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 |
|---|---|---|---|---|
bluesky bluesky-ai | 29d | The AI isn’t going to be able to run on however much RAM is in your refrigerator and IoT botnets are absolutely nothing new. This feels lik… | The AI isn’t going to be able to run on however much RAM is in your refrigerator and IoT botnets are absolutely nothing new. This feel… | ↗Not briefed |
hn hn-ai | 29d | Creatine might protect the brain from gut inflammation | — | ↗Not briefed |
rss the-decoder | 29d | Gemini 4 Argon is now available worldwide as Google's most capable frontier model | Google is making Gemini 4 Argon available worldwide through its API, the Gemini app, and Google Antigravity. The frontier model is agg… | ↗Not briefed |
bluesky bluesky-ai | 29d | If anyone’s freaking out, I break down a lot of the smokescreen in this chain. tl;dr: Anthropic jailbroke a known vulnerable android tv, a… | If anyone’s freaking out, I break down a lot of the smokescreen in this chain. tl;dr: Anthropic jailbroke a known vulnerable android… | ↗Not briefed |
rss theverge-ai | 29d | I don’t miss the camera on the Ray-Ban Meta Audio glasses | The other day I whizzed past several people while running in the park, my playlist blasting at full volume. The next morning, I smiled… | ↗Not briefed |
bluesky bluesky-ai | 29d | Quick rundown of CAI's many scandals: | Quick rundown of CAI's many scandals: | ↗Not briefed |
arxiv arxiv-ai | 29d | ChunkAct: Action Chunking for Agentic Robotics | Agentic robotics places a large language model (LLM) agent in the robot control loop, where it composes perception, motion, and vision… | ↗Not briefed |
arxiv arxiv-ai | 29d | Cepheid: Decoupling Scale from Structure for Cross-Family Motion Completion | Sparse-to-dense motion completion recovers dense mesh motion from sparse vertex displacements. Residual priors trained on one geometry… | ↗Not briefed |
arxiv arxiv-ai | 29d | What Matters in Quality-Aware VLA Pretraining: A Benchmark and Empirical Study | How should demonstration quality be used during vision-language-action (VLA) pretraining to improve performance after target-task fine… | ↗Not briefed |
arxiv arxiv-ai | 29d | Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards | Long-horizon agents have drawn growing attention for their ability to work autonomously -- from hours to days -- on economically valua… | ↗Not briefed |
arxiv arxiv-ai | 29d | Data Free Training of Diffusion Models for Lattice Field Theories | Sampling equilibrium field configurations remains a major computational bottleneck in lattice field theory, motivating the development… | ↗Not briefed |
arxiv arxiv-ai | 29d | DoubtLess: Training-Free Contrastive Decoding for Efficient Reasoning | Reasoning models achieve strong performance through long chains of thought (CoT), but often spend more tokens than needed to solve a p… | ↗Not briefed |
arxiv arxiv-ai | 29d | AesCode: Aesthetic Code Generation with Decoupled Cross-Modal Rewards🔧 Modal🔧 GPT-5 | Generating information-rich visual artifacts such as slides, posters, and dashboards as code is appealing because the outputs are stru… | ↗Not briefed |
arxiv arxiv-ai | 29d | Hybrid Reasoning Agents | How can we build an agent that thinks beyond human language, yet still acts through it in the external environment? Modern LLM agents… | ↗Not briefed |
arxiv arxiv-ai | 29d | Exact Topological Compliance: Generating Persistence-Equivalent Graphs | Persistent homology summarizes the topological evolution of a graph as persistence diagrams (PDs). Generating graphs whose topology ma… | ↗Not briefed |
arxiv arxiv-ai | 29d | Two Clocks: Why Larger Rollout Batches Learn Slower but Fail on Schedule in Reinforcement Learning with Verifiable Rewards | Reinforcement learning with verifiable rewards (RLVR) relies on costly model-generated responses, motivating larger rollout batches an… | ↗Not briefed |
hn hn-ai | 29d | MountX: A self-hosted dynamic language compiling to C99 | — | ↗Not briefed |
