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 |
|---|---|---|---|---|
arxiv arxiv-ai | 10d | Latent Imagination for Parameterized Action Spaces | Parameterized-action MDPs (PAMDPs) ask an agent to choose a discrete action together with continuous parameters whose meaning depends… | ↗Not briefed |
bluesky bluesky-ai | 10d | “And the real problem is with complexity. Complexity actually doesn’t merely accrue. It can actually explode.” @bcantrill.bsky.social from … | “And the real problem is with complexity. Complexity actually doesn’t merely accrue. It can actually explode.” @bcantrill.bsky.social… | ↗Not briefed |
bluesky bluesky-ai | 10d | www.wired.com/story/startu... Radical Numerics announced Omnii Defense that is designed to flag natural and AI-engineered pathogens using a… | www.wired.com/story/startu... Radical Numerics announced Omnii Defense that is designed to flag natural and AI-engineered pathogens us… | ↗Not briefed |
hn hn-ai | 10d | In Memoriam: Tim Peters 1952–2026 | — | ↗Not briefed |
rss simon-willison | 10d | llm 0.37.2 | — | ↗Not briefed |
arxiv arxiv-ai | 10d | Where Is Uncertainty Observed? A Mechanistic Understanding of Uncertainty Quantification in LLMs | Knowing when a large language model's (LLM's) response can be trusted is critical for its safe use. Uncertainty quantification (UQ) ca… | ↗Not briefed |
bluesky bluesky-ai | 10d | ‘In October, AI token spend grew 2.1% month-over-month for OpenAI and 0.5% for Anthropic.’ econlab.substack.com/p/ai-index-n... | ‘In October, AI token spend grew 2.1% month-over-month for OpenAI and 0.5% for Anthropic.’ econlab.substack.com/p/ai-index-n... | ↗Not briefed |
bluesky bluesky-ai | 10d | “OpenAI and Anthropic are the only two companies really driving substantial growth in demand, so we could see a supply shock and a period o… | “OpenAI and Anthropic are the only two companies really driving substantial growth in demand, so we could see a supply shock and a per… | ↗Not briefed |
hn hn-ai | 10d | Launch HN: Pomerado (YC F26) – Turn any website into an MCP for agents | Hi HN, we’re Alan and Akshay from Pomerado (<a href="https://pomerado.ai/">https://pomerado.ai/</a>). We… | ↗Not briefed |
bluesky bluesky-ai | 10d | I would not, personally, feel comfortable assuming that whichever LLM they used was not running on an Azure server. | I would not, personally, feel comfortable assuming that whichever LLM they used was not running on an Azure server. | ↗Not briefed |
hn hn-ai | 10d | Show HN: Sonamu – Tree-Based Wayland Compositor | I use computers for everything and constantly have hundreds of tabs open. Existing paradigms limited me to grouping tabs by applicatio… | ↗Not briefed |
bluesky bluesky-ai | 10d | Orbiting ESA satellite gets voice-activated Doom 🎮 Amateur radio operators in Oslo launched Doom on the ESA's OPS-SAT PRETTY satellite us… | Orbiting ESA satellite gets voice-activated Doom 🎮 Amateur radio operators in Oslo launched Doom on the ESA's OPS-SAT PRETTY satelli… | ↗Not briefed |
bluesky bluesky-ai | 10d | Bring your own model is GA in Visual Studio 18.12, and it works with no GitHub account. Microsoft Foundry, OpenAI, Anthropic, Ollama, or a…🔧 Ollama🔧 OpenRouter🔧 LM Studio | Bring your own model is GA in Visual Studio 18.12, and it works with no GitHub account. Microsoft Foundry, OpenAI, Anthropic, Ollama,… | ↗Not briefed |
bluesky bluesky-ai | 10d | 🇬🇧 BoE's Bailey: two AI risks 🛡️ Cyber 🫧 A share bubble 🔗 https://www.trade2spike.com/ai-in-action/bank-of-england-bailey-zurich-ai-bu… | 🇬🇧 BoE's Bailey: two AI risks 🛡️ Cyber 🫧 A share bubble 🔗 https://www.trade2spike.com/ai-in-action/bank-of-england-bailey-zurich-… | ↗Not briefed |
arxiv arxiv-ai | 10d | DISCERN: Preserve at Write, Resolve at Read for Persistent LLM-Agent Memory | Enterprise agents accumulate memory across dozens of sessions in which facts contradict, get superseded, and blur as similarly named e… | ↗Not briefed |
bluesky bluesky-ai | 10d | I just remembered that goodAI is run by Keen Software House, makers of Space Engineers, whose CEO sadly has fully succumbed to LLM and vibe… | I just remembered that goodAI is run by Keen Software House, makers of Space Engineers, whose CEO sadly has fully succumbed to LLM and… | ↗Not briefed |
