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
2moafter a few more days with deepseek v4.1 flash instead of GPT-6 we are now at 395tps prefill and 16tps decode waowafter a few more days with deepseek v4.1 flash instead of GPT-6 we are now at 395tps prefill and 16tps decode waow
↗Not briefed
hn
hn-ai
2moI Turned Google Glass into a Claude Code Terminal—
↗Not briefed
hn
hn-ai
2moClaude Mods, Functional Hooks in Claude Code and Desktop—
↗Not briefed
bluesky
bluesky-ai
2mo“Editor’s Note: Anthropic is a financial supporter of NPR.” Sept. 15, 2026“Editor’s Note: Anthropic is a financial supporter of NPR.” Sept. 15, 2026
↗Not briefed
hn
hn-ai
2moShow HN: MergeMe – 1 Slack card per PR/MR. Supports GitHub/GitLab(+ self hosted)Hi HN!<p>I built MergeMe because I worked in quite a few teams where it would take too long to pick up &amp; review PRs&#x2F;MRs, peop…
↗Not briefed
arxiv
arxiv-ai
2moAraMIP: Extending MIPVU Towards Metaphor Identification in ArabicMetaphor research has gained increasing attention due to its relevance to linguistic creativity, language use, cognitive processes, an…
↗Not briefed
bluesky
bluesky-ai
2moLawmakers panic over AI safety after rogue agent hacks rival platforms, exposing regulatory gaps. 🌐 Evidence: OpenAI breach, whistleblower…Lawmakers panic over AI safety after rogue agent hacks rival platforms, exposing regulatory gaps. 🌐 Evidence: OpenAI breach, whistleb…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic co-founder Jack Clark discusses growing concerns about artificial intelligence, recent warning signs and what he says companies s…Anthropic co-founder Jack Clark discusses growing concerns about artificial intelligence, recent warning signs and what he says compan…
↗Not briefed
bluesky
bluesky-ai
2moThe LLM that Natalie uses still can't quite get Trump's tone right. Swapping out the em-dashes for hypens was a nice touch, though. Very co…The LLM that Natalie uses still can't quite get Trump's tone right. Swapping out the em-dashes for hypens was a nice touch, though. Ve…
↗Not briefed
bluesky
bluesky-ai
2moKind of crazy that Fable 5.1 is the only recent Anthropic model that isn't an absolute pain to use and interact with. Doesn't seem sustaina…Kind of crazy that Fable 5.1 is the only recent Anthropic model that isn't an absolute pain to use and interact with. Doesn't seem sus…
↗Not briefed
bluesky
bluesky-ai
2moOpenAI is deep into developing RSI & RL’ing multiagent systems here Noam Brown describes a multiagent system that doesn’t do top-down mana…OpenAI is deep into developing RSI & RL’ing multiagent systems here Noam Brown describes a multiagent system that doesn’t do top-down…
↗Not briefed
arxiv
arxiv-ai
2moFROD: Feature Matching Residual Denoising Oracle Bone DecipherOracle bone script (OBS), one of the earliest Chinese writing systems, plays an important role in the study of Chinese etymology. Trad…
↗Not briefed
arxiv
arxiv-ai
2moEasy to Catch a Liar, Hard to Clear an Honest One: Language Models Diagnosing a Corrupted Reward Channel from a Verified RecordAn agent that learns from rewards has to trust whatever reports those rewards. When the reports suddenly change, either the world chan…
↗Not briefed
arxiv
arxiv-ai
2moPsychological Effects of Cultural Upheavals from Millions of Song Lyrics Over 100 YearsCultural upheavals impact many aspects of social life, and many studies have investigated their impact on language patterns. However,…
↗Not briefed
bluesky
bluesky-ai
2moBecause she hate-follows you on an alt and therefore is aware of the thread you did on the class counsel in the Anthropic suit, is my guess.Because she hate-follows you on an alt and therefore is aware of the thread you did on the class counsel in the Anthropic suit, is my…
↗Not briefed
hn
hn-ai