arxiv arxiv-ai | 29d | MFG: Multi-turn Feedback Guided Molecular Optimization | Multi-objective molecular optimization is a critical process in drug discovery that aims to modify a given molecule to simultaneously… | ↗Not briefed |
arxiv arxiv-ai | 29d | The Last Remote Sensing Model Built by Humans? Casting Any Earth Observation Task as Recursive Self-Improvement | Earth observation (EO) models are still built by hand. For each new task, experts design a training recipe and tune it against a bench… | ↗Not briefed |
arxiv arxiv-ai | 29d | LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models | Robot policies are expected to generalize across deployment environments, including variations in the robot itself: cameras are bumped… | ↗Not briefed |
arxiv arxiv-ai | 29d | Who Improves the Robot? Toward Physical Recursive Self-Improvement | Robots are increasingly autonomous in task execution, yet their improvement remains largely human-driven. This study investigates Phys… | ↗Not briefed |
hn hn-ai | 29d | As Federal Court Calls ALPRs Mass Surveillance, EFF Gets Disappointing Ruling | — | ↗Not briefed |
arxiv arxiv-ai | 29d | Fast Rates for Offline Q-Learning with KL Regularization | This paper investigates offline Q-learning for finite-horizon Markov decision processes with Kullback-Leibler (KL) regularization. Rec… | ↗Not briefed |
arxiv arxiv-ai | 29d | Scaling Online Optimal Transport Minibatches for Flow Matching | A central challenge in scaling flow models is maintaining stability in training and integration. In flow matching, this stability depe… | ↗Not briefed |
bluesky bluesky-ai | 29d | This afternoon I had to remind an Anthropic model that 30+45 is not 3045. No, they are not months away from hijacking my Alexa or something. | This afternoon I had to remind an Anthropic model that 30+45 is not 3045. No, they are not months away from hijacking my Alexa or some… | ↗Not briefed |
rss techmeme | 29d | Sources: Decagon, which offers AI customer service agents and was last valued at $4.5B, hired a bank and is in early talks with buyers, including ServiceNow (Katie Roof/Business Insider) | Katie Roof / Business Insider: Sources: Decagon, which offers AI customer service agents and was last valued at $4.5B, hired a bank an… | ↗Not briefed |
arxiv arxiv-ai | 29d | RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks | It is difficult to accurately model the physics of the interaction between a robot and the objects it manipulates, including contact m… | ↗Not briefed |
bluesky bluesky-ai | 29d | Good! Let the dominoes fall! The graph in the article is interesting. Character AI had more users spending time on it than ChatGPT, DeepSe…🔧 ChatGPT | Good! Let the dominoes fall! The graph in the article is interesting. Character AI had more users spending time on it than ChatGPT, D… | ↗Not briefed |
arxiv arxiv-ai | 29d | OmniCompass: A Holistic Benchmark for Omni-Modal Video Grounding and Understanding🔧 Modal | With the rapid development of multimodal large language models (MLLMs), evaluating their video understanding capabilities has become i… | ↗Not briefed |
bluesky bluesky-ai | 29d | hey anthropic where's my check. you just keep doing crimes but you haven't sent me my check for stealing my work to do crimes. I want my ch… | hey anthropic where's my check. you just keep doing crimes but you haven't sent me my check for stealing my work to do crimes. I want… | ↗Not briefed |
arxiv arxiv-ai | 29d | Joint Neural Posterior Estimation and Model Selection for Chemical Reaction Networks | Simulation-based Bayesian inference for Chemical Reaction Networks (CRNs) must simultaneously solve two computationally demanding prob… | ↗Not briefed |
arxiv arxiv-ai | 29d | Read the Room Before Your Model Generates | Large language models (LLMs) generally rely on sampling temperature to regulate the amount of randomness in their outputs. Mainstream… | ↗Not briefed |
hn hn-ai | 29d | Home and Factory Robots | — | ↗Not briefed |
