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 | 28d | Show HN: Model Atlas – An LLM leaderboard you can reweight | — | ↗Not briefed |
arxiv arxiv-ai | 28d | The Physics Is the Network: Fast, Training-Free Neural Emulation | Neural emulation need not require learning a substitute for known physics. We investigate PolyNets that couple a continuous-time neura… | ↗Not briefed |
bluesky bluesky-ai | 28d | Amazing sentence: "OpenAI lawyers urge the Florida court to sidestep the complex issues, arguing that in this instance, there’s no need f… | Amazing sentence: "OpenAI lawyers urge the Florida court to sidestep the complex issues, arguing that in this instance, there’s no n… | ↗Not briefed |
bluesky bluesky-ai | 28d | Should news organizations accept philanthropic funding from OpenAI? @dicktofel.bsky.social says maybe not. I say definitely not. There are … | Should news organizations accept philanthropic funding from OpenAI? @dicktofel.bsky.social says maybe not. I say definitely not. There… | ↗Not briefed |
hn hn-ai | 28d | Someone on Reddit is running Qwen 27B at ~20 tok/s on a jailbroken PS5's GPU | — | ↗Not briefed |
arxiv arxiv-ai | 28d | The Wealth of Agents: Scaling Agentic AI Across Developers | Today's agentic systems are typically designed, built, and deployed by single developers with full control over all participating agen… | ↗Not briefed |
bluesky bluesky-ai | 28d | Also lol at SQL, get tf out of here, mate, if you use a LLM to write SQL (or any code, really) for you, you have no place nearby a keyboard. | Also lol at SQL, get tf out of here, mate, if you use a LLM to write SQL (or any code, really) for you, you have no place nearby a key… | ↗Not briefed |
arxiv arxiv-ai | 28d | VASPO: Self-Distillation with Privileged Value Estimation | On-policy self-distillation provides dense supervision for reasoning models via a privileged self-teacher. However, this supervision r… | ↗Not briefed |
arxiv arxiv-ai | 28d | Bayesian Learning to Select Fair Classifiers | Evaluations of classifiers (AI or human or agents) often assume a batch setting in which every agent produces labels on every example… | ↗Not briefed |
hn hn-ai | 28d | Show HN: botsay – cowsay for the AI Era | Does this need to exist?<p>Probably not.<p>Does it add whimsy to the sterile, cold terminal environment that has become our chosen hom… | ↗Not briefed |
bluesky bluesky-ai | 28d | I am the world's worst shopkeeper, but today's news from Anthropic makes it an *excellent* time to invest in a Just enough Internet tea tow… | I am the world's worst shopkeeper, but today's news from Anthropic makes it an *excellent* time to invest in a Just enough Internet te… | ↗Not briefed |
bluesky bluesky-ai | 28d | New term: “diarrhea engineering/programming” For when you have something that runs like shit, but now, well, it *still* runs like shit, b… | New term: “diarrhea engineering/programming” For when you have something that runs like shit, but now, well, it *still* runs like sh… | ↗Not briefed |
hn hn-ai | 28d | Rapl: Enjoyable, composable, hackable N-dimensional arrays for Rust | — | ↗Not briefed |
arxiv arxiv-ai | 28d | A Two-Level Clinical Novelty Prioritization Framework for Disease Trajectories Discovered from Longitudinal Electronic Health Records | Graph-based analysis of electronic health records (EHRs) yields disease-trajectory catalogs, but structural support does not establish… | ↗Not briefed |
bluesky bluesky-ai | 28d | "I've never had a relationship with an AI model in that way. That was... until I downloaded Muse." | "I've never had a relationship with an AI model in that way. That was... until I downloaded Muse." | ↗Not briefed |
bluesky bluesky-ai | 28d | separate from my big thread, can you like. not use an LLM to make graphs of your data or if you are just at least make it not look like slo… | separate from my big thread, can you like. not use an LLM to make graphs of your data or if you are just at least make it not look lik… | ↗Not briefed |
