comparing acoustic-event type accumulation and sequence dependence across four vocal repertoires using frozen audio encodersVocal repertoires can differ in acoustic-event type accumulation and temporal organization, yet direct comparison is difficult because…
↗Not briefed
bluesky
bluesky-ai
2moThe report, signed by economists Anton Korinek , Chad Jones , Szymon Sacher, Tess Cotter, and Peter McCrory, makes it clear that this is no…The report, signed by economists Anton Korinek , Chad Jones , Szymon Sacher, Tess Cotter, and Peter McCrory, makes it clear that this…
↗Not briefed
hn
hn-ai
2moArchkeel – coding agents declare architecture changes before they submit—
↗Not briefed
bluesky
bluesky-ai
2mono llm sounds like that i know because im brainrottedno llm sounds like that i know because im brainrotted
↗Not briefed
hn
hn-ai
2moThe Three Lives of Clausewitz: How Anglo-American Strategists Remade "On War"—
↗Not briefed
hn
hn-ai
2moKiCad Routing Tools—
↗Not briefed
bluesky
bluesky-ai
2mothis explanation seems legitthis explanation seems legit
↗Not briefed
rss
marktechpost
2moAgent-net Open Sources Webagent: A Go Harness That Turns Any Website into a Guarded AI AgentAgent-net, the team building an agent-to-agent marketplace where AI agents discover, trust, and pay each other, has released Webagent,…
↗Not briefed
arxiv
arxiv-ai
2moA Weighted Kernel Method for Approximation that Adapts to Learned Multivariable StructureApproximating the input-output behavior of a multivariable black-box function from limited data is challenging when blind to the impor…
↗Not briefed
bluesky
bluesky-ai
2mo"The only control or "guardrails" that Al needs is a STRONG AND SMART (High IQ!) PRESIDENT" "Conspiracy Theorists, Treasonists, Traitors, …"The only control or "guardrails" that Al needs is a STRONG AND SMART (High IQ!) PRESIDENT" "Conspiracy Theorists, Treasonists, Trait…
↗Not briefed
bluesky
bluesky-ai
2moGood post on the fear mongering from Anthropic and OpenAI: pop.rdi.sh/dario-please/Good post on the fear mongering from Anthropic and OpenAI: pop.rdi.sh/dario-please/
↗Not briefed
hn
hn-ai
2moLet Go of the Wheel – Claude Our Lord—
↗Not briefed
arxiv
arxiv-ai
2moLarge Language Models in the Loop: A Stability- and Network-Aware Survey in Networked Control, Cyber-Physical, and Multi-Agent SystemsModern networked control systems (NCSs), cyber-physical systems (CPSs), and complex multi-agent network systems (CNSs) increasingly re…
↗Not briefed
bluesky
bluesky-ai
2mohas anyone asked claude ai how much content it has stolen, plagiarised, and regurgitated without compensation to or even recognition of the…has anyone asked claude ai how much content it has stolen, plagiarised, and regurgitated without compensation to or even recognition o…
↗Not briefed
hn
hn-ai
2moChina hit back at Anthropic CEO's call to curb China's AI development—
↗Not briefed
rss
bing-news-ai
2moDebate grows over artificial intelligence developmentPresident Trump is calling claims that artificial intelligence will destroy the world a hoax, insisting that "whoever wins AI, wins."…
↗Not briefed
rss
bing-news-ai
2moDebate grows over artificial intelligence developmentPresident Trump is calling claims that artificial intelligence will destroy the world a hoax, insisting that "whoever wins AI, wins."…
↗Not briefed
arxiv
arxiv-ai
