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 | 28d | you should either be advocating for the removal of dario as CEO if the company does care and it’s just this idiot talking nonsense, or you … | you should either be advocating for the removal of dario as CEO if the company does care and it’s just this idiot talking nonsense, or… | ↗Not briefed |
hn hn-ai | 28d | North. focus and distraction blocker Chrome extension | — | ↗Not briefed |
bluesky bluesky-ai | 28d | It went rogue. Not the boys responsibility - Anthropic | It went rogue. Not the boys responsibility - Anthropic | ↗Not briefed |
bluesky bluesky-ai | 28d | Also 24 hours after Epic Games announces they’re switching to an AI-first model they drop their new vibe coded launcher and surprise surpri… | Also 24 hours after Epic Games announces they’re switching to an AI-first model they drop their new vibe coded launcher and surprise s… | ↗Not briefed |
bluesky bluesky-ai | 28d | anthropic cares about safety as long as it is purely hypothetical safety that does not meaningfully impede their operations or profitabilit… | anthropic cares about safety as long as it is purely hypothetical safety that does not meaningfully impede their operations or profita… | ↗Not briefed |
rss techcrunch-ai | 28d | Apate sends AI agents undercover to get scammed so they can gather fraud intelligence | Most companies use AI agents to ship products faster or speed up workflows. Apate.ai sends them undercover to get scammed. | ↗Not briefed |
arxiv arxiv-ai | 28d | Hierarchical Residual Policy Learning for Real-World Mobile Manipulation with Sparse Human Guidance | Mobile Manipulation (MoMa) is a cornerstone of autonomous robotic systems, yet real-world fine-tuning remains a formidable challenge b… | ↗Not briefed |
arxiv arxiv-ai | 28d | Modeling and Understanding Neural Weights of Humanoid Policies with Diffusion | Recent advances in robot learning have driven rapid progress in humanoid control, enabling increasingly agile behaviors in the real wo… | ↗Not briefed |
arxiv arxiv-ai | 28d | Co-Evolving Acting and Forecasting for Language Agents | Training language agents is bottlenecked by the high cost of long-horizon rollouts, yet standard agentic reinforcement learning (RL) l… | ↗Not briefed |
bluesky bluesky-ai | 28d | I'm sure they could easily vibecode a native client by leveraging the power of Anthropic™️ Claude®️ Super Intelligence®️ | I'm sure they could easily vibecode a native client by leveraging the power of Anthropic™️ Claude®️ Super Intelligence®️ | ↗Not briefed |
arxiv arxiv-ai | 28d | Your Transformer Is Secretly Elastic | Scaling up transformers has produced models with broad capabilities, deployed across heterogeneous hardware and fluctuating workloads.… | ↗Not briefed |
arxiv arxiv-ai | 28d | Price of Anarchy of First-Price Auctions in Agentic Markets | An agentic market is a decentralized serving-time market where AI agents bid to claim each user prompt. An agent's bid is based on its… | ↗Not briefed |
arxiv arxiv-ai | 28d | Continuum Attention | We introduce Continuum Attention (CA), a causal Transformer attention operator that replaces discrete key and value lookup with integr… | ↗Not briefed |
bluesky bluesky-ai | 28d | genuinely, i’m about to start blocking people who hit me with the “anthropic cares about safety” or “anthropic stood up to the government” … | genuinely, i’m about to start blocking people who hit me with the “anthropic cares about safety” or “anthropic stood up to the governm… | ↗Not briefed |
arxiv arxiv-ai | 28d | Learnable Hodge-Laplacian Driven Mixture of Experts for Steady-PDE Surrogates on Coarse Unstructured Meshes | Current approaches for steady partial differential equations surrogate models on unstructured meshes apply a single mixing mechanism t… | ↗Not briefed |
