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
rss
techmeme
24dThe PostgreSQL Global Development Group delays the general availability of PostgreSQL 19 after LLM-assisted code reviews uncovered architectural defects (Lindsay Clark/The Register)Lindsay Clark / The Register: The PostgreSQL Global Development Group delays the general availability of PostgreSQL 19 after LLM-assis…
↗Not briefed
hn
hn-ai
24dHow do you know your agent's memory layer is helping, not hurting?—
↗Not briefed
hn
hn-ai
24dNVidia Refused GTC Berlin Entry to NeoCloud Providers who also use AMD—
↗Not briefed
bluesky
bluesky-ai
24dBrave shields AI chats with TEEs 🔒 Brave Software has set up end-to-end encrypted conversations for its Leo AI assistant using Trusted Ex…Brave shields AI chats with TEEs 🔒 Brave Software has set up end-to-end encrypted conversations for its Leo AI assistant using Trust…
↗Not briefed
bluesky
bluesky-ai
24dI'm not very active on Reddit these days but like... most of the subs I frequent are either anti-AI as a whole or at least have "no genAI" …I'm not very active on Reddit these days but like... most of the subs I frequent are either anti-AI as a whole or at least have "no ge…
↗Not briefed
bluesky
bluesky-ai
24dbut the upside about being in this hellscape of people's weaponized incompentence, i have never felt more like i have to speak for myself, …but the upside about being in this hellscape of people's weaponized incompentence, i have never felt more like i have to speak for mys…
↗Not briefed
bluesky
bluesky-ai
24dI think I am at the point where I might just end up muting the words ai, generative, LLMs and their model names so I can pretend none of th…I think I am at the point where I might just end up muting the words ai, generative, LLMs and their model names so I can pretend none…
↗Not briefed
bluesky
bluesky-ai
24dOpenAI has added four more cases of misbehaving AI agents to a disclosures page it stood up following the Hugging Face hack in July, in an …🔧 Hugging FaceOpenAI has added four more cases of misbehaving AI agents to a disclosures page it stood up following the Hugging Face hack in July, i…
↗Not briefed
bluesky
bluesky-ai
24dThe AHM is literally demanding that ppl submitting to journals be allowed to opt out of LLM-based fraud and plagiarism detection systems, w…The AHM is literally demanding that ppl submitting to journals be allowed to opt out of LLM-based fraud and plagiarism detection syste…
↗Not briefed
bluesky
bluesky-ai
24dWhite House announces plans to replace Katie Zacharia with a vintage Big Mouth Billy Bass retrofitted with on device AI. Its BassGPT model …White House announces plans to replace Katie Zacharia with a vintage Big Mouth Billy Bass retrofitted with on device AI. Its BassGPT m…
↗Not briefed
hn
hn-ai
24dAnthropic's Valuation Collapsed by $500B—
↗Not briefed
hn
hn-ai
24dWe Got Duo Support in NetNewsWire—
↗Not briefed
bluesky
bluesky-ai
24dit definitely bothered people on wall street considering this week china has been unveiling cheaper more powerful chips and llm models and …it definitely bothered people on wall street considering this week china has been unveiling cheaper more powerful chips and llm models…
↗Not briefed
hn
hn-ai
24dShow HN: Tinywm – a scrolling window manager for macOS, inspired by Niri—
↗Not briefed
hn
hn-ai
24dLLM says something, my tool track if it followed its promise on tools—
↗Not briefed
bluesky
bluesky-ai
24dAlso related to emails, ANTHROPIC I DON'T WANT A SINGLE MORE EMAIL FROM YOU if it isn't "HERE IS YOUR MONEY" or better yet "OOOPS WE ARE DE…Also related to emails, ANTHROPIC I DON'T WANT A SINGLE MORE EMAIL FROM YOU if it isn't "HERE IS YOUR MONEY" or better yet "OOOPS WE A…
↗Not briefed
bluesky
bluesky-ai
24d#Xiomi released a 311B parameter #MoE #model, with 15B active, #FP8. #1M context window, Text, Multimodal: Image, Video, Audio. With built-…#Xiomi released a 311B parameter #MoE #model, with 15B active, #FP8. #1M context window, Text, Multimodal: Image, Video, Audio. With b…
↗Not briefed
hn
hn-ai
24dAnthropic assigned highest junk credit rating in June—
↗Not briefed
rss
google-research
24dGrounding AI agents in biological reality: Building blocks for autonomous wet labsGeneral Science
↗Not briefed
hn
hn-ai
24dHistory of Cdrom Technology—
↗Not briefed
bluesky
bluesky-ai
24dAi skepticism is justified. We need regulations. I tried Claude code recently and the capabilities are scary.Ai skepticism is justified. We need regulations. I tried Claude code recently and the capabilities are scary.