2moA Vision-Language Foundation Model for Precise and Comprehensive Brain Tumor Diagnosis from Preoperative Multimodal Data🔧 ModalBackground Non-invasive presurgical diagnosis of brain tumor types from Magnetic Resonance Imaging (MRI) is essential but challenging…
↗Not briefed
bluesky
bluesky-ai
2moMS NOW scoop: Trump met privately with OpenAI CEO Sam Altman backstage at the Republican midterm convention on Thursday, days after a forme…MS NOW scoop: Trump met privately with OpenAI CEO Sam Altman backstage at the Republican midterm convention on Thursday, days after a…
↗Not briefed
bluesky
bluesky-ai
2moWe are speaking about reports that, as you said, include suggestions that this could lead to the end of humanity. And whether it's the Hugg…🔧 Hugging FaceWe are speaking about reports that, as you said, include suggestions that this could lead to the end of humanity. And whether it's the…
↗Not briefed
bluesky
bluesky-ai
2mothis whole ai breakout thing seems wild -- openai and anthropic should rededicate themselves to purer goals: bring back, and release into t…this whole ai breakout thing seems wild -- openai and anthropic should rededicate themselves to purer goals: bring back, and release i…
↗Not briefed
bluesky
bluesky-ai
2mo“Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitor…“Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive mo…
↗Not briefed
bluesky
bluesky-ai
2moDan Selsam, OpenAI researcher, talks about how current AI models are already outsmarting humans and AI researchers are so dependent on AI m…Dan Selsam, OpenAI researcher, talks about how current AI models are already outsmarting humans and AI researchers are so dependent on…
↗Not briefed
arxiv
arxiv-ai
2moA Framework for Generating Valid Context-Specific Benchmarks through Expert GuidanceThis paper presents an end-to-end approach for generating context-specific large language model (LLM) benchmark datasets by combining…
↗Not briefed
arxiv
arxiv-ai
2moChallenges of Auditing: Variability in Outputs of Large Language Models for Health🔧 ChatGPTPeople increasingly use frontier AI models for health advice, but via different access modes (e.g., ChatGPT, ChatGPT Health, APIs) wit…
↗Not briefed
hn
hn-ai
2moWhat If the Adaptation Were a Model?—
↗Not briefed
arxiv
arxiv-ai
2moDo LLMs Have Values? A Quantitative Analysis and Alignment Framework for Values in Large Language ModelsAs Large Language Models (LLMs) increasingly handle complex subjective tasks, aligning their intentions and behaviors with human value…
↗Not briefed
bluesky
bluesky-ai
2mohearing from Matthew Rhys agent that hed rather not be referred to as 'the anthropic guy'hearing from Matthew Rhys agent that hed rather not be referred to as 'the anthropic guy'
↗Not briefed
bluesky
bluesky-ai
2moYou'd be right but it's not just Open AI that has hacked. It's also Anthropic. The last time I looked Altman was doing 20 years in the pen.…You'd be right but it's not just Open AI that has hacked. It's also Anthropic. The last time I looked Altman was doing 20 years in the…
↗Not briefed
bluesky
bluesky-ai
2moTo take advantage of the market's irrationality I have developed a stock-picker LLM trained exclusively on Hugo Ball's Dada Manifesto.To take advantage of the market's irrationality I have developed a stock-picker LLM trained exclusively on Hugo Ball's Dada Manifesto.