bluesky bluesky-ai | 28d | Anthropic could never make something as glorious as a slime mold | Anthropic could never make something as glorious as a slime mold | ↗Not briefed |
hn hn-ai | 28d | HN: RPDio Hyper – 120fps desktop streaming in Rust | — | ↗Not briefed |
arxiv arxiv-ai | 28d | VascuSR: Iterative Anatomy-Aware Diffusion for CT Super-Resolution | Generative super-resolution (SR) can recover sharp detail in computed tomography (CT) images from severely downsampled observations, b… | ↗Not briefed |
bluesky bluesky-ai | 28d | Watching a podcast with a former OpenAI safety researcher and TIL that AI models are multilingual. Seems obvious given they’re trained in … | Watching a podcast with a former OpenAI safety researcher and TIL that AI models are multilingual. Seems obvious given they’re traine… | ↗Not briefed |
arxiv arxiv-ai | 28d | ReaDy: Recurrent States as Dynamic Context for Speculative Drafting in Hybrid Linear Attention Large Language Models | Speculative decoding accelerates autoregressive generation by drafting multiple tokens per step, yet feature-based drafters rely on ta… | ↗Not briefed |
arxiv arxiv-ai | 28d | Graph-Enhanced Disease Trajectory Discovery and Temporal Endpoint Forecasting from Longitudinal Electronic Health Records | Longitudinal electronic health records (EHRs) can reveal disease-progression structure, while useful forecasting requires leakage-cont… | ↗Not briefed |
hn hn-ai | 28d | He linked Saudi Arabia to SV. Then OpenAI found a bug wrapped in toilet paper | — | ↗Not briefed |
bluesky bluesky-ai | 28d | あと、当たり前ですが、パランティアのMavenなど、実際の戦争兵器として利用されているLLMに関しては、提供者と利用者に重大な道義的・法的責任があることは言うまでもありません。 | あと、当たり前ですが、パランティアのMavenなど、実際の戦争兵器として利用されているLLMに関しては、提供者と利用者に重大な道義的・法的責任があることは言うまでもありません。 | ↗Not briefed |
arxiv arxiv-ai | 28d | Hindsight-Privileged RLVR: Single-Rollout Token Credit Without Group Sampling | Reinforcement learning with verifiable rewards (RLVR) trains language models from checked task outcomes, but assigning a terminal rewa… | ↗Not briefed |
hn hn-ai | 28d | Almost everyone misunderstands Chrome extension manifest version 3 | — | ↗Not briefed |
bluesky bluesky-ai | 28d | Another banger is “I can’t eat output of the LLM” well ok I don’t eat meat but I do type stuff into the box and I’m getting paid so in a be… | Another banger is “I can’t eat output of the LLM” well ok I don’t eat meat but I do type stuff into the box and I’m getting paid so in… | ↗Not briefed |
bluesky bluesky-ai | 28d | Think a lot of other fields could learn from the thoughtful ways in which mathematicians are processing the changes being wrought by AI. … | Think a lot of other fields could learn from the thoughtful ways in which mathematicians are processing the changes being wrought by A… | ↗Not briefed |
bluesky bluesky-ai | 28d | The former employee took more than 100 confidential files before leaving to join Apple, OpenAI alleges. www.theverge.com/ai-artificia... | The former employee took more than 100 confidential files before leaving to join Apple, OpenAI alleges. www.theverge.com/ai-artificia.… | ↗Not briefed |
hn hn-ai | 28d | Show HN: A faster WordPress core written in C/C++ (Live Demo) | Over the last 20 years, I repeatedly got frustrated with how slow WordPress (WP) is, especially once you stack a few plugins on it.<p>… | ↗Not briefed |
arxiv arxiv-ai | 28d | Quantifying privacy risks in foundation model representations of brain MRI | Pretrained representations are increasingly used to share and reuse information from medical images without exposing the original scan… | ↗Not briefed |
bluesky bluesky-ai | 28d | 【注意：立場表明】 私は生成AI（LLM）企業の環境負荷への責任回避的な態度や、データセットの著作権上の問題を強く批判していますが、LLM技術の利用そのものを全面的に否定しているわけではありません。実際に医療・福祉・障害者支援等での活用可能性についても肯定的に論じています。 … | 【注意：立場表明】 私は生成AI（LLM）企業の環境負荷への責任回避的な態度や、データセットの著作権上の問題を強く批判していますが、LLM技術の利用そのものを全面的に否定しているわけではありません。実際に医療・福祉・障害者支援等での活用可能性についても肯定的に論じてい… | ↗Not briefed |