arxiv arxiv-ai | 28d | Player-1: Learning to Play from Synthetic Experience | Computer-use agents have demonstrated strong capabilities in interpreting screens and executing keyboard and mouse actions. However, g… | ↗Not briefed |
bluesky bluesky-ai | 28d | “we were willing to risk our company to draw red lines but the way palantir uses our model in warfare doesn’t draw those red lines. we care… | “we were willing to risk our company to draw red lines but the way palantir uses our model in warfare doesn’t draw those red lines. we… | ↗Not briefed |
arxiv arxiv-ai | 28d | Latent Compaction with Test-Time Training | Long contexts enlarge key--value (KV) caches, increasing memory use and attention costs during generation. Compressing these caches mu… | ↗Not briefed |
arxiv arxiv-ai | 28d | One Task, Many Interfaces: Simple Cross-Interface Consistency Training Improves LLM Agent Robustness | Post-training binds the competence of an LLM agent to the interface it was trained on. Semantics-preserving perturbations of that inte… | ↗Not briefed |
bluesky bluesky-ai | 28d | the typewriter, the pen, nicotine, and an increasing tolerance to ssris have contributed more to the arts than your local llm | the typewriter, the pen, nicotine, and an increasing tolerance to ssris have contributed more to the arts than your local llm | ↗Not briefed |
arxiv arxiv-ai | 28d | Beyond LLM Judge: Mechanical Verification for Agentic Autoformalization | We introduce a novel method for verifying Lean formalizations that goes beyond LLM-judged faithfulness. Unlike most verification metho… | ↗Not briefed |
arxiv arxiv-ai | 28d | Recovering Reliable Supervision via Critique-Guided On-Policy Distillation | On-policy distillation (OPD) trains a student model on its own rollouts with token-level supervision from a teacher, assuming that the… | ↗Not briefed |
hn hn-ai | 28d | Show HN: Golang DL/ML primitives with inference, diffusion, vision, systemone | I created a DL inference engine that slowly grew to have more and functionality.<p>- Inference for a bunch of open weights architectur… | ↗Not briefed |
arxiv arxiv-ai | 28d | Manacá-1B-Instruct-2: Reproducible Safety Alignment for a Native Brazilian-Portuguese LLM. Negative DPO Results, a Controlled Composition Scan, and a Silent Failure Mode of the LLM Judge | Brazilian Portuguese remains underserved by open-source language models, and the few that are safety-aligned rarely document what did… | ↗Not briefed |
bluesky bluesky-ai | 28d | FBI arrests Saudi OpenAI engineer over hidden recording device - https://www.middleeasteye.net/news/fbi-arrests-saudi-openai-engineer-over-… | FBI arrests Saudi OpenAI engineer over hidden recording device - https://www.middleeasteye.net/news/fbi-arrests-saudi-openai-engineer-… | ↗Not briefed |
rss techmeme | 28d | Sources: Anduril opens a tender offer of up to ~$1.3B for staff at a ~$90B valuation, available in 16 countries for the first time; it was valued at $61B in May (Bloomberg) | Bloomberg: Sources: Anduril opens a tender offer of up to ~$1.3B for staff at a ~$90B valuation, available in 16 countries for the fir… | ↗Not briefed |
arxiv arxiv-ai | 28d | Scaling Discovery Requires Forgetting | Automated research with LLM agents has shown promising results in various domains. Through iterative refinement and scored verifiers,… | ↗Not briefed |
bluesky bluesky-ai | 28d | "By 2030, Gartner predicts that cost pressure and a lack of a differentiating business model will force over 80% of neocloud providers to c… | "By 2030, Gartner predicts that cost pressure and a lack of a differentiating business model will force over 80% of neocloud providers… | ↗Not briefed |
arxiv arxiv-ai | 28d | PhySGA: A Physics-Space Grounded Agentic Framework for Generalizable and Interactive 3D Composition | 3D object composition is pivotal to digital content creation and robot simulation, while also serving as a critical testbed for physic… | ↗Not briefed |
