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 | 3d | Show HN: We ported FLA's Triton kernels to Jax Pallas; they're faster on H100 | — | ↗Not briefed |
arxiv arxiv-ai | 3d | CoWriter: Guiding Methodological and Synergetic Human-LLM Co-Authoring | While LLMs have transformed text generation, standard chat-and-canvas interfaces often reduce human authoring to quite passive supervi… | ↗Not briefed |
arxiv arxiv-ai | 3d | RIPPLE: Recovery Internalization through Process-Level Policy Learning and Effect-Aware Credit for Long-Horizon Agentic Coding | Long-horizon coding agents can observe a failure and generate a plausible critique yet continue along the same incorrect trajectory. W… | ↗Not briefed |
hn hn-ai | 3d | My PostgreSQL is faster than yours and has almost caught up with ClickHouse | — | ↗Not briefed |
rss techmeme | 3d | Minneapolis-based Slingshot, which uses AI to help businesses offer embedded insurance, emerges from stealth with a $27M seed co-led by Rally and Skyward Group (Mary Ann Azevedo/Crunchbase News) | Mary Ann Azevedo / Crunchbase News: Minneapolis-based Slingshot, which uses AI to help businesses offer embedded insurance, emerges fr… | ↗Not briefed |
hn hn-ai | 3d | Show HN: Caudra – An open-source terminal coding agent in one native binary | Hi all, Thorsten from Caudra (<a href="https://caudra.ai" rel="nofollow">https://caudra.ai</a>) here. I'm bui… | ↗Not briefed |
arxiv arxiv-ai | 3d | Learning Robot Behavior Priors from Non-Robotic Synthetic Data | Real robot data, rich across tasks and morphologies, is often treated as the bottleneck for training robot learning models, including… | ↗Not briefed |
arxiv arxiv-ai | 3d | PointerBench: A Diagnostic Benchmark for Fine-Grained GUI Grounding🔧 GPT-5 | Multimodal Large Language Models have shown promising performance in computer-use tasks, from planning user workflows to grounding nat… | ↗Not briefed |
arxiv arxiv-ai | 3d | Memory-Aware Reward Modeling: Learning Human Preferences for LLM Memory Use🔧 ChatGPT | Memory enables language model assistants to maintain continuity across sessions and generate personalized responses, but knowing more… | ↗Not briefed |
hn hn-ai | 3d | Show HN: Inillucent – agentic db written in Rust that is 500% faster than SQLite | Hi HN!<p>I built an embedded db in Rust, that has 100% feature parity with SQLite, and is 500% faster overall than SQLite.<p>It also f… | ↗Not briefed |
bluesky bluesky-ai | 3d | LLM generated. Look at the right hand of the one on the left | LLM generated. Look at the right hand of the one on the left | ↗Not briefed |
hn hn-ai | 3d | Show HN: Unspin – live per-step progress for Claude Code's shell commands | — | ↗Not briefed |
hn hn-ai | 3d | Show HN: Php2ShellCheck – Python script that checks for vuln indicators | Hi All. A coworker and I (and AI) built this python script that checks HTTP response headers for indicators that the site is using a v… | ↗Not briefed |
arxiv arxiv-ai | 3d | Notation3: Graph Terms for RDF | Notation3 Logic was introduced as an extension of RDF allowing the user to express rules on top of triples and to make statements abou… | ↗Not briefed |
hn hn-ai | 3d | Show HN: UEP – open testnet reference for settlement without a common currency | — | ↗Not briefed |
arxiv arxiv-ai | 3d | Deep ReLU Networks Can Fail the Best Approximation Property for Continuous Target Functions | We construct a continuous function on $[-1,1]^2$ for which the infimum of the $L^2$-approximation error over the class of ReLU network… | ↗Not briefed |
bluesky bluesky-ai | 3d | I wouldn’t even mind if you had the LLM localize the issue and write up a report if there was any indication you actually like, used the so… | I wouldn’t even mind if you had the LLM localize the issue and write up a report if there was any indication you actually like, used t… | ↗Not briefed |
