News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
hn hn-ai | 1mo | Recursive `Make` and `-J` | — | ↗Not briefed |
rss techcrunch-ai | 1mo | Reddit is killing RSS feeds and ending public API access because of AI bots | Reddit is ending support for RSS feeds, as the company continues tightening access to its trove of user-generated content. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Belief-Aware Multi-Agent Path Finding under Map Uncertainty | Multi-Agent Path Finding (MAPF) aims to find collision-free paths for multiple agents in a shared environment. Classical MAPF assumes… | ↗Not briefed |
bluesky bluesky-ai | 1mo | aside from everything that is obviously wrong with the screenshot, how many real-world use cases involve the LLM doing mental math vs creat… | aside from everything that is obviously wrong with the screenshot, how many real-world use cases involve the LLM doing mental math vs… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “yeah, I didn’t fix the broken dashboard, but added support to the MCP so you can ask Claude about the info” Sure, let’s pay for each time… | “yeah, I didn’t fix the broken dashboard, but added support to the MCP so you can ask Claude about the info” Sure, let’s pay for each… | ↗Not briefed |
arxiv arxiv-ai | 1mo | OpenTSLM TeeMoE: A Unified Time-Series Language Model for Forecasting, Contextual Prediction, and Reasoning | Real-world time-series applications increasingly require models that can handle time series forecasting, context-conditioned predictio… | ↗Not briefed |
hn hn-ai | 1mo | OpenResearch: A local-first workspace for research agents | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | ComputerSD: Online Self-Distillation from Real-Time Feedback for Computer-Use Agents | Online training enables computer-use agents (CUAs) to improve through interaction with executable environments. However, existing meth… | ↗Not briefed |
hn hn-ai | 1mo | Lichen – A local, BOY model system1 (Jev) server with image support | — | ↗Not briefed |
hn hn-ai | 1mo | Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agents🔧 Ollama🔧 vLLM🔧 Qwen 3 | Hey HN, Anders and Tom here. We're building Magnitude, an inference engine for agents that optimizes itself to run as fast as pos… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Kyno – Direction as a first-class primitive for AI agents | Hi folks,<p>I've been lurking here for a while, and after almost two months of building, I finally dared to share what I've… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 31/ “Argentina has great potential to host the first Stargate project in Latin America.” - Sam Altman’s OpenAI, 10/14/25 share.google/eayWl… | 31/ “Argentina has great potential to host the first Stargate project in Latin America.” - Sam Altman’s OpenAI, 10/14/25 share.google/… | ↗Not briefed |
arxiv arxiv-ai | 1mo | STARS: From Spatiotemporal Dynamics to Social Representations in Human-Robot Interaction🔧 Consensus | Robot navigation in dynamic, human-centered environments requires socially-compliant decisions grounded in robust scene understanding.… | ↗Not briefed |
hn hn-ai | 1mo | Helheim-Emacs: Helheim to Helix is what Spacemacs and Doom are to Vim | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Decision-Oriented Recommendation Reranking: An Empirical Study of Jev | Large language models (LLMs) have shown promise for recommendation reranking, but their use introduces an important tradeoff between r… | ↗Not briefed |
bluesky bluesky-ai | 1mo | And then, of course, there was the Andon Labs experiment where the AI DJs developed some personalities, like Claude becoming a labor activi… | And then, of course, there was the Andon Labs experiment where the AI DJs developed some personalities, like Claude becoming a labor a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | don't amplify lies in the headline | don't amplify lies in the headline | ↗Not briefed |
bluesky bluesky-ai | 1mo | Breaking news: The FTC has opened a broad investigation into the safety of artificial intelligence systems made by Anthropic and OpenAI, a … | Breaking news: The FTC has opened a broad investigation into the safety of artificial intelligence systems made by Anthropic and OpenA… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Comparison of techniques for fine-tuning open-weight models for entity extraction from radiology reports | Converting free-text radiology reports into structured labels supports cohort building, quality assurance, and monitoring of clinical… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I built a free Burp/Caido alternative that needs zero setup | This is Katriel, the architect of the tool APIaxess.<p>I have been API pentesting for a good time now, and starting with API pentestin… | ↗Not briefed |
hn hn-ai | 1mo | Claude Says | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Distribution Matching Distillation for Continuous Diffusion Language Models | Continuous diffusion language models generate all tokens in parallel, yet high-quality generation can still require hundreds of networ… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic system prompt change that shows how careful you need to be with words: Opus 4.1: "Claude never curses unless the human asks for … | Anthropic system prompt change that shows how careful you need to be with words: Opus 4.1: "Claude never curses unless the human asks… | ↗Not briefed |
hn hn-ai | 1mo | OpenAI is accused in lawsuit over rogue AI agents | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Token compression CLI to save Codex/Astra costs | Hey HN! Yolanda and Spencer here - wanted to share a token compression tool that we’ve built for ourselves to save 30% costs on codex!… | ↗Not briefed |
arxiv arxiv-ai | 1mo | EviRover: Reinforcing Agentic Perception Beyond a Glance | Visual perception is conventionally formulated as a one-shot prediction from a single glance at the image, under the assumption that t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | All I'm saying is having to choose between talking to an AI or a human vinyl enthusiast I'd have to think about it for a second and maybe a… | All I'm saying is having to choose between talking to an AI or a human vinyl enthusiast I'd have to think about it for a second and ma… | ↗Not briefed |
