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 | 2d | Here's a cute little starter home you can buy in San Mateo County! It even got a price cut today! Buy it today before the Anthropic folks s… | Here's a cute little starter home you can buy in San Mateo County! It even got a price cut today! Buy it today before the Anthropic fo… | ↗Not briefed |
hn hn-ai | 2d | Optimizing AI Agents: Prompts vs. Weights (GEPA, GRPO) – Stanford CS329Z [video] | — | ↗Not briefed |
arxiv arxiv-ai | 2d | Kernel Ridge Attention: Nonlinear Memory with Selective Forgetting | Kernel Ridge Attention (KRA) is a recurrent sequence mixer that combines learned nonlinear features with covariance-guided memory upda… | ↗Not briefed |
bluesky bluesky-ai | 2d | Talking with my parents and their friends recently made me realize the people using AI for everything are not the ones we'd expect. You can…🔧 ChatGPT | Talking with my parents and their friends recently made me realize the people using AI for everything are not the ones we'd expect. Yo… | ↗Not briefed |
arxiv arxiv-ai | 2d | SixthView: A Benchmark for Hidden-View Reconstruction and 3D Assembly Reasoning | Inferring the unseen side of a 3D assembly requires reasoning about part placement and occlusion. Different assemblies can match the o… | ↗Not briefed |
arxiv arxiv-ai | 2d | EVA-RSI: Recursive Self-Improvement of Video Evaluators for Judging Temporal Quality | Video evaluators play an important role in video generation, yet they miss motion defects, limiting the quality of generated videos. R… | ↗Not briefed |
rss techmeme | 2d | Microsoft makes Word, Excel, and PowerPoint generally available inside Copilot, which CEO Satya Nadella calls "the new OS for work" (Nick Indge/Ars Technica) | Nick Indge / Ars Technica: Microsoft makes Word, Excel, and PowerPoint generally available inside Copilot, which CEO Satya Nadella cal… | ↗Not briefed |
arxiv arxiv-ai | 2d | Risk-Controlled Correction for Frozen Multimodal Models under Missing Modalities | A deployed multimodal predictor may face missing modalities after its weights can no longer be changed. This raises a question that tr… | ↗Not briefed |
hn hn-ai | 2d | Ask HN: Can world models lead to racisim? Are there any Alignment frameworks? | I have never felt or believed in Dooms day theories about AI. But below is one of only 2 occasions where I am feeling AI can be too da… | ↗Not briefed |
arxiv arxiv-ai | 2d | Accelerating Distributed Digital Ising Machines | Ising machines, dedicated accelerators built around the dynamics of Ising-like spin models, have emerged as a leading platform for com… | ↗Not briefed |
hn hn-ai | 2d | ABFS: Another Build File Service | — | ↗Not briefed |
hn hn-ai | 2d | SiliconForest: Darwin/XNU on Apple silicon | — | ↗Not briefed |
arxiv arxiv-ai | 2d | NullTok: empirical-null token selection for training-free long video understanding | Long video understanding remains a daunting challenge for existing multimodal large language models (MLLMs), mainly due to the large n… | ↗Not briefed |
arxiv arxiv-ai | 2d | A Rank-One Bypass of Fixed Linear Erasure | Linear concept erasure can make a target unreadable to a specified family of linear classifiers applied to a model's current internal… | ↗Not briefed |
arxiv arxiv-ai | 2d | Infinity-Wasserstein Distributionally Robust Principal Component Analysis: Exact Reformulation, Finite-Sample Guarantee, and Efficient Algorithm | Principal component analysis (PCA) is a fundamental and ubiquitous technique for dimension reduction and representation learning, yet… | ↗Not briefed |
arxiv arxiv-ai | 2d | In Situ Training of Optical Neural Networks by Coding Agents | Optical neural networks promise fast, energy-efficient computing, but every device departs from its design and must be trained with th… | ↗Not briefed |
bluesky bluesky-ai | 2d | This is going to upset you! Stop clicking "Agree" when a website asks! This is a threat to democracy! youtu.be/82Xv_bPVtpU?... | This is going to upset you! Stop clicking "Agree" when a website asks! This is a threat to democracy! youtu.be/82Xv_bPVtpU?... | ↗Not briefed |