bluesky bluesky-ai | 10d | Startup Radical Numerics has developed a model that it says can identify natural and AI-engineered pathogens without an internet connection. | Startup Radical Numerics has developed a model that it says can identify natural and AI-engineered pathogens without an internet conne… | ↗Not briefed |
hn hn-ai | 10d | Ask HN: Does codex, Claude $20 subscription just seem unworkable now? | — | ↗Not briefed |
bluesky bluesky-ai | 10d | "Protestors and speakers were VERY aware of Schumer's daughter's job as a lobbyist for OpenAI" Just one reason Schumer supports the BAAJA … | "Protestors and speakers were VERY aware of Schumer's daughter's job as a lobbyist for OpenAI" Just one reason Schumer supports the B… | ↗Not briefed |
bluesky bluesky-ai | 10d | ‘OpenAI told investors Tuesday that its adjusted operating margin improved from minus 150% to minus 109% from the second to the third quart… | ‘OpenAI told investors Tuesday that its adjusted operating margin improved from minus 150% to minus 109% from the second to the third… | ↗Not briefed |
hn hn-ai | 10d | Show HN: Web Crawl Platformer Game | Hey! If you’re a fan of platformer games, and maybe even a fan of web crawlers, we have a treat for you!<p>Follow along with CCBot as… | ↗Not briefed |
rss techmeme | 10d | Sources: self-driving simulation software maker Applied Intuition is in talks to raise at least $1.5B led by General Catalyst at a ~$30B pre-money valuation (Bloomberg) | Bloomberg: Sources: self-driving simulation software maker Applied Intuition is in talks to raise at least $1.5B led by General Cataly… | ↗Not briefed |
arxiv arxiv-ai | 10d | Reliable Near-Field Source Enumeration and Positioning Informed by Two-Stage MUSIC | Accurate and reliable near-field multi-source localization is essential for location-aware processing in future wireless systems. Howe… | ↗Not briefed |
rss wired-ai | 10d | A Startup Wants to Fight AI Bioweapons With AI | Radical Numerics has developed a model that it says can identify natural and AI-engineered pathogens without an internet connection. | ↗Not briefed |
bluesky bluesky-ai | 10d | I was asking a DeepSeek where the sketchier chinese sites were and it was being a pain, this was basically a shortcut out of arguing with i… | I was asking a DeepSeek where the sketchier chinese sites were and it was being a pain, this was basically a shortcut out of arguing w… | ↗Not briefed |
hn hn-ai | 10d | Show HN: An audit for AI agents headed into air-gapped networks | — | ↗Not briefed |
bluesky bluesky-ai | 10d | As a Montanan, I’m particularly ashamed of Jon Tester for his involvement with Anthropic and Palantir. | As a Montanan, I’m particularly ashamed of Jon Tester for his involvement with Anthropic and Palantir. | ↗Not briefed |
bluesky bluesky-ai | 10d | If you’re fooled so easily, why should we still listen to you regarding this topic? | If you’re fooled so easily, why should we still listen to you regarding this topic? | ↗Not briefed |
bluesky bluesky-ai | 10d | I do think House House is sincere and mean well, but I think most of the folks calling this out are not saying that they didn't mean well b… | I do think House House is sincere and mean well, but I think most of the folks calling this out are not saying that they didn't mean w… | ↗Not briefed |
arxiv arxiv-ai | 10d | Terms of Care: Designing for Participation in Data Governance by Disability Communities | Improving disability representation in AI-generated media requires community-specific datasets, yet openly sharing such data can expos… | ↗Not briefed |
bluesky bluesky-ai | 10d | Once they go public everyone whose copyrighted works were stolen for training has standing to sue. Good luck to anthropic haha they'll have… | Once they go public everyone whose copyrighted works were stolen for training has standing to sue. Good luck to anthropic haha they'll… | ↗Not briefed |
arxiv arxiv-ai | 10d | Orbit: An Evidence-Grounded Benchmark for Evaluating LLMs on IT Operations Reasoning | LLMs now support IT operations (ITOps), but benchmarks rarely show why a model fails. We present ORBIT, a benchmark of ITOps questions… | ↗Not briefed |
bluesky bluesky-ai | 10d | This is why I can't trust "locally run open weight models." I can't trust that you didn't use corporate AI to make your local AI, I can't t… | This is why I can't trust "locally run open weight models." I can't trust that you didn't use corporate AI to make your local AI, I ca… | ↗Not briefed |