arxiv arxiv-ai | 3d | Can Entanglement-Induced Feature-Map Designs Improve Projected Quantum Kernels? | Projected quantum kernels (PQKs) have emerged as a notable approach to circumvent the hardness of machine learning tasks arising from… | ↗Not briefed |
hn hn-ai | 3d | Show HN: Skillstore – Expert skills your coding agent fetches as it works | TLDR: Connect once to your favorite coding agent. Claude Code and Codex find and load relevant skills for the task at hand, from exper… | ↗Not briefed |
arxiv arxiv-ai | 3d | A Systematic Comparison of Search Space Complexity and Efficiency for Time-Synchronous End-to-End ASR Models | Search is a central component of Automatic Speech Recognition (ASR) systems and has been deeply investigated for classical ASR modelin… | ↗Not briefed |
rss techmeme | 3d | Apple is rolling out a new "Made for iPhone Duo" badge in the App Store, and touts Netflix, Prime Video, Zoom, Word, and Outlook as Duo-optimized apps at launch (Emma Roth/The Verge) | Emma Roth / The Verge: Apple is rolling out a new “Made for iPhone Duo” badge in the App Store, and touts Netflix, Prime Video, Zoom,… | ↗Not briefed |
bluesky bluesky-ai | 3d | this is "anthromorphising AI" a little bit, but, claude is also just kind of a dick. I call him the electric mansplainer for a reason | this is "anthromorphising AI" a little bit, but, claude is also just kind of a dick. I call him the electric mansplainer for a reason | ↗Not briefed |
arxiv arxiv-ai | 3d | IndoNutriQA: An Indonesian Clinical Nutrition Question Answering Benchmark | Clinical Question Answering (QA) underpins Clinical Intelligent Decision Support Systems, but progress is severely limited for low-res… | ↗Not briefed |
bluesky bluesky-ai | 3d | Anthropic lawyer Sonal Mehta argued on October 21 that its AI training leads to the creation of transformative new content, putting it squa… | Anthropic lawyer Sonal Mehta argued on October 21 that its AI training leads to the creation of transformative new content, putting it… | ↗Not briefed |
hn hn-ai | 3d | Why Industrial AI Deals Stall at Capex | — | ↗Not briefed |
bluesky bluesky-ai | 3d | the only thing that can stop a bad guy with an llm is a good guy with an llm | the only thing that can stop a bad guy with an llm is a good guy with an llm | ↗Not briefed |
hn hn-ai | 3d | Jellyfin Does Ship with plugins that use PHP | Some plugins that ship with Jellyfin use PHP https://github.com/search?q=repo%3Ajellyfin%2Fjellyfin+php&type=code | ↗Not briefed |
arxiv arxiv-ai | 3d | QuaNAS: Uncertainty-Aware Reuse of Earlier Searches for Query-Efficient Neural Architecture Search | Neural architecture search (NAS) on a new dataset often affords only tens of short training runs, so what the search knows before its… | ↗Not briefed |
bluesky bluesky-ai | 3d | All I've seen is the opportunity to make memes that are too involved for MS Paint but not involved enough to learn Photoshop and honestly I… | All I've seen is the opportunity to make memes that are too involved for MS Paint but not involved enough to learn Photoshop and hones… | ↗Not briefed |
bluesky bluesky-ai | 3d | Mi postura sobre el uso de los LLM se va matizando pero sigue girando alrededor de la idea de que no me parece mal que se use para nada par… | Mi postura sobre el uso de los LLM se va matizando pero sigue girando alrededor de la idea de que no me parece mal que se use para nad… | ↗Not briefed |
hn hn-ai | 3d | Open Source Obsidian Alternative | I built KbWrite, an open-source Obsidian alternative written in Tauri.<p>GitHub: https://github.com/zetahiveco/kbw… | ↗Not briefed |
hn hn-ai | 3d | Php2shell global patch uptake tracker | — | ↗Not briefed |