hn hn-ai | 2d | Intellij-Elm – Elm Plugin for IntelliJ Platform IDEs | — | ↗Not briefed |
hn hn-ai | 2d | Optimjack from the Outside: The Hijacked Script Was on Our Radar | — | ↗Not briefed |
bluesky bluesky-ai | 2d | Some of the issues I’m seeing could maybe be resolved with a more structured and considered approach to using an LLM—and actually looking a… | Some of the issues I’m seeing could maybe be resolved with a more structured and considered approach to using an LLM—and actually look… | ↗Not briefed |
arxiv arxiv-ai | 2d | Asking but Missing: Decision Coverage in Natural-Language Data Querying | Large language models are increasingly used to translate natural language questions about structured data into executable programs. A… | ↗Not briefed |
arxiv arxiv-ai | 2d | Grounding Radiology Reports in 3D Chest CT: Results of the ReXGrounding Challenge | Segmenting each reported finding in a radiology scan would allow findings to be measured and tracked over time and let AI systems gene… | ↗Not briefed |
arxiv arxiv-ai | 2d | LAGR: Language-Aware Grasp Re-Ranking for Task Oriented Grasping | Grasping is a fundamental capability for robotic manipulation, with the appropriate grasp depending on the downstream task. For exampl… | ↗Not briefed |
hn hn-ai | 2d | Btrfs Seeks to Deal with AI and Newbies Without Common Sense | — | ↗Not briefed |
arxiv arxiv-ai | 2d | Neural Memory Orchestration | Modern HPC workloads increasingly stall on the memory hierarchy rather than on compute, and page faults from mmap'd files, swap, and t… | ↗Not briefed |
hn hn-ai | 2d | Inside the $900B data centre bet behind Anthropic's IPO | — | ↗Not briefed |
arxiv arxiv-ai | 2d | Beyond Cross-Modal Reconstruction: Participant-Level Evaluation of sEMG-Kinematic Coupling in Post-Stroke Gait🔧 Modal | Accurate cross-modal reconstruction is often treated as evidence of physiological coupling, although it may simply reflect temporal pa… | ↗Not briefed |
arxiv arxiv-ai | 2d | RoboBuilder: Agentic Robot Co-Design with Real-world Video Feedback in the Loop | Can AI agents design robots that work in the real world? We introduce RoboBuilder, a benchmark in which coding agents co-design a robo… | ↗Not briefed |
bluesky bluesky-ai | 2d | もう一回やってみよっ♪ 頑張りは報われるよ♡ 【 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 #AI… | もう一回やってみよっ♪ 頑張りは報われるよ♡ 【 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… | ↗Not briefed |
rss wired-ai | 2d | Over a Third of Gen Z-ers Rely on AI for Friendship | Young people who have online-only friends or use AI as a social companion face higher odds of loneliness than those who do not, a new… | ↗Not briefed |
arxiv arxiv-ai | 2d | A Fast, Programmable GPU-to-CPU Invocation Layer for GPU-Initiated Communication🔧 vLLM | Modern GPU workloads increasingly rely on device-initiated communication to let GPU kernels communicate directly without synchronizing… | ↗Not briefed |
arxiv arxiv-ai | 2d | ReflectScene:MLLM-Guided Closed-Loop Diffusion for 3D Indoor Scene Generation🔧 Modal | Text-conditioned 3D indoor scene generation requires layouts satisfying geometric, functional, and semantic constraints. However, exis… | ↗Not briefed |
rss techmeme | 2d | YouTube is rolling out an update to include "co-views", or the estimated number of people watching together on TV, in the public view counts for videos (Emma Roth/The Verge) | Emma Roth / The Verge: YouTube is rolling out an update to include “co-views”, or the estimated number of people watching together on… | ↗Not briefed |
arxiv arxiv-ai | 2d | RePlan-VLA: Progress-Aware Replanning for Long-Horizon Vision-Language-Action Models | Vision-Language-Action (VLA) models have emerged as a promising paradigm for general-purpose robotic manipulation, yet their performan… | ↗Not briefed |