↗Not briefed
rss
the-decoder
24dChatGPT becomes a full Chrome browser assistant that reads and acts on your open tabs🔧 ChatGPTOpenAI's updated Chrome extension lets ChatGPT access open browser tabs, sum up content, and use "Computer Use" in Work mode to fill o…
↗Not briefed
hn
hn-ai
24dShow HN: Webhix – Self-hosted webhook inspector in a single binary—
↗Not briefed
bluesky
bluesky-ai
24dIt exists to be scraped, indexed and swallowed by AI models. VOLUME PASSES FOR EVIDENCE Take a single propaganda claim. Produce 20,000 cop…It exists to be scraped, indexed and swallowed by AI models. VOLUME PASSES FOR EVIDENCE Take a single propaganda claim. Produce 20,00…
↗Not briefed
bluesky
bluesky-ai
24dfor weeks i was like holy crap everyone selling on vinted is a fucking model until i asked someone a question about sizing (girl wearing th…for weeks i was like holy crap everyone selling on vinted is a fucking model until i asked someone a question about sizing (girl weari…
↗Not briefed
bluesky
bluesky-ai
24d#Meta releases #Muse Spark “Open” #MoE #model 816B parameters, 42B active. Multilingual, Multimodal, Coding, tool calling. Not actually #…#Meta releases #Muse Spark “Open” #MoE #model 816B parameters, 42B active. Multilingual, Multimodal, Coding, tool calling. Not actua…
↗Not briefed
bluesky
bluesky-ai
24dFinally free from the security assisted shackles of my contractor lanyard now I've a badge. Also I know the AI bubble is about to burst but…Finally free from the security assisted shackles of my contractor lanyard now I've a badge. Also I know the AI bubble is about to burs…
↗Not briefed
bluesky
bluesky-ai
24dyoud think anthropic and openai would come up with a jevlike before random other companies like cloudflareyoud think anthropic and openai would come up with a jevlike before random other companies like cloudflare
↗Not briefed
hn
hn-ai
24dThe Vertex Cover Problem on Cubic Bridgeless Graphs is NP-complete and also in P—
↗Not briefed
bluesky
bluesky-ai
24d@mistralai.bsky.social released #Timeseries forecasting up to 209M parameters #model #OpenSource MIT modified lic #AI #LLM #SLM@mistralai.bsky.social released #Timeseries forecasting up to 209M parameters #model #OpenSource MIT modified lic #AI #LLM #SLM
↗Not briefed
bluesky
bluesky-ai
24dThat's not you defining it, that's an AI getting shit wrong. Also, it's a good example of how the way you ask questions matters when you'r…That's not you defining it, that's an AI getting shit wrong. Also, it's a good example of how the way you ask questions matters when…
↗Not briefed
bluesky
bluesky-ai
24dprefiro deixar LLM editar meus textos. sem brincadeira.prefiro deixar LLM editar meus textos. sem brincadeira.
