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
1molol, someone on the school tech Google group used Claude to code an extension to hide AI in search results for students. Best use of AI yet.lol, someone on the school tech Google group used Claude to code an extension to hide AI in search results for students. Best use of A…
↗Not briefed
bluesky
bluesky-ai
1moHow many more are they not telling us about? If ever there was a time to promote/teach/stress critical thinking and independent thought, it…How many more are they not telling us about? If ever there was a time to promote/teach/stress critical thinking and independent though…
↗Not briefed
hn
hn-ai
1moExclusive – Anthropic sets up biology lab as it ramps AI drug program—
↗Not briefed
arxiv
arxiv-ai
1moPRISM-BN: A Controlled Corpus and Benchmark for Text-to-Parameterized Bayesian Network Extraction🔧 GPT-5Probabilistic Graphical Models (PGMs), especially Bayesian Networks (BNs), expose directed structure and probabilistic parameters, mak…
↗Not briefed
hn
hn-ai
1moNew England RetroFest and the Maine Vintage Computer Society—
↗Not briefed
bluesky
bluesky-ai
1moYes! Can we also talk about how LLM and bots are being pushed as an accessibility tool in schools!?! Because it’s all connected across the …Yes! Can we also talk about how LLM and bots are being pushed as an accessibility tool in schools!?! Because it’s all connected across…
↗Not briefed
arxiv
arxiv-ai
1moAccelerating Dense LLMs via L0-regularized Mixture-of-ExpertsLarge language models (LLMs) achieve strong performance but suffer from slow and costly inference. Existing acceleration methods often…
↗Not briefed
hn
hn-ai
1moTalus: eBPF ransomware detector for Linux (Rust)—
↗Not briefed
bluesky
bluesky-ai
1moAsked whether OpenAI should face legal ramifications for the Hugging Face incident, Lina Khan tells @cwarzel.bsky.social, “It seems that th…🔧 Hugging FaceAsked whether OpenAI should face legal ramifications for the Hugging Face incident, Lina Khan tells @cwarzel.bsky.social, “It seems th…
↗Not briefed
bluesky
bluesky-ai
1moThe hiring of Jessica Schumer is another example of powerful people and families, regardless of party, who are deeply intertwined with the …The hiring of Jessica Schumer is another example of powerful people and families, regardless of party, who are deeply intertwined with…
↗Not briefed
bluesky
bluesky-ai
1moWith the latest external hack of OpenAI via Claude, closed models continue to be the tip of the iceberg on AI risks, not open models. They …With the latest external hack of OpenAI via Claude, closed models continue to be the tip of the iceberg on AI risks, not open models.…
↗Not briefed
arxiv
arxiv-ai
1moFrom Code Archival to Knowledge Graph: Bridging Software Heritage, COAR Notify and WikidataSoftware is a first-class scientific object, yet validated links between source code and the scholarly record remain largely absent fr…
↗Not briefed
arxiv
arxiv-ai
1moSamsone: A Family of Open Small Audio Language Models for On-Device InferenceThe success of Large Audio Language Models has driven the development of massive multimodal networks exceeding billions of parameters.…
↗Not briefed
bluesky
bluesky-ai
1mo言葉による指示やスケッチで建物や家具を自由自在に3D化できるAI搭載3Dモデリングツール「Cartesian」 https://gigazine.net/news/20260918-cartesian-formas-ai-3d-model/言葉による指示やスケッチで建物や家具を自由自在に3D化できるAI搭載3Dモデリングツール「Cartesian」 https://gigazine.net/news/20260918-cartesian-formas-ai-3d-model/
↗Not briefed
hn
hn-ai
1moWhere Waterfox Stands on AI and LLMs—
↗Not briefed
arxiv
arxiv-ai
1moMulti-Domain Clustering via Measure QuantizationClustering is a fundamental task in data analysis, typically addressed through centroid-based methods such as K-means. In this work, w…
↗Not briefed
arxiv
arxiv-ai