bluesky bluesky-ai | 28d | email match is the usual culprit: if your openai account email matches your github profile email, the dots build that account-graph link an… | email match is the usual culprit: if your openai account email matches your github profile email, the dots build that account-graph li… | ↗Not briefed |
rss theverge-ai | 28d | My AI doppelganger tells me what to eat | This is Optimizer, a weekly newsletter sent from Verge senior reviewer Victoria Song that dissects and discusses the latest gizmos and… | ↗Not briefed |
hn hn-ai | 28d | The G = 9 Gate – Fermion-Boson Transitions and Dark Matter | — | ↗Not briefed |
bluesky bluesky-ai | 28d | it *is* making software more secure, though: winbuzzer.com/2026/07/06/a... | it *is* making software more secure, though: winbuzzer.com/2026/07/06/a... | ↗Not briefed |
hn hn-ai | 28d | GPT-2 on the GPU | — | ↗Not briefed |
bluesky bluesky-ai | 28d | My unpopular opinion is that I think AI assisted coding is not equivalent to "AI art". Coders have been using LLMs for assistance for like … | My unpopular opinion is that I think AI assisted coding is not equivalent to "AI art". Coders have been using LLMs for assistance for… | ↗Not briefed |
arxiv arxiv-ai | 28d | Average-Reward Policy Identification from a Single Continuing Trajectory | In reinforcement learning, an agent must explore an unknown environment to learn how to make good decisions. Unlike methods that can f… | ↗Not briefed |
hn hn-ai | 28d | Separating HBM from GPUs to Expand Memory Capacity; Optical Links May Be Key | — | ↗Not briefed |
arxiv arxiv-ai | 28d | Decoupled Marginal Sharpening for Training-Free Inference-Time Scaling | We study how to improve LLM reasoning at inference time without updating model weights or using external reward models. Recent reward-… | ↗Not briefed |
hn hn-ai | 28d | Linus Torvalds's first LLM-assisted patch series | — | ↗Not briefed |
arxiv arxiv-ai | 28d | Lost in Translation: Why Lower Bias Scores Don't Mean Fairer Models | Bias audits of multilingual language models usually compare a model's average stereotype score in English with its average in another… | ↗Not briefed |
arxiv arxiv-ai | 28d | AdaRSI: On-Demand Recursive Self-Improvement for Embodied Agents | Pretrained vision-language-action models (VLAs) often fail midway through unseen long-horizon tasks. Recursive self-improvement (RSI)… | ↗Not briefed |
bluesky bluesky-ai | 28d | I know from very reliable sources that OpenAI are using human translators for their own content, too. Just saying. | I know from very reliable sources that OpenAI are using human translators for their own content, too. Just saying. | ↗Not briefed |
arxiv arxiv-ai | 28d | SafeAdapt-Poker: Counterfactual Risk Calibration for Budgeted Adaptation under Opponent Shift | Opponent histories can mislead after behavioral shifts. SafeAdapt-Poker formulates budgeted adaptation under opponent shift as a dual-… | ↗Not briefed |
arxiv arxiv-ai | 28d | ChronoCulture: Evaluating Cultural Awareness Across Time in Large Language Models | Cultural attitudes and social norms evolve over time, yet most cultural benchmarks for large language models (LLMs) treat culture as a… | ↗Not briefed |
hn hn-ai | 28d | Metagenomic testing: technique that finds almost every infection, even unknown | — | ↗Not briefed |
bluesky bluesky-ai | 28d | Is epic trying to pivot from making and selling games to being a marketing front for Anthropic or something? Why are they doing all this? | Is epic trying to pivot from making and selling games to being a marketing front for Anthropic or something? Why are they doing all th… | ↗Not briefed |
hn hn-ai | 28d | OpenAI: Employee who left for Apple took files about unreleased device | — | ↗Not briefed |