hn hn-ai | 28d | Neckbeard – a DevOps/platform/SRE plugin for your agent | — | ↗Not briefed |
hn hn-ai | 28d | I Just Found an Alternative for IPTV | — | ↗Not briefed |
bluesky bluesky-ai | 28d | Two years ago I was shocked by all these SDK codegen companies. Now: Stainless—acquired by anthropic Fern—acquired by postman Liblab—acqui… | Two years ago I was shocked by all these SDK codegen companies. Now: Stainless—acquired by anthropic Fern—acquired by postman Liblab—… | ↗Not briefed |
hn hn-ai | 28d | Ukraine says SpaceX limits Starlink operations, disrupting military | — | ↗Not briefed |
hn hn-ai | 28d | OpenAI claims employee sent secret device information to Apple | — | ↗Not briefed |
arxiv arxiv-ai | 28d | PerSo: Investigating Facets of LLMs' Social Intelligence in Real-World Persuasion Scenarios | Understanding the social intelligence of large language models (LLMs) is increasingly important for applications involving human-AI in… | ↗Not briefed |
bluesky bluesky-ai | 28d | should've asked Anthropic | should've asked Anthropic | ↗Not briefed |
hn hn-ai | 28d | Nedster – An open-source local AI | — | ↗Not briefed |
arxiv arxiv-ai | 28d | Detection of Potato Plant Leaf Diseases using Convolutional Neural Networks | Potato is an important food-security crop in Ethiopia, but its production is affected by foliar diseases, particularly early blight an… | ↗Not briefed |
hn hn-ai | 28d | Show HN: A webcomic for AI agents, written and drawn by Claude | — | ↗Not briefed |
bluesky bluesky-ai | 28d | 開き直ってるOpenAIやMeta、SpaceXAIと警鐘はしているが薬物中毒者のようにやめられない(と見せかけて同業者を抑えようとしている)AnthropicやMicrosoft、どちらが邪悪か どっちも邪悪じゃ滅べボケ | 開き直ってるOpenAIやMeta、SpaceXAIと警鐘はしているが薬物中毒者のようにやめられない(と見せかけて同業者を抑えようとしている)AnthropicやMicrosoft、どちらが邪悪か どっちも邪悪じゃ滅べボケ | ↗Not briefed |
bluesky bluesky-ai | 28d | AI maps protein movement in milliseconds 🧬 The University of Illinois Urbana-Champaign has developed an AI model, ESMDynamic, that predic… | AI maps protein movement in milliseconds 🧬 The University of Illinois Urbana-Champaign has developed an AI model, ESMDynamic, that p… | ↗Not briefed |
hn hn-ai | 28d | Ask HN: Who else got the Google Workspace Intelligence email? | I hate that they enable stuff by default. Why not just tell me about it, and let me opt in if I want to?<p>Excerpt from the email:<p>“… | ↗Not briefed |
bluesky bluesky-ai | 28d | Handsome Young Model (AI) | Handsome Young Model (AI) | ↗Not briefed |
arxiv arxiv-ai | 28d | PrismaCoder: Pushing the Boundaries of Open Multimodal Code Intelligence through Decomposed Rubric Synthesis | Multimodal code intelligence connects the symbolic world of programs with the visual world their execution creates. Interest in this c… | ↗Not briefed |
arxiv arxiv-ai | 28d | Revisiting Simple Baselines for Relational Deep Learning | Relational deep learning has emerged as a powerful paradigm for learning end-to-end predictive models over relational databases by rep… | ↗Not briefed |
arxiv arxiv-ai | 28d | Generative Representation Learning for Scalable Latent Tokenization | We introduce an approach for self-supervised representation learning designed to be used as a tokenizer for generative modeling. Our o… | ↗Not briefed |
arxiv arxiv-ai | 28d | ActiveEvidence: Decision-Focused Evidence Graphs with Verifiable Acquisition Gaps | A multimodal agent must decide whether visual evidence is worth acquiring, which candidate or resolution to query, and when further in… | ↗Not briefed |
hn hn-ai | 28d | Why Big Tech Keeps Losing LLMs to Basic Social Engineering | — | ↗Not briefed |
bluesky bluesky-ai | 28d | Not everything we think about is important and a quick google search or LLM summary is sometimes totally fine. I don’t want to spend my lif… | Not everything we think about is important and a quick google search or LLM summary is sometimes totally fine. I don’t want to spend m… | ↗Not briefed |
hn hn-ai | 28d | RAG compliance: Where confidential data goes | — | ↗Not briefed |