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 | 17d | ZeroRAG: Adaptive Semantic Peer Routing for Decentralized Retrieval-Augmented Generation | Retrieval-augmented generation (RAG) is a standard approach to enhance the reliability of large language model (LLM) outputs by retrie… | ↗Not briefed |
arxiv arxiv-ai | 17d | AraEngage: Interpretable Linguistic Correlates of Engagement in Arabic Facebook News | Engagement features developed for English are routinely applied to Arabic, but unvocalized Arabic changes what a lexical feature measu… | ↗Not briefed |
arxiv arxiv-ai | 17d | How Finite Brains Generalize Without Bound: Spatial Memory as a Universal Computer | A child who learns to add small numbers adds numbers of any length. LSTMs and Transformers trained on short sequences fail on longer o… | ↗Not briefed |
arxiv arxiv-ai | 17d | Adaptive Semantic Branching: Disagreement-Localized Cross-Trajectory Regeneration for Reasoning Candidate Expansion | Reasoning trajectories generated for the same problem often arrive at conflicting answers. These disagreements reveal alternative inte… | ↗Not briefed |
hn hn-ai | 17d | Computer use is GA in Gemini API | — | ↗Not briefed |
arxiv arxiv-ai | 17d | Same Answer, Different Future: Matched Predictions Do Not Imply Matched Belief Updating in Language Models | Language models can produce similar probabilistic predictions after observing outcomes, reading summaries, receiving instructions, or… | ↗Not briefed |
arxiv arxiv-ai | 17d | GuidedWAM: Training-Free Energy Guidance for World Action Models in Autonomous Driving | World action models (WAMs) for autonomous driving couple the driving policy and the visual world model in a single generative network… | ↗Not briefed |
bluesky bluesky-ai | 17d | Willing to admit now that I wrote answers 1, 2, and 1. savage.love/savagelove/2... | Willing to admit now that I wrote answers 1, 2, and 1. savage.love/savagelove/2... | ↗Not briefed |
arxiv arxiv-ai | 17d | EDGAR: Efficient Diffusion-Driven Gaussian Avatar Refinement via Synthetic Failure | Recent advances in Gaussian splatting have enabled efficient and photorealistic head avatar reconstruction, yet monocular video-based… | ↗Not briefed |
rss techmeme | 17d | Sources detail Dario Amodei's AI beliefs, shaped by forums like GiveWell and based on EA and "rationalism", and his disputes with Sam Altman and Greg Brockman (New York Times) | New York Times: Sources detail Dario Amodei's AI beliefs, shaped by forums like GiveWell and based on EA and “rationalism”, and his di… | ↗Not briefed |
rss the-decoder | 17d | Anthropic loses product lead Catherine Wu behind Claude Code to OpenAI | Catherine Wu, the founding product lead for Claude Code, is leaving Anthropic for OpenAI, where she'll head up product management for… | ↗Not briefed |
arxiv arxiv-ai | 17d | Structure from Collision via Hypothesis-Space Reasoning | Knowledge of the hidden interior structure of an object, including the presence and location of cavities, is essential for predictive… | ↗Not briefed |
hn hn-ai | 17d | Show HN: Ergon - build AI agents in code or on a canvas that generates Python | — | ↗Not briefed |
arxiv arxiv-ai | 17d | Persistent Retrieval for Conversational Agents: A Study of Retrieval Efficiency | Conversational agents can retrieve information directly or delegate retrieval to a subagent. This design choice affects how retrieval… | ↗Not briefed |
hn hn-ai | 17d | WebsiteBench: Can AI Agents Rebuild Websites Through Browser Exploration? | — | ↗Not briefed |
rss techmeme | 17d | Discord is moving its desktop app from Electron to Rust for improved speed and memory safety; its CTO shares an image of the Windows app using just 290MB of RAM (Tom Warren/The Verge) | Tom Warren / The Verge: Discord is moving its desktop app from Electron to Rust for improved speed and memory safety; its CTO shares a… | ↗Not briefed |
bluesky bluesky-ai | 17d | Hi Meerkat, to avoid the people scrutiny they now call #datacenters AI compute campuses also for #tax avoidance #IRS #OpenAI #Anthropic #AI… | Hi Meerkat, to avoid the people scrutiny they now call #datacenters AI compute campuses also for #tax avoidance #IRS #OpenAI #Anthropi… | ↗Not briefed |