hn hn-ai | 29d | Show HN: Hev is a Jev-compatible API, answered by humans | Burned a few tokens during the long weekend to give you real decisions powered by the most trustworthy humans on the internet: strange… | ↗Not briefed |
arxiv arxiv-ai | 29d | Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs | Pre-trained Large Language Model (LLMs) are effective on many tasks in zero-shot and few-shot settings. For machine translation of low… | ↗Not briefed |
rss techmeme | 29d | Pluto Security, which monitors AI agents and vibe-coded apps, exits stealth with $42M, including a $34M Series A, and says it has hit $2M+ in ARR since April (Ram Iyer/TechCrunch) | Ram Iyer / TechCrunch: Pluto Security, which monitors AI agents and vibe-coded apps, exits stealth with $42M, including a $34M Series… | ↗Not briefed |
arxiv arxiv-ai | 29d | Self-Evolving Agent Harness | Harness self-evolution revises an agent's software scaffolding, including prompts, memory, workflows, and executable components, while… | ↗Not briefed |
bluesky bluesky-ai | 29d | I see Anthropic reinvented Mononoke | I see Anthropic reinvented Mononoke | ↗Not briefed |
bluesky bluesky-ai | 29d | The most indebted corporations in America right now are, at their core, computer virus companies. They think the computer viruses are their… | The most indebted corporations in America right now are, at their core, computer virus companies. They think the computer viruses are… | ↗Not briefed |
arxiv arxiv-ai | 29d | Statistical Vignettes on the Grokking Phenomenon | In this work, we study grokking in a simple, analytically tractable setting whose rich phenomenology allows us to disentangle the effe… | ↗Not briefed |
bluesky bluesky-ai | 29d | some day they will invent an LLM that understands brevity | some day they will invent an LLM that understands brevity | ↗Not briefed |
hn hn-ai | 29d | Puter's Mission to Open-Source Everything | — | ↗Not briefed |
arxiv arxiv-ai | 29d | Depth Memory | Residual connections propagate information through a network by additively accumulating layer outputs. Recently proposed alternatives,… | ↗Not briefed |
arxiv arxiv-ai | 29d | Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning | In cooperative Multi-Agent Reinforcement Learning (MARL), agents must learn to coordinate their actions to achieve shared objectives.… | ↗Not briefed |
arxiv arxiv-ai | 29d | SkillScaffold: Scaffolding Agentic Reinforcement Learning with Policy-Adaptive Skill Credit Assignment and Selective Injection | Agentic reinforcement learning (RL) enables large language models (LLMs) to learn from multi-turn interactions with verifiable outcome… | ↗Not briefed |
bluesky bluesky-ai | 29d | "What I think is most important to understand about that time is that the LLM released by OpenAI had essentially nothing to do with what th… | "What I think is most important to understand about that time is that the LLM released by OpenAI had essentially nothing to do with wh… | ↗Not briefed |
arxiv arxiv-ai | 29d | Registration is a Representation Problem: Solving Whole-Body PET/CT Registration | Raw image intensities are a poor surrogate for similarity across images in correspondence problems. Further, while auxiliary segmentat… | ↗Not briefed |
bluesky bluesky-ai | 29d | Literally "look at us, guys, we're doing this!" www-cnbc-com.cdn.ampproject.org/c/s/www.cnbc... | Literally "look at us, guys, we're doing this!" www-cnbc-com.cdn.ampproject.org/c/s/www.cnbc... | ↗Not briefed |
bluesky bluesky-ai | 29d | For years, researchers have described how generative AI would change influence ops. But something more interesting is happening now: AI is … | For years, researchers have described how generative AI would change influence ops. But something more interesting is happening now: A… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Pippin – macOS High-Performance Screen Sharing Client for Linux | Hi HN, I built pippin, a native Rust client for macOS's "High-Performance" Screen Sharing mode on Linux. There are lots… | ↗Not briefed |