↗Not briefed
hn
hn-ai
24dDesigning more efficient quantum circuits with Claude—
↗Not briefed
hn
hn-ai
24dClojurific: ClojureScript compiler fork, optimized for modern JavaScript—
↗Not briefed
arxiv
arxiv-ai
24dRethinking the design space of Vanilla Pixel Diffusion TransformersPixel-space diffusion models have recently re-emerged as a compelling alternative to latent diffusion, avoiding the lossy latent bottl…
↗Not briefed
bluesky
bluesky-ai
24dI'm beginning to anticipate that some Trump shenanigans truly crash the economy and Anthropic researchers come pouring out of their office …I'm beginning to anticipate that some Trump shenanigans truly crash the economy and Anthropic researchers come pouring out of their of…
↗Not briefed
arxiv
arxiv-ai
24dBehavioral Couplings: From Formation to Behavioral Emergence in Language Models🔧 ElicitFine-tuning large language models (LLMs) can induce behaviors beyond those explicitly targeted. For example, training a model to respo…
↗Not briefed
arxiv
arxiv-ai
24dAI-Driven Reproducibility and Analytical Robustness in Scientific DiscoveryReproducible and analytically robust research findings are essential to scientific progress, but evaluating these properties typically…
↗Not briefed
arxiv
arxiv-ai
24dDiscoveryRSI: Recursively Learning to Discover Valuable HypothesesScientific discovery requires developing worthwhile hypotheses into practical investigations, yet learning from this evolving research…
↗Not briefed
bluesky
bluesky-ai
24dLeider auch in diesem Clip der übliche Fehler, Next-Token-Prediction mit dem tatsächlich gelernten Wissen des Modells gleichzusetzen. Trot…Leider auch in diesem Clip der übliche Fehler, Next-Token-Prediction mit dem tatsächlich gelernten Wissen des Modells gleichzusetzen.…
↗Not briefed
bluesky
bluesky-ai
24dtelling the robot girl that I installed an LLM during maintenance so I can dismiss all of her complaints as hallucinationstelling the robot girl that I installed an LLM during maintenance so I can dismiss all of her complaints as hallucinations
↗Not briefed
bluesky
bluesky-ai
24dOpenAI put Arizonans at risk by failing to implement adequate user protections or transparency measures, despite clear indications that min…OpenAI put Arizonans at risk by failing to implement adequate user protections or transparency measures, despite clear indications tha…
↗Not briefed
arxiv
arxiv-ai
24dTask-aligned spectral inductive biases in parameterized quantum circuitsParameterized quantum circuits are widely investigated in quantum machine learning, yet how their architecture can provide task-aligne…
↗Not briefed
arxiv
arxiv-ai
24dBeyond the Answer: Student-Conditioned Trace Construction for Reasoning DistillationA reasoning trace can accompany a correct answer without providing effective supervision for a particular student. We study how to con…
↗Not briefed
arxiv
arxiv-ai
24dWaiting Buys Exactly Two Moments: Endogenous Identification in Prediction-Driven AllocationA planner with a fixed budget of preventive help must decide whom to help and when. Waiting reveals risk but removes intervention oppo…
↗Not briefed
rss
techcrunch-ai
24dIndia becomes the latest country to report rogue AI agent activityThe agents got in through credentials and connection settings left exposed in app configurations, public code repositories, and overly…
↗Not briefed
bluesky
bluesky-ai
24dLaunched this week on TrustList: Nemotron 3.5 Super VL NVIDIA's open-weights vision-language model, released 15 October 2026: 120B paramet…Launched this week on TrustList: Nemotron 3.5 Super VL NVIDIA's open-weights vision-language model, released 15 October 2026: 120B pa…
↗Not briefed
bluesky
bluesky-ai
24di'm an LLM user this applies to literally all my knowledge everi'm an LLM user this applies to literally all my knowledge ever
↗Not briefed
rss
techmeme
24dThree CoreWeave co-founders have sold $3.3B in stock since its March 2025 IPO, including $2.4B in 2026 alone; CoreWeave stock fell 60%+ from its June 2025 peak (Dylan Sloan/Bloomberg)Dylan Sloan / Bloomberg: Three CoreWeave co-founders have sold $3.3B in stock since its March 2025 IPO, including $2.4B in 2026 alone;…
↗Not briefed
arxiv
arxiv-ai
24dQuality-of-Transmission Estimation in Quantum Networks: Compute, Simulate, or Learn?Future quantum networks will rely on control-plane decisions such as routing, scheduling, entanglement provisioning, purification, and…
↗Not briefed