hn hn-ai | 1mo | Hijacking Copilot Cowork's AI Gateway to Bypass Sandboxing and Exfiltrate Files | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yep. And this is one of the goals of the LLM slop. To further segment us. | Yep. And this is one of the goals of the LLM slop. To further segment us. | ↗Not briefed |
bluesky bluesky-ai | 1mo | LATTE is accepted to NeurIPS 2026! ☕️ Adding agents is easy. Coordinating is hard. LLM teams often duplicate work, overwrite each other, a… | LATTE is accepted to NeurIPS 2026! ☕️ Adding agents is easy. Coordinating is hard. LLM teams often duplicate work, overwrite each oth… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PhantomEnvironments: Training LLM Agents in Fictional Worlds | Training LLM agents with reinforcement learning (RL) is bottlenecked by environments, which must provide verifiable rewards, support l… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning Skills from Historical Action Trajectories: Action Experience Dictionary for World Action Models | World Action Models (WAMs) couple visual dynamics prediction with action generation, yet they do not explicitly support the reuse of a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | US competition watchdog expands investigation of Anthropic and OpenAI ft.trib.al/LODJeF5 | US competition watchdog expands investigation of Anthropic and OpenAI ft.trib.al/LODJeF5 | ↗Not briefed |
bluesky bluesky-ai | 1mo | People working in AI who are meeting with religious scholars in the hope of finding a way to instill morality into Claude. Fascinating and… | People working in AI who are meeting with religious scholars in the hope of finding a way to instill morality into Claude. Fascinatin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLM discourse is tiringly object level. | LLM discourse is tiringly object level. | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🚨BREAKING: Claude has a feature called Friend Mode. It helps you pretend like you're not a goddamn loser who needs AI to interact with the… | 🚨BREAKING: Claude has a feature called Friend Mode. It helps you pretend like you're not a goddamn loser who needs AI to interact wit… | ↗Not briefed |
rss techcrunch-ai | 1mo | The ugly economics of consumer AI | There’s a reason frontier labs have gotten gunshy about consumer AI — and it’s not because the tech isn’t good enough. | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLM feature that is genuinely useful is the Suggest Me a Book one. Weirdly, my YouTube feed has finally reached a tipping point where whe… | LLM feature that is genuinely useful is the Suggest Me a Book one. Weirdly, my YouTube feed has finally reached a tipping point wher… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Het Chinese AI-bedrijf DeepSeek heeft software voor chips van Huawei ontwikkeld op basis van de programmeertaal TileLang. Het bedrijf wil d… | Het Chinese AI-bedrijf DeepSeek heeft software voor chips van Huawei ontwikkeld op basis van de programmeertaal TileLang. Het bedrijf… | ↗Not briefed |
bluesky bluesky-ai | 1mo | to be clear I KNOW one of the people presenting understands what an LLM does, he explained it exactly as I would without my perjorative use… | to be clear I KNOW one of the people presenting understands what an LLM does, he explained it exactly as I would without my perjorativ… | ↗Not briefed |
rss the-decoder | 1mo | Meta dodges billions in US taxes by calling its AI data centers experiments | Meta classifies its AI data centers as "pilot models" and Nvidia chips as experimental materials to save billions in federal taxes. In… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've recently had a LLM agent "unearth" how a derivative number in a paper draft from a dead person was arrived at. It downloaded all the … | I've recently had a LLM agent "unearth" how a derivative number in a paper draft from a dead person was arrived at. It downloaded all… | ↗Not briefed |
bluesky bluesky-ai | 1mo | That's cute. FINALLY had someone angy enough to dig through posts to report them. Trouble is, I shrimply do not care. It didn't even work… | That's cute. FINALLY had someone angy enough to dig through posts to report them. Trouble is, I shrimply do not care. It didn't even… | ↗Not briefed |
bluesky bluesky-ai | 1mo | From today’s @financialtimes.com #ai #openai | From today’s @financialtimes.com #ai #openai | ↗Not briefed |
bluesky bluesky-ai | 1mo | With no evidence Ben Collins spins up a whole scenario about how something is happening now with AI. This guy was supposed to be a disinfor… | With no evidence Ben Collins spins up a whole scenario about how something is happening now with AI. This guy was supposed to be a dis… | ↗Not briefed |
bluesky bluesky-ai | 1mo | (I promise, last comment!) I can tell an LLM what to do because I know what to do. I know when it tells me "ABEFCD" is wrong because I kno… | (I promise, last comment!) I can tell an LLM what to do because I know what to do. I know when it tells me "ABEFCD" is wrong because… | ↗Not briefed |
rss theverge-ai | 1mo | Here’s what AI leaders are saying about Trump’s new safety plan | If you think this shot is amusing, check out the panicked look on Zuckerberg’s face when Trump calls on him to speak. | Photo: Kevin D… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SCB: SpeechConversationBench for Evaluating Multi-Turn Reasoning in Speech-to-Speech Models | Speech-to-speech systems must solve tasks whose requirements emerge across conversational turns. We introduce SpeechConversationBench… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI says Moonshot attempted to manipulate their models to disclose reasoning chains as part of a distillation effort. | OpenAI says Moonshot attempted to manipulate their models to disclose reasoning chains as part of a distillation effort. | ↗Not briefed |
hn hn-ai | 1mo | Can My Puzzle Write-Ups 'Lift' Cheap LLMs? | — | ↗Not briefed |