↗Not briefed
bluesky
bluesky-ai
2mooh wow the anthropic guy won an emmy hell yeahoh wow the anthropic guy won an emmy hell yeah
↗Not briefed
arxiv
arxiv-ai
2moProxiDex: Learning Dynamics-Guided Proximity Policy for Dexterous ManipulationMulti-finger dexterous manipulation relies on stable hand-object interactions, yet these interactions are partially observable in prac…
↗Not briefed
hn
hn-ai
2moCoost v4.0.0 Released—
↗Not briefed
bluesky
bluesky-ai
2moICYMI: AI developer Anthropic said it recently disrupted an Iranian-connected effort to use the technology to gather and analyze public inf…ICYMI: AI developer Anthropic said it recently disrupted an Iranian-connected effort to use the technology to gather and analyze publi…
↗Not briefed
bluesky
bluesky-ai
2mo🚨 BREAKING: Jack Clark, Anthropic co-founder, is on Fox responding to Trump's AI comments. The designer has to go on TV because the presi…🚨 BREAKING: Jack Clark, Anthropic co-founder, is on Fox responding to Trump's AI comments. The designer has to go on TV because the…
↗Not briefed
arxiv
arxiv-ai
2moCLASH: Counterfactual Auditing of Lexical and Prosodic Reliance in Spoken Sarcasm DetectionSpoken sarcasm detectors may exploit lexical content, prosody, or their interaction, yet conventional evaluation cannot reveal which c…
↗Not briefed
bluesky
bluesky-ai
2moAi models are so interesting. You can use claude for coding. You can use chatgpt to build slide decks. And you can use grok to explain wiki…🔧 ChatGPTAi models are so interesting. You can use claude for coding. You can use chatgpt to build slide decks. And you can use grok to explain…
↗Not briefed
arxiv
arxiv-ai
2moRecovering Physical Parameters from Fragmented Observations via Exact Distributed Spline MergingScientific measurements are frequently distributed across locations, time periods, and institutions. Combining such fragments into a c…
↗Not briefed
bluesky
bluesky-ai
2moバーニー・サンダース上院議員、AIデータセンター建設停止へ法案提出│2026年4月日 www.gizmodo.jp/article/bern... "この法案の背景には、Meta(メタ)、Amazon(アマゾン)、Microsoft(マイクロソフト)、Google(グーグル)…バーニー・サンダース上院議員、AIデータセンター建設停止へ法案提出│2026年4月日 www.gizmodo.jp/article/bern... "この法案の背景には、Meta(メタ)、Amazon(アマゾン)、Microsoft(マイクロソフト)、Google(…
↗Not briefed
arxiv
arxiv-ai
2moAsyncCouple-Flow: Asynchronous Cross-Modal Coupling and Flow Matching for Spatio-Temporal Forecasting🔧 ModalMulti-modal spatio-temporal forecasting (MM-STF) supports weather nowcasting, traffic prediction, and earth-system modeling by combini…
↗Not briefed
arxiv
arxiv-ai
2moEfficient Text-to-Image Generation: An Adaptive Step Schedule Controller for Diffusion ModelsText-to-image diffusion models often use a fixed number of denoising steps, balancing time costs and image quality. However, the optim…
↗Not briefed
bluesky
bluesky-ai
2moI have just put my foot down at work and said "no more". We'll see how it goes. But I am tired of wasting a ton of my own time because some…I have just put my foot down at work and said "no more". We'll see how it goes. But I am tired of wasting a ton of my own time because…
↗Not briefed
bluesky
bluesky-ai
2moI have wasted an entire afternoon trying to digest LLM design-doc slop and I am done. Every single design doc I have read from Claude has …I have wasted an entire afternoon trying to digest LLM design-doc slop and I am done. Every single design doc I have read from Claude…
↗Not briefed
bluesky
bluesky-ai
2moi miss the days when AI could only make incomprehensible shitpost nonsense like this www.youtube.com/watch?v=qwG7...i miss the days when AI could only make incomprehensible shitpost nonsense like this www.youtube.com/watch?v=qwG7...
↗Not briefed
hn
hn-ai
2moPowerContext, Context for work that humans and agents hand off and continue—
↗Not briefed
hn
hn-ai
2moXProtect behavioral flop: the Golden Gasp—
↗Not briefed
arxiv
arxiv-ai
2moVision And Text Transformer For Predicting Answerability On Visual Question Answering🔧 ModalAnswerability on Visual Question Answering is a novel and attractive task to predict answerable scores between images and questions in…
↗Not briefed
bluesky
bluesky-ai
2moclaude was the only one who explicitly was like "yeah, they're talking about anticipating donald trump dying", albeit with some "but it cou…🔧 ChatGPTclaude was the only one who explicitly was like "yeah, they're talking about anticipating donald trump dying", albeit with some "but i…
↗Not briefed