arxiv arxiv-ai | 10d | Outside the Purview of the Reward: Capability Drift in Multilingual Post-Training | Post-training can improve rewarded behavior while weakening capabilities users still need. We study two GRPO branches from a shared su… | ↗Not briefed |
bluesky bluesky-ai | 10d | > typos so bad that the LLM remembers them 4 compactions later | > typos so bad that the LLM remembers them 4 compactions later | ↗Not briefed |
arxiv arxiv-ai | 10d | A Reliability Study of Explainable AI for Freight Mode Classification | Machine learning models are increasingly used to support freight transportation decisions, but their predictions are rarely accompanie… | ↗Not briefed |
bluesky bluesky-ai | 10d | Claude’s web traffic surged 24% between September and October, a much bigger month-over-month increase than any of its closest AI rivals. | Claude’s web traffic surged 24% between September and October, a much bigger month-over-month increase than any of its closest AI riva… | ↗Not briefed |
bluesky bluesky-ai | 10d | Wait schumers daughter is on a lobbying firm for OpenAI 😩 | Wait schumers daughter is on a lobbying firm for OpenAI 😩 | ↗Not briefed |
arxiv arxiv-ai | 10d | Curiosity-0826-4B: Generative Medical Event Sequence Modeling Across Pediatric and Adult Populations | We introduce Curiosity-0826-4B, an updated version of the generative medical event foundation model trained on Epic Cosmos that simula… | ↗Not briefed |
bluesky bluesky-ai | 10d | In all seriousness “a game labelled non-AI used a bit of LLMs in coding” raises a taxonomy problem. “LLMs for some debugging and code“ is … | In all seriousness “a game labelled non-AI used a bit of LLMs in coding” raises a taxonomy problem. “LLMs for some debugging and code… | ↗Not briefed |
bluesky bluesky-ai | 10d | Anthropic giving Claude Ultra to 3000 CNCF Maintainers is pretty cool. https://claude.ai/cncf #kubecon | Anthropic giving Claude Ultra to 3000 CNCF Maintainers is pretty cool. https://claude.ai/cncf #kubecon | ↗Not briefed |
hn hn-ai | 10d | Anthropic donates 3000 ultra licenses to CNCF maintainers | — | ↗Not briefed |
hn hn-ai | 10d | Show HN: Haumea, an agent harness for tech support | Hi All,<p>I created this harness inspired by pi agent runtime, and dsh superagent ticketing system, to help in day to day tech support… | ↗Not briefed |
bluesky bluesky-ai | 10d | “Digital conquistador” is quite wonderful. #LLM | “Digital conquistador” is quite wonderful. #LLM | ↗Not briefed |
bluesky bluesky-ai | 10d | small illustration of why we should take AI chatbot answers with a grain of salt. these are the answers i got after asking about the heigh…🔧 ChatGPT | small illustration of why we should take AI chatbot answers with a grain of salt. these are the answers i got after asking about the… | ↗Not briefed |
bluesky bluesky-ai | 10d | seeing some VERY confused people trying to assert that code generators and LLM chatbots are something distinct from "generative AI", and th… | seeing some VERY confused people trying to assert that code generators and LLM chatbots are something distinct from "generative AI", a… | ↗Not briefed |
bluesky bluesky-ai | 10d | There's a bunch of "part of Lamplight" bots that link to "goodAI.network." I was like OK maybe its some kind of machine learning not traine… | There's a bunch of "part of Lamplight" bots that link to "goodAI.network." I was like OK maybe its some kind of machine learning not t… | ↗Not briefed |
arxiv arxiv-ai | 10d | Efficient Knowledge Transfer in Federated Bayesian Optimization through Neural Network Surrogates | Federated Bayesian Optimization (FedBO) has emerged as a powerful paradigm for enhancing Bayesian Optimization (BO) in distributed, pr… | ↗Not briefed |
bluesky bluesky-ai | 10d | The desire to remove Schumer is deep and includes skepticism that one daughter started working for OpenAI this fall in an obvious influence… | The desire to remove Schumer is deep and includes skepticism that one daughter started working for OpenAI this fall in an obvious infl… | ↗Not briefed |
bluesky bluesky-ai | 10d | Again, this is not the same thing. You're assuming that the LLM database you're building is so far superior to anything else that others ha… | Again, this is not the same thing. You're assuming that the LLM database you're building is so far superior to anything else that othe… | ↗Not briefed |