arxiv arxiv-ai | 2d | TOGBench: Benchmarking and Improving Tiny-Object Generation Fidelity in Aerial Imagery | Advances in deep generative models now enable the synthesis of realistic imagery. However, aerial imagery is dominated by tiny objects… | ↗Not briefed |
arxiv arxiv-ai | 2d | Online State-Representation Adaptation under Synthetic Interaction-Cue Remapping | When cues change what they predict, continuing to update a learned mapping may help an agent choose the appropriate action. We tested… | ↗Not briefed |
arxiv arxiv-ai | 2d | NarrowSpec: Draft Hidden Width Need Not Match the Target in Speculative Decoding | Speculative decoding is widely used to speed up generation in large language models (LLMs), with a small draft model proposing tokens… | ↗Not briefed |
bluesky bluesky-ai | 2d | A recent trend I’ve seen in web search (for video game information) is top ranking search result pages that are AI “nega-wikis”: pages auto… | A recent trend I’ve seen in web search (for video game information) is top ranking search result pages that are AI “nega-wikis”: pages… | ↗Not briefed |
hn hn-ai | 2d | Show HN: AgentShield Zero-dep secret redactor and shell firewall for Claude | — | ↗Not briefed |
arxiv arxiv-ai | 2d | A System 1 Decision Model for Medical Multiple-Choice Questions Built on Frozen Language Models | A System 1 model answers a multiple-choice question in a single pass. A frozen language model turns the question and each option into… | ↗Not briefed |
bluesky bluesky-ai | 2d | Kaum ein Unternehmen wächst schneller als das KI-Labor. Jetzt können Privatanleger teilhaben. Sollten Sie das auch? | Kaum ein Unternehmen wächst schneller als das KI-Labor. Jetzt können Privatanleger teilhaben. Sollten Sie das auch? | ↗Not briefed |
bluesky bluesky-ai | 2d | The average Anthropic employee will be worth $72M after the IPO. This is a significant windfall from a company that didn’t even exist in 20… | The average Anthropic employee will be worth $72M after the IPO. This is a significant windfall from a company that didn’t even exist… | ↗Not briefed |
bluesky bluesky-ai | 2d | turns out Intel isn't being saved by the ai boom, and boy howdy who could've foreseen that allowing a massive shortage in components for yo… | turns out Intel isn't being saved by the ai boom, and boy howdy who could've foreseen that allowing a massive shortage in components f… | ↗Not briefed |
rss techmeme | 2d | LanceDB, which makes a multimodal lakehouse for AI training, built on the open format Lance, raised a $57.5M Series B led by Bessemer and Innovation Endeavors (Kyt Dotson/SiliconANGLE) | Kyt Dotson / SiliconANGLE: LanceDB, which makes a multimodal lakehouse for AI training, built on the open format Lance, raised a $57.5… | ↗Not briefed |
arxiv arxiv-ai | 2d | Learning Martial Arts Motion Skills for Animating Opponents in VR Sparring Games | Animating virtual opponents is an important yet challenging task in game development. In martial arts games, opponent motions must not… | ↗Not briefed |
arxiv arxiv-ai | 2d | Evidence-Based Credit Assignment for Search-Augmented Language Models | Reinforcement learning from terminal answer rewards enables search-augmented language models to interleave reasoning and retrieval wit… | ↗Not briefed |
hn hn-ai | 2d | How are large companies not ripshit about the frontier labs? | I am utterly baffled that the entire world is not coming after Anthropic and OpenAI with pitchforks. We are essentially funding:<p>1.T… | ↗Not briefed |
bluesky bluesky-ai | 2d | www.counterpunch.org/2026/11/16/i... | www.counterpunch.org/2026/11/16/i... | ↗Not briefed |
bluesky bluesky-ai | 2d | みずほ、社内AIを米ChatGPTから中国deepseekに乗り換え「コストが40倍違う」🔧 ChatGPT | みずほ、社内AIを米ChatGPTから中国deepseekに乗り換え「コストが40倍違う」 | ↗Not briefed |
arxiv arxiv-ai | 2d | Metrics-in-the-Loop: Complexity Feedback Reshapes Agent-Written Code | Coding agents now write whole codebases in long autonomous runs, guided mainly by whether their tests pass. Yet passing tests alone do… | ↗Not briefed |