1moRethinking Human-Aligned Evaluation: An Analysis of Semantic Metrics Beyond WERWord Error Rate (WER), the most commonly used metric for Automatic Speech Recognition (ASR), treats every lexical deviation from the r…
↗Not briefed
rss
openai-news
1moIntroducing the Australian Youth Safety BlueprintOpenAI introduces the Australian Youth Safety Blueprint, a six-pillar roadmap for safer AI experiences that protect and empower young…
↗Not briefed
hn
hn-ai
1moSwitzerland Is Scrapping ILS as GPS Spoofing Hits Crews Daily—
↗Not briefed
arxiv
arxiv-ai
1moWhen Steering Fails in Latent Reasoning: A Latent-to-Language Transition GapActivation steering has become a widely used approach for controlling language models during explicit chain-of-thought (CoT) reasoning…
↗Not briefed
bluesky
bluesky-ai
1moIt’s one thing to argue that we can do things with the intentional stance — sure, of course. But it’s another thing completely to have noth…It’s one thing to argue that we can do things with the intentional stance — sure, of course. But it’s another thing completely to have…
↗Not briefed
hn
hn-ai
1moShow HN: Big Calendar – an iOS calendar with big fonts and readable eventsI&#x27;ve been using google calendar for over 10 years, but since turning 40 I&#x27;ve found the fonts too small.<p>So I built an opin…
↗Not briefed
arxiv
arxiv-ai
1moOutcome-Conditioned End-Effector Geometry Across Vision-Language-Action PoliciesVision-language-action (VLA) policies solve the same manipulation task through different action interfaces, but task success alone doe…
↗Not briefed
arxiv
arxiv-ai
1moBeyond Gaussian Worlds: Latent Geometry Matters for JEPAsRecent Joint-Embedding Predictive Architectures (JEPAs) prevent representation collapse by constraining learned representations to fol…
↗Not briefed
bluesky
bluesky-ai
1mo今もこの国を動かしている。AIが選ぶ「日本の偉人TOP3」【ChatGPT/Gemini/Claude】 youtu.be/7bjd68_WxmA?... @YouTubeより🔧 ChatGPT今もこの国を動かしている。AIが選ぶ「日本の偉人TOP3」【ChatGPT/Gemini/Claude】 youtu.be/7bjd68_WxmA?... @YouTubeより
↗Not briefed
arxiv
arxiv-ai
1moAnalysing the Linearity of Linguistic Relations in Language Model Embedding SpacesWe propose a framework to analyse how strongly different linguistic relations are linearly encoded in language model embedding spaces.…
↗Not briefed
bluesky
bluesky-ai
1moMy favorite part so far is that the authenticity checker we have for bios works - we can see if you copy/paste anything from an LLM and als…My favorite part so far is that the authenticity checker we have for bios works - we can see if you copy/paste anything from an LLM an…
↗Not briefed
bluesky
bluesky-ai
1moEhrlich gesagt, so ähnliche Fragen frage ich mich _die_ _ganze_ _Zeit_ beim Thema KI, z.B. auch in Richtung Urheberrecht. Warum gelten für …Ehrlich gesagt, so ähnliche Fragen frage ich mich _die_ _ganze_ _Zeit_ beim Thema KI, z.B. auch in Richtung Urheberrecht. Warum gelten…
↗Not briefed
hn
hn-ai
1moShow HN: AI agents design an open chip, the best one gets fabricated—
↗Not briefed
hn
hn-ai
1moOpenAI close to solving Millennium Prize problem Hodge Conjecture—
↗Not briefed
arxiv
arxiv-ai
1moConfigurable Multi-Stage Vision Pipeline for Crop Disease and Pest DiagnosisFarmer.Chat is Digital Green's farm advisory service for smallholder farmers. When something looks wrong with a crop, the farmer takes…
↗Not briefed
arxiv
arxiv-ai
1moSynthDemo-RL: Breaking the Zero-Reward Barrier in VLA Adaptation with LLM-Guided Synthetic DemonstrationsFine-tuning Vision-Language-Action (VLA) models commonly relies on human teleoperation demonstrations, while reinforcement learning (R…
↗Not briefed
bluesky
bluesky-ai
1moQue los trabajadores de OpenAI digan que la IA puede acabar con la Humanidad, es un poco como cuando los cristianos te dicen que Dios te va…Que los trabajadores de OpenAI digan que la IA puede acabar con la Humanidad, es un poco como cuando los cristianos te dicen que Dios…