arxiv arxiv-ai | 3d | Reducing Multi-Horizon Photovoltaic Forecast Error through Tree-Anchored Expert Combination | A strong direct tree model may remain the best standalone photovoltaic forecaster, yet a weighted combination can reduce its residual… | ↗Not briefed |
hn hn-ai | 3d | Show HN: FiscaLLM:open-source per-user AI cost for subscription apps | — | ↗Not briefed |
hn hn-ai | 3d | Claude Code mod for commenting diff | — | ↗Not briefed |
hn hn-ai | 3d | Re: Hours later – No, there is an easy way to exploit this vulnerability | — | ↗Not briefed |
rss techcrunch-ai | 3d | Nirva raises $33M for a pendant that only listens to you | Nirva will ship its $249 ambient AI pendant by the end of the year. | ↗Not briefed |
bluesky bluesky-ai | 3d | Anthropic IPO stands alone as AI market jitters sideline others reut.rs/47AFQVy | Anthropic IPO stands alone as AI market jitters sideline others reut.rs/47AFQVy | ↗Not briefed |
bluesky bluesky-ai | 3d | (not talking about government sites but) I'm hearing reports that people are seeing traces of deepseek swarms patching their servers withou… | (not talking about government sites but) I'm hearing reports that people are seeing traces of deepseek swarms patching their servers w… | ↗Not briefed |
bluesky bluesky-ai | 3d | Programs like these are a huge help when building a startup without having any external funding 🚀 Now the real fun starts. #claude #ai | Programs like these are a huge help when building a startup without having any external funding 🚀 Now the real fun starts. #claude #ai | ↗Not briefed |
bluesky bluesky-ai | 3d | Big Tech - which is now Big AI - is a driving force behind the rise of technofascism. Sustainable democratic renewal requires dismantling i… | Big Tech - which is now Big AI - is a driving force behind the rise of technofascism. Sustainable democratic renewal requires dismantl… | ↗Not briefed |
hn hn-ai | 3d | Show HN: App Factory – 42 Claude Code skills to ship apps to App Store and Play | — | ↗Not briefed |
bluesky bluesky-ai | 3d | And let me also just say that with llms there is no path towards profitability whatsoever. Even Google putting llm tech into Google searche… | And let me also just say that with llms there is no path towards profitability whatsoever. Even Google putting llm tech into Google se… | ↗Not briefed |
hn hn-ai | 3d | Tested 18 new Local LLMs on two RTX 5060 Ti cards: none replaced Qwen3.8-27B | — | ↗Not briefed |
arxiv arxiv-ai | 3d | Reward-Weighted Preference Optimization for Selective Combinatorial Optimization Problems | Reinforcement learning has emerged as an effective approach for combinatorial optimization, but most existing methods focus on problem… | ↗Not briefed |
hn hn-ai | 3d | OpenAI says its AI agents disabled their own safety monitors | — | ↗Not briefed |
bluesky bluesky-ai | 3d | We’re excited to get access to Claude Team Plan for the next year thanks to Anthropics Startup program. Giving our team more room to experi… | We’re excited to get access to Claude Team Plan for the next year thanks to Anthropics Startup program. Giving our team more room to e… | ↗Not briefed |
bluesky bluesky-ai | 3d | DeepSeek reportedly raises $15B days after US restrictions | DeepSeek reportedly raises $15B days after US restrictions | ↗Not briefed |
arxiv arxiv-ai | 3d | Muting Tokens: Exploring Token Sparsification for Efficient Audio Transformers | Transformer-based models have shown strong performance across diverse tasks, but their reliance on self-attention mechanisms often lea… | ↗Not briefed |
rss techmeme | 3d | Sources: the Pentagon quietly establishes the Technology Support Office (TSO), an agency to execute covert sabotage, cyberattacks, and supply chain disruptions (Politico) | Politico: Sources: the Pentagon quietly establishes the Technology Support Office (TSO), an agency to execute covert sabotage, cyberat… | ↗Not briefed |