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 | 1mo | PlantShade: Predicting Plant Shadows for Lighting-Aware Robotic Agricultural Operation | Plant growth and agricultural production form the foundation of a country's sustainable development and directly impact human liveliho… | ↗Not briefed |
arxiv arxiv-ai | 1mo | How Much of a Real Workload Can LLM-Generated GPU Kernels Actually Reach? | Language models can now write GPU kernels that outperform PyTorch. We evaluate five model configurations on KernelBench level 1 and fi… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Vim-like, hyper-efficient, LLM power tool - 20-80k tokens not 350k+ | I built a focused TUI tool for efficient parallelized tasks, early independent testers expressed 90%+ API savings and more direct resu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FedeRage: Provably Convergent Agnostic Federated Learning under General Client Drift | Federated learning (FL) enables collaborative model training without sharing raw data, but its performance degrades under non-IID data… | ↗Not briefed |
bluesky bluesky-ai | 1mo | President Donald Trump's posts come after a series of warnings by AI experts about the technologies' capabilities and the July 2026 cyberat…🔧 Hugging Face | President Donald Trump's posts come after a series of warnings by AI experts about the technologies' capabilities and the July 2026 cy… | ↗Not briefed |
bluesky bluesky-ai | 1mo | im already seeing it now, waterfox is trying to bank on Firefox's pro AI dipshittery and is going to slowly put in more fucking LLM bullshi… | im already seeing it now, waterfox is trying to bank on Firefox's pro AI dipshittery and is going to slowly put in more fucking LLM bu… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Jev uses 100% synthetic training data | Jev uses 100% synthetic training data | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anyone who believes the LLM-based AI tech is truly conscious is either very stupid or certifiably in the grip of complete delusion | Anyone who believes the LLM-based AI tech is truly conscious is either very stupid or certifiably in the grip of complete delusion | ↗Not briefed |
bluesky bluesky-ai | 1mo | i'm not swiping this | i'm not swiping this | ↗Not briefed |
bluesky bluesky-ai | 1mo | i think Jev is enough of an improvement that it is worth actively building towards it in places where that's possible and that this meaning… | i think Jev is enough of an improvement that it is worth actively building towards it in places where that's possible and that this me… | ↗Not briefed |
rss techcrunch-ai | 1mo | Is the AI safety debate about safety or control? | Not everyone agrees with Amodei's call for globally coordinated action for AI safety. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Can a fully open model help make AI more “prosocial” through a game? Soham Padia built Steering Arena, where players try to elicit kind & …🔧 Elicit | Can a fully open model help make AI more “prosocial” through a game? Soham Padia built Steering Arena, where players try to elicit ki… | ↗Not briefed |
hn hn-ai | 1mo | Astra for Law | — | ↗Not briefed |
hn hn-ai | 1mo | Microsoft, OpenAI lose fight to hide internal docs admitting scraping is theft | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | storage.courtlistener.com/recap/gov.us... | storage.courtlistener.com/recap/gov.us... | ↗Not briefed |
hn hn-ai | 1mo | Killer AI Is Here. We're Using It in Iran | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Following multiple reports of AI agent swarms hacking websites and coordinating via secret message boards, along with a dire message for hu… | Following multiple reports of AI agent swarms hacking websites and coordinating via secret message boards, along with a dire message f… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: HSL-Noise – A noise generator mapped to the HSL color space | I felt that the classical "colors of noise" (Pink, Brownian, White...) were very limiting when describing a very specific no… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Physically Based Rendering in the Latent Space | Image diffusion models have shown impressive image generation capabilities but are often hard to control, in contrast to classical com… | ↗Not briefed |
bluesky bluesky-ai | 1mo | très bon papier. prenez le temps de le lire. www.arretsurimages.net/chroniques/c... | très bon papier. prenez le temps de le lire. www.arretsurimages.net/chroniques/c... | ↗Not briefed |
bluesky bluesky-ai | 1mo | What happened Judicial documents declassified this week in the lawsuit by The New York Times against OpenAI and Microsoft (initiated three … | What happened Judicial documents declassified this week in the lawsuit by The New York Times against OpenAI and Microsoft (initiated t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I honestly wonder if that's part of why Anthropic has Catholic clergy on its "Soul Document" team. Catholicism argues there's a divinely in… | I honestly wonder if that's part of why Anthropic has Catholic clergy on its "Soul Document" team. Catholicism argues there's a divine… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A U.S. government website used Alibaba’s Qwen AI model to help users search proposed federal regulations, days after the FBI accused Alibab… | A U.S. government website used Alibaba’s Qwen AI model to help users search proposed federal regulations, days after the FBI accused A… | ↗Not briefed |
bluesky bluesky-ai | 1mo | futurism.com/future-socie... ““If true, it chimes with an old sci-fi prediction and with what we know of human nature,” Thomas wrote. “If w… | futurism.com/future-socie... ““If true, it chimes with an old sci-fi prediction and with what we know of human nature,” Thomas wrote.… | ↗Not briefed |
hn hn-ai | 1mo | After the Write Hole (Btrfs) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | US government website used AI search tool from China that FBI said copied Anthropic reut.rs/4reNXQx | US government website used AI search tool from China that FBI said copied Anthropic reut.rs/4reNXQx | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI's new legal brain hits the market ⚖️ OpenAI unveiled Astra for Law: GPT-6 Astra, a model tailored for legal work with a dedicated s… | OpenAI's new legal brain hits the market ⚖️ OpenAI unveiled Astra for Law: GPT-6 Astra, a model tailored for legal work with a dedica… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Lightweight Plug-in Gate for Transformer-Based Time-Series Forecasters | Covariate-rich time-series forecasting requires deciding how external variables enter the target forecasting path. Existing Transforme… | ↗Not briefed |
