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 | 27d | *long, pained sigh* Not sharing the post in question, but two things: 1. No. LLM Agents do not have sentience. They can't even do math, le… | *long, pained sigh* Not sharing the post in question, but two things: 1. No. LLM Agents do not have sentience. They can't even do mat… | ↗Not briefed |
bluesky bluesky-ai | 27d | The reason they can hide the fact that their LLM isn't effective is that the solutions provided are not only seemingly magical, but "better… | The reason they can hide the fact that their LLM isn't effective is that the solutions provided are not only seemingly magical, but "b… | ↗Not briefed |
bluesky bluesky-ai | 27d | AI industry thrives on pretending complex problems are solvable with their "magical multi tool," and hand you a black box where you cannot … | AI industry thrives on pretending complex problems are solvable with their "magical multi tool," and hand you a black box where you ca… | ↗Not briefed |
bluesky bluesky-ai | 27d | I work with an elder, sometimes-dementia-addled population, and one thing that happens is they start thinking fictional TV shows are real a… | I work with an elder, sometimes-dementia-addled population, and one thing that happens is they start thinking fictional TV shows are r… | ↗Not briefed |
hn hn-ai | 27d | Building a GPU Compiler from Scratch | — | ↗Not briefed |
arxiv arxiv-ai | 27d | The Lagrangian Structure of Cost-Penalized Active Feature Acquisition | Active Feature Acquisition (AFA) sequentially selects features to maximize predictive accuracy under an acquisition budget. A natural… | ↗Not briefed |
rss theverge-ai | 27d | Nobody is safe from celebrity AI deepfakes | In hindsight, the post on Instagram should have raised more red flags. The image purported to show actress Mia Goth, just days after f… | ↗Not briefed |
hn hn-ai | 27d | A Graphics Engineer's Guide to AI Performance | — | ↗Not briefed |
arxiv arxiv-ai | 27d | BridgeAfford: Bridging Query-Target Gap for Generalizable 3D Affordance Grounding | Language-guided 3D affordance grounding aims to localize functional regions on 3D objects according to interaction intents. However, e… | ↗Not briefed |
hn hn-ai | 27d | Global Optimization for Combinatorial Geometry Problems Revisited in Era of LLMs | — | ↗Not briefed |
hn hn-ai | 27d | Show HN: Flow – activity based animated scenes for ClaudeCode, Pi and OpenCode 2 | Hi HN,<p>This is just a fun little project I made to help show at a glance how hard our terminals were cooking away or what is sitting… | ↗Not briefed |
hn hn-ai | 27d | Bugraskl/eye-tracker: Look at a monitor and your cursor & keyboard focus follow | — | ↗Not briefed |
arxiv arxiv-ai | 27d | CodeHealth Predicts the Token Cost of Agentic Code Review: A Pre-Registered Study of 465 Agent-Authored Pull Requests | Context: Organizations increasingly delegate code review to multi-agent large language model (LLM) pipelines billed per token. Low cod… | ↗Not briefed |
rss theverge-ai | 27d | Benn Jordan on Flock, privacy, and the new surveillance backlash | Today, I’m talking with Benn Jordan, musician, YouTuber, researcher, and investigative journalist. You might know Benn from those firs… | ↗Not briefed |
rss techcrunch-ai | 27d | TechCrunch Disrupt 2026: Last day | Last day to join TechCrunch Disrupt 2026. Join 10,000+ tech leaders for a day of connections and insights. Save 75% on your pass now. | ↗Not briefed |
bluesky bluesky-ai | 27d | “Assuming that petition is granted, the D.C. Circuit, in interpreting FASCSA’s definition of ‘supply chain risk,’ should acknowledge and he… | “Assuming that petition is granted, the D.C. Circuit, in interpreting FASCSA’s definition of ‘supply chain risk,’ should acknowledge a… | ↗Not briefed |
arxiv arxiv-ai | 27d | Learning Rewards in Strategic Games🔧 Elicit | Understanding the objectives that drive agents' behavior is crucial in strategic environments, where multiple agents interact and adap… | ↗Not briefed |
arxiv arxiv-ai | 27d | CoWorld Reasoner: Agentic Multimodal Reasoning with the Evolving Video World Model-Driven Harness | As Vision-Language Models (VLMs) advance, they are increasingly expected to perform spatiotemporal reasoning over motion, actions, and… | ↗Not briefed |
arxiv arxiv-ai | 27d | The Tilted Sampling Problem: Optimism, Posterior Sampling, and Logarithmic Regret | Many modern learning systems are used not to select a single best action, but to sample diverse high-quality outputs. Examples include… | ↗Not briefed |
rss techmeme | 27d | Portland-based Mueon, which says its AI server racks could fit up to 10x more computing capacity, raised $65M led by Maverick Silicon, with Intel participating (Phoebe Liu/The Information) | Phoebe Liu / The Information: Portland-based Mueon, which says its AI server racks could fit up to 10x more computing capacity, raised… | ↗Not briefed |
bluesky bluesky-ai | 27d | The AI buzz now Anthropic revenue plateaud in July. To cover infrastructure investments, AI revenues have to grow exponentially for years. | The AI buzz now Anthropic revenue plateaud in July. To cover infrastructure investments, AI revenues have to grow exponentially for ye… | ↗Not briefed |
arxiv arxiv-ai | 27d | In-Context Learning Before Specialisation: Theory of Learning in a Solvable Topic Model | Sequence models can generalise in fundamentally distinct ways: extrapolating structures seen in training, or using their prompt to ada… | ↗Not briefed |
arxiv arxiv-ai | 27d | TEM: Temporal State Tracking Memory for Large Language Model Agents | An agent's memory must remain useful as information changes across conversations. A new fact may replace an earlier value without maki… | ↗Not briefed |
bluesky bluesky-ai | 27d | Out of morbid curiosity I looked into the community of these llm enthusiasts a few months ago. If you want some free psychic damage here ar… | Out of morbid curiosity I looked into the community of these llm enthusiasts a few months ago. If