bluesky bluesky-ai | 17d | We had an opt-in policy: it was our existing body of law against theft and copyright infringement. Then Google, Microsoft, OpenAI, Anthropi… | We had an opt-in policy: it was our existing body of law against theft and copyright infringement. Then Google, Microsoft, OpenAI, Ant… | ↗Not briefed |
arxiv arxiv-ai | 17d | DISCERN: Distributional Synthetic Replay for Retain-Free Incremental Machine Unlearning | Machine unlearning aims to remove the influence of specific training data from a trained model without retraining from scratch. In rea… | ↗Not briefed |
arxiv arxiv-ai | 17d | When Bigger Becomes Brittle: Why Compression Hurts More as Language Models Scale | Quantization and pruning reduce the cost of deploying language models, but their effects are usually reported as benchmark scores. Suc… | ↗Not briefed |
hn hn-ai | 17d | OpenAI kicks off charm offensive to change voters' view of data centers | — | ↗Not briefed |
hn hn-ai | 17d | Marvell to buy Lumentum in $115B all-stock deal | — | ↗Not briefed |
bluesky bluesky-ai | 17d | Hideo Kojima says AI lacks what makes the creative process fun. As an example he brings up model kits or Lego, the final product isn't the … | Hideo Kojima says AI lacks what makes the creative process fun. As an example he brings up model kits or Lego, the final product isn't… | ↗Not briefed |
bluesky bluesky-ai | 17d | i’m learning a lot about this acronym today. like it was coined by jim cramer??? and he recently updated it to MANGOS (meta, anthropic, nvi… | i’m learning a lot about this acronym today. like it was coined by jim cramer??? and he recently updated it to MANGOS (meta, anthropic… | ↗Not briefed |
arxiv arxiv-ai | 17d | When, Where, How: Active In-Context Learning for Tabular Foundation Models | Tabular Foundation Models (TFMs) adapt to a new task only through the labeled rows in their context. However, labels are expensive to… | ↗Not briefed |
arxiv arxiv-ai | 17d | InnerLogic-Surg: An Open Surgical Foundation Model via Continual Pretraining | Vision foundation models are the backbone of modern surgical image analysis, yet most surgical systems still rely on encoders pretrain… | ↗Not briefed |
rss techmeme | 17d | Sources: Micron suspends construction of its $3B to $4.5B Fab 11E facility in Taiwan, set for 2028, due to mounting uncertainties over a local labor dispute (Bloomberg) | Bloomberg: Sources: Micron suspends construction of its $3B to $4.5B Fab 11E facility in Taiwan, set for 2028, due to mounting uncerta… | ↗Not briefed |
bluesky bluesky-ai | 17d | まだ帰らないの…? …えへっ あ、おつかれさま♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #小説 #… | まだ帰らないの…? …えへっ あ、おつかれさま♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #… | ↗Not briefed |
bluesky bluesky-ai | 17d | Related, you can't get good information if the information environment is poisoned by AI-slop misinformation. Fighting back against LLM gen… | Related, you can't get good information if the information environment is poisoned by AI-slop misinformation. Fighting back against LL… | ↗Not briefed |
arxiv arxiv-ai | 17d | Network Effect Decomposition for Causal Effect Estimation under Network Interference | Estimating causal effects from observational network data requires accounting for both individual treatment and neighborhood exposure.… | ↗Not briefed |
rss techcrunch-ai | 17d | Agility is the first pure humanoid robot company on the stock market. Are you buying? | Agility arrives on the Nasdaq today as the first publicly-traded pure-play humanoid robotics company. | ↗Not briefed |
rss wired-ai | 17d | AI’s Next Trick? Speaking Dolphin | Denise Herzing, the founder and director of the Wild Dolphin Project, is using AI to help understand audio and video footage of the ma… | ↗Not briefed |
hn hn-ai | 17d | Show HN: OOO, but you decide what gets recommended | Hi HN!<p>I recently saw HackerTube, which Andrej Karpathy posted here, and really liked it. Instead of YouTube’s algorithm, it uses vi… | ↗Not briefed |
arxiv arxiv-ai | 17d | Which Signals of Code Correctness Survive Distribution Shift? Hidden-State and Causal Predictors of Program and Verdict Correctness | Large language models increasingly judge code, yet without external feedback such judgments are unreliable, and verifying them by exec… | ↗Not briefed |