↗Not briefed
hn
hn-ai
1moHolistic, a Jenkins dashboard for the office TV—
↗Not briefed
rss
the-decoder
1mo42 leading mathematicians warn that AI existential risk is real and urgent42 Fellows of the Royal Society, including Fields Medal winners Martin Hairer and Peter Scholze, warn of existential AI risks in an op…
↗Not briefed
bluesky
bluesky-ai
1moBon, c'est une jolie hallucination de l'IAg, assez classique : elle prend le nom d'un auteur qui existe bien (ici une journaliste du Monde)…Bon, c'est une jolie hallucination de l'IAg, assez classique : elle prend le nom d'un auteur qui existe bien (ici une journaliste du M…
↗Not briefed
bluesky
bluesky-ai
1mo“A Microsoft executive called the training of AI models the ‘largest theft of labor in human history.’ In another instance…that same execut…“A Microsoft executive called the training of AI models the ‘largest theft of labor in human history.’ In another instance…that same e…
↗Not briefed
arxiv
arxiv-ai
1moRiemannian Neural Hamiltonian Flows: Geodesic Symplectic Transport and InterpretabilityHamiltonian normalizing flows are attractive generative models because their phase-space maps are invertible and volume preserving, bu…
↗Not briefed
hn
hn-ai
1moMy Proxmox LVM pool died and I'm fine with that—
↗Not briefed
hn
hn-ai
1moContextClip – Clean web docs into token-optimized LLM prompts—
↗Not briefed
bluesky
bluesky-ai
1moただのメカ好きが作ったシリーズ 🔫GUN CAROUSEL SERIES M24 SWS CUSTOM✨ 1988年に米陸軍で採用された、Remington Model 700系をベースとする7.62×51mmボルトアクション狙撃銃「M24 SWS」。🔭 今回は大型スコ…ただのメカ好きが作ったシリーズ 🔫GUN CAROUSEL SERIES M24 SWS CUSTOM✨ 1988年に米陸軍で採用された、Remington Model 700系をベースとする7.62×51mmボルトアクション狙撃銃「M24 SWS」。🔭 今回…
↗Not briefed
bluesky
bluesky-ai
1moAI "security" is a sick joke. If you use #AI you are making youself and others vulnerable. It's only a matter of time before this costs …AI "security" is a sick joke. If you use #AI you are making youself and others vulnerable. It's only a matter of time before this c…
↗Not briefed
hn
hn-ai
1moSaudis Tell European Refiners They'll Get No Crude Next Month—
↗Not briefed
bluesky
bluesky-ai
1moMore people should remember that OpenAI and Anthropic did not invent the tech they useMore people should remember that OpenAI and Anthropic did not invent the tech they use
↗Not briefed
bluesky
bluesky-ai
1moInterviewer: "Hey, uhm, why did you not do basic cybersecurity?" OpenAI: "well, you see, [sci-fi scenario] so we might as well not bother …Interviewer: "Hey, uhm, why did you not do basic cybersecurity?" OpenAI: "well, you see, [sci-fi scenario] so we might as well not bo…
↗Not briefed
bluesky
bluesky-ai
1moCar l'un des moyens de détecter l'usage de LLM, ce sont les détournements de source : les LLM ont tendance à citer une source pour appuyer …Car l'un des moyens de détecter l'usage de LLM, ce sont les détournements de source : les LLM ont tendance à citer une source pour app…
↗Not briefed
bluesky
bluesky-ai
1moBon, bravo au LLM, il a réussi à sortir une URL correcte, me dis-je, mi-ironique, mi-blasé. Passons aux fameux articles de presse sans URL…Bon, bravo au LLM, il a réussi à sortir une URL correcte, me dis-je, mi-ironique, mi-blasé. Passons aux fameux articles de presse san…
↗Not briefed
bluesky
bluesky-ai
1moExecutives from OpenAI and Microsoft were reportedly worried about ChatGPT training that scraped millions of news articles.🔧 ChatGPTExecutives from OpenAI and Microsoft were reportedly worried about ChatGPT training that scraped millions of news articles.
↗Not briefed
bluesky
bluesky-ai
1mo1Password couldn’t figure it out but Anthropic could Christ1Password couldn’t figure it out but Anthropic could Christ
↗Not briefed
bluesky
bluesky-ai
1moANTHROPIC FIGURED IT OUT? Anthropic? I gotta say this one hurts.ANTHROPIC FIGURED IT OUT? Anthropic? I gotta say this one hurts.
↗Not briefed