bluesky bluesky-ai | 1mo | futurism.com/artificial-i... “The incidents include a still-unreleased model inserting “jailbreak-like instructions” into its own notes to … | futurism.com/artificial-i... “The incidents include a still-unreleased model inserting “jailbreak-like instructions” into its own note… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I think distillation is the right frame, but what's being distilled is the chain of actions that an llm goes through as it runs multiple se… | I think distillation is the right frame, but what's being distilled is the chain of actions that an llm goes through as it runs multip… | ↗Not briefed |
hn hn-ai | 1mo | Fastembed – Rust library for generating vector embeddings, reranking locally | — | ↗Not briefed |
rss techmeme | 1mo | Andrew Ng says renewed AI extinction warnings are "much more science fiction than science" and the latest "wave of PR" is probably intended to shape regulation (Micah Barkley/Bloomberg) | Micah Barkley / Bloomberg: Andrew Ng says renewed AI extinction warnings are “much more science fiction than science” and the latest “… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI is targeting the nation's top law firms with new advanced legal AI tools. This move intensifies its rivalry with Anthropic and marks… | OpenAI is targeting the nation's top law firms with new advanced legal AI tools. This move intensifies its rivalry with Anthropic and… | ↗Not briefed |
hn hn-ai | 1mo | Rdap-Java – A ~5KB Java 11 RDAP client to check domain availability | — | ↗Not briefed |
rss techcrunch-ai | 1mo | UN turns to Google to make its global data ready for AI agents | The shift comes after a UNICEF test found leading AI models struggled to accurately retrieve global development statistics. | ↗Not briefed |
rss wired-ai | 1mo | The AI ‘Slowdown’ Is an Antitrust Mess | By framing their efforts as a “slowdown” rather than an industry-wide push for better security standards, AI labs may have set themsel… | ↗Not briefed |
bluesky bluesky-ai | 1mo | are we sure OpenAI isn't a plot by Sam Altman to sign onto his verification scheme app | are we sure OpenAI isn't a plot by Sam Altman to sign onto his verification scheme app | ↗Not briefed |
bluesky bluesky-ai | 1mo | been using kagi as my search engine for a bit over a year now, and i love how i can manually prioritize or downrank results, or even report… | been using kagi as my search engine for a bit over a year now, and i love how i can manually prioritize or downrank results, or even r… | ↗Not briefed |
bluesky bluesky-ai | 1mo | being at a physical game jam with no real ai policy and walking around the tables seeing every other person with a claude sidebar open was … | being at a physical game jam with no real ai policy and walking around the tables seeing every other person with a claude sidebar open… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Stiefel-AdamW: Geometry-Aware AdamW for Linear Factorization Blocks | A pervasive structural pattern in modern deep learning is the linear factorization block: a submodule of the form $W = BA$ in which tw… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "The Pentagon’s Maven Smart System, built by Palantir, also uses Anthropic’s Claude AI to 'semi-autonomously rank targets by strategic impo… | "The Pentagon’s Maven Smart System, built by Palantir, also uses Anthropic’s Claude AI to 'semi-autonomously rank targets by strategic… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ‘“OpenAI has disclosed six reports of “unexpected or concerning” behavior in artificial-intelligence models as the debate on AI safety beco… | ‘“OpenAI has disclosed six reports of “unexpected or concerning” behavior in artificial-intelligence models as the debate on AI safety… | ↗Not briefed |
bluesky bluesky-ai | 1mo | There's a network of doomsday AI sex cults competing to become gods (Meta, X), create gods (Anthropic, OpenAI), or worship traditional gods… | There's a network of doomsday AI sex cults competing to become gods (Meta, X), create gods (Anthropic, OpenAI), or worship traditional… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I wonder how much LLM traffic is people just playing: either 'blow me down, it can do that', or 'LoL look at how bad this is' without the… | I wonder how much LLM traffic is people just playing: either 'blow me down, it can do that', or 'LoL look at how bad this is' withou… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Brilliant and necessary move @gavinnewsom.bsky.social, sir. Right now there is a rogue AI model that is preventing starter homes from being… | Brilliant and necessary move @gavinnewsom.bsky.social, sir. Right now there is a rogue AI model that is preventing starter homes from… | ↗Not briefed |
bluesky bluesky-ai | 1mo | You are also referring to “model collapse”, which happens when an LLM starts training on other AI generated data. It eventually devolves t… | You are also referring to “model collapse”, which happens when an LLM starts training on other AI generated data. It eventually devol… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Excel macros are Turing-Complete. Therefore, given a sufficiently large spreadsheet, you could write an LLM in Excel. And if an LLM can re… | Excel macros are Turing-Complete. Therefore, given a sufficiently large spreadsheet, you could write an LLM in Excel. And if an LLM c… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: A map of 68 programs and jobs in AI safety research | Nine sections covering: research programs, research orgs, evaluators and red-teaming, policy and governance, career-changers, funding,… | ↗Not briefed |
hn hn-ai | 1mo | ClickHouse's New TimeSeries Engine: Drop-In Prometheus Replacement | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Guide to working in the open. A guide to working in the open, based on UK Government culture form a decade ago. I've created for myself t… | Guide to working in the open. A guide to working in the open, based on UK Government culture form a decade ago. I've created for mys… | ↗Not briefed |