hn hn-ai | 17d | Show HN: Hyperbutler – Docker on Windows 11 Without WSL or Docker Desktop | — | ↗Not briefed |
hn hn-ai | 17d | Show HN: Vivonco – match patient's cancer reports to targeted drugs and trials | Hi HN. I made a lightweight app where you can add patient's cancer reports (pathology, scans, gene tests, even phone photos) and… | ↗Not briefed |
hn hn-ai | 17d | LLM pricing agents can converge to different price regimes | — | ↗Not briefed |
arxiv arxiv-ai | 17d | Responsible GenAI Use During Active Armed Conflict: Evidence from Sudan's Emergency Response Rooms | Armed conflict can increase the value of Generative AI (GenAI) while undermining the conditions required for its responsible use. Yet… | ↗Not briefed |
rss techmeme | 17d | Amazon partners with Uber to make Uber Eats the "official" US restaurant delivery partner for Amazon Prime subscribers, after previously partnering with Grubhub (Andrew J. Hawkins/The Verge) | Andrew J. Hawkins / The Verge: Amazon partners with Uber to make Uber Eats the “official” US restaurant delivery partner for Amazon Pr… | ↗Not briefed |
bluesky bluesky-ai | 17d | Heh, burde ha sagt mig selv at det allerede findes.. Øko-venlige/etiske AI-udbydere: www.greenpt.com www.ecoia.ai basebox.ai | Heh, burde ha sagt mig selv at det allerede findes.. Øko-venlige/etiske AI-udbydere: www.greenpt.com www.ecoia.ai basebox.ai | ↗Not briefed |
bluesky bluesky-ai | 17d | OpenAI, Anthropic, and SpaceXSI have all reached out to Husk, an influencer who regularly goes viral by exposing the shortcomings of chatbo… | OpenAI, Anthropic, and SpaceXSI have all reached out to Husk, an influencer who regularly goes viral by exposing the shortcomings of c… | ↗Not briefed |
arxiv arxiv-ai | 17d | FARR: Forecast-Aware Residual Refinement with Matched Endpoint Evidence for Long-Horizon Time-Series Forecasting | Large language models use auxiliary conditioning to incorporate additional information into time-series forecasts. However, auxiliary… | ↗Not briefed |
arxiv arxiv-ai | 17d | PlayGen2: Learning Real-Time Playable Worlds with Explicit 3D State | Building a playable world requires more than generating plausible video: it demands responsive control, persistent spatial structure,… | ↗Not briefed |
bluesky bluesky-ai | 17d | I decided to write a bit about the Amiga LLM-ish port I did. An Amiga WHDLoad installer for the Atari ST version of "Get Dexter!" gemini:/… | I decided to write a bit about the Amiga LLM-ish port I did. An Amiga WHDLoad installer for the Atari ST version of "Get Dexter!" gem… | ↗Not briefed |
rss wired-ai | 17d | He Trolls AI Labs for a Living. Now They All Want to Work With Him | OpenAI, Anthropic, and SpaceXSI have all reached out to Husk, an influencer who regularly goes viral by exposing the shortcomings of c… | ↗Not briefed |
arxiv arxiv-ai | 17d | Do Recommenders Need to Think in Words? Latent Reasoning for Efficient LLM-Based Recommendation | LLM-based recommender systems increasingly generate chain-of-thought (CoT) text before each item prediction to activate the LLM's reas… | ↗Not briefed |
bluesky bluesky-ai | 17d | Hopefully the 'l33t coders' who are mashing queries into LLM coding bots like Claude to do fancy projects have a backup plan for when the V… | Hopefully the 'l33t coders' who are mashing queries into LLM coding bots like Claude to do fancy projects have a backup plan for when… | ↗Not briefed |
rss techmeme | 17d | Study: UK consumer AI spending passes £1B annually, with 5.1% of adults paying for subscriptions as of July 2026; Claude is rapidly catching up to ChatGPT (Bloomberg)🔧 ChatGPT | Bloomberg: Study: UK consumer AI spending passes £1B annually, with 5.1% of adults paying for subscriptions as of July 2026; Claude is… | ↗Not briefed |
hn hn-ai | 17d | Overlooked Linux Mv Option Replaces Three Commands with One | — | ↗Not briefed |
arxiv arxiv-ai | 17d | CORE: Coarse-to-Fine Search with Context Management for Deep Research | Deep research agents answer complex questions through repeated search. They rely on a harness that runs their tool calls and returns e… | ↗Not briefed |