2moShow HN: EACL is a situated ReBAC authorization library for Datomic and DatahikeEACL is a situated ReBAC authorization library inspired by SpiceDB &amp; Google Zanzibar, built in Clojure and backed by Datomic Pro,…
↗Not briefed
bluesky
bluesky-ai
2moOpen AI/Chat GPT began emailing me out of the blue. A LOT. I’ve never used it and never subscribed to it. When I asked how it got my email,…Open AI/Chat GPT began emailing me out of the blue. A LOT. I’ve never used it and never subscribed to it. When I asked how it got my e…
↗Not briefed
hn
hn-ai
2moAustralia Killed the Duck Curve (With Batteries) [video]—
↗Not briefed
bluesky
bluesky-ai
2moLeading LLM companies are invoking unproven 'extinction' warnings to justify measures that could protect their market position and expand t…Leading LLM companies are invoking unproven 'extinction' warnings to justify measures that could protect their market position and exp…
↗Not briefed
hn
hn-ai
2moKubernetes for Haters—
↗Not briefed
arxiv
arxiv-ai
2moMemorisation bias in medical AIMedical AI models hold immense potential to improve patient outcomes, but they are also known to unintentionally memorise individual r…
↗Not briefed
hn
hn-ai
2moGerman companies rely almost exclusively on AI models and services from the USA—
↗Not briefed
rss
techmeme
2moAt Dreamforce, Salesforce announces Koa, a specialized CRM reasoning model built on Nvidia's Nemotron architecture, designed to power AI agents in Agentforce (Mike Wheatley/SiliconANGLE)Mike Wheatley / SiliconANGLE: At Dreamforce, Salesforce announces Koa, a specialized CRM reasoning model built on Nvidia's Nemotron ar…
↗Not briefed
bluesky
bluesky-ai
2moI can't believe I'm about to jump into Will discourse... But I think he's more right with "miraculous" than all the replies. Yes, we can …I can't believe I'm about to jump into Will discourse... But I think he's more right with "miraculous" than all the replies. Yes, we…
↗Not briefed
arxiv
arxiv-ai
2moGrounding SWE-Agent Decisions in Architecture-0 Design: Navigating Unknown Unknowns through Physical Mapping🔧 ConsensusAutonomous Software Engineering Agents (SWE-Agents) excel in deterministic coding tasks but struggle with Architecture 0, the nascent…
↗Not briefed
hn
hn-ai
2moShow HN: I made an 8 bit game that teaches beginners to pwn basic C programsI&#x27;ve been hosting small CTF for my alma mater high school every summer for the last few years, usually in the form of a sequence…
↗Not briefed
hn
hn-ai
2moGoogle SearchGuard. Anti-Bot System Exploring—
↗Not briefed
bluesky
bluesky-ai
2motl:dw AI company's focus on cyber apocalypse nonsense while ignoring actual harms to people (ai psychosis, self harm, encouragement of mass…tl:dw AI company's focus on cyber apocalypse nonsense while ignoring actual harms to people (ai psychosis, self harm, encouragement of…
↗Not briefed
bluesky
bluesky-ai
2moMne (skryté) použitie AI vadí iba na 3 miestach: 1. V produktoch, ktoré spracúvajú moje dáta (tu by som chcel opt-in model) 2. Vo vedeckých…Mne (skryté) použitie AI vadí iba na 3 miestach: 1. V produktoch, ktoré spracúvajú moje dáta (tu by som chcel opt-in model) 2. Vo vede…
↗Not briefed
bluesky
bluesky-ai
2moTalking on WNYC's Brian Lehrer show rn about AI fears, Anthropic/OpenAI's calls for slowing down, and Trump/Nvidia's counter-calls for putt…Talking on WNYC's Brian Lehrer show rn about AI fears, Anthropic/OpenAI's calls for slowing down, and Trump/Nvidia's counter-calls for…
↗Not briefed
hn
hn-ai
2moShow HN: KaozKit – JavaScript LLM agents on an engine built for microcontrollers—
↗Not briefed
rss
techcrunch-ai
2mo4 days left to exhibit at TechCrunch Disrupt 2026Last day to exhibit at Disrupt is Sept 18. 4 days left. Get your startup in front of 10,000+ founders, investors, operators and tech l…
↗Not briefed
arxiv
arxiv-ai
2moFluxVLA Engine: A One-Stop VLA Engineering Platform for Embodied IntelligenceVision-language-action (VLA) models, world-action models (WAMs), and offline reinforcement learning methods are rapidly expanding the…
↗Not briefed
hn
hn-ai
2moHeart damage may occur from getting too much light at night—
↗Not briefed
hn
hn-ai
2moShow HN: Panel – A research workspace where the agent can build its own panes—
↗Not briefed
hn
hn-ai
2moWe built a graph database service for agent memory—
↗Not briefed
arxiv
arxiv-ai
2moCross-Domain Inference for Human Localization: Applying Wi-Fi RSSI Data to CSI-Trained ModelsWi-Fi signal data can be used to compromise the privacy of individuals. While many existing approaches rely on Channel State Informati…
↗Not briefed
hn
hn-ai
2moSovereign: A Unified GPU Inference Substrate (Fractal Memory, Manifold Routing)—
↗Not briefed
rss
techmeme
2moOpenAI backs a provision of the FRONTIER Act, a bipartisan House proposal, that would require AI companies to embed outside evaluators to ensure model safety (Brendan Bordelon/Politico)Brendan Bordelon / Politico: OpenAI backs a provision of the FRONTIER Act, a bipartisan House proposal, that would require AI companie…
↗Not briefed
hn
hn-ai
2moEffects of music training on the child's brain and cognitive development—
↗Not briefed
bluesky
bluesky-ai
2moone of the weirdest tics of LLM-generated posters is how they always have three meaningless declarative sentence fragments at the bottom li…one of the weirdest tics of LLM-generated posters is how they always have three meaningless declarative sentence fragments at the bott…
↗Not briefed
rss
gary-marcus
2moBREAKING: Secret US AI evaluation framework has been partly revealedVery partly
↗Not briefed
arxiv
arxiv-ai
2moMyoFlow: Anchor-Tied Rectified Flow for HD-sEMG Gesture Recognition Across Sessions and SubjectsHigh-density surface electromyography (HD-sEMG) gesture recognition supports prosthetic control, assistive robotics, and rehabilitatio…
↗Not briefed
arxiv
arxiv-ai
2moEnd-to-End Latency-Minimizing and Load-Balanced Request Scheduling for Edge LLM Inference in Agentic AI ServicesLarge language model (LLM)-powered agentic AI services increasingly demand low-latency inference, motivating the deployment of LLMs ac…
↗Not briefed
bluesky
bluesky-ai
2mopeople are still going to be complaining about Bernie Bros while being chased through the bombed-out streets of New York by Anthropic hunte…people are still going to be complaining about Bernie Bros while being chased through the bombed-out streets of New York by Anthropic…
↗Not briefed
hn
hn-ai
2moLMAX Disruptor – High Performance Inter-Thread Messaging Library—
↗Not briefed
bluesky
bluesky-ai
2mo12 years after Time gave us a tipping point now AI gets one i guess? damn i guess ai has some rough years in front of them. remind me to te…12 years after Time gave us a tipping point now AI gets one i guess? damn i guess ai has some rough years in front of them. remind me…
↗Not briefed
hn
hn-ai
2moHow to process incoming call audio on Android / iOS?- use case is blocking spam calls<p>- when an incoming call arrives, the caller s audio is taken in real time and processed through an…
↗Not briefed
bluesky
bluesky-ai
2moChina released an open AI model and the Internet turned it into UNCENSORED, BLACK HAT, AI No guardrails, no babysitter and potentially capa…China released an open AI model and the Internet turned it into UNCENSORED, BLACK HAT, AI No guardrails, no babysitter and potentially…
↗Not briefed
bluesky
bluesky-ai
2moRussian spies ran a campaign against Kyiv on Anthropic’s Claude, and the AI chatbot’s maker caught them at it The operators automated phish…Russian spies ran a campaign against Kyiv on Anthropic’s Claude, and the AI chatbot’s maker caught them at it The operators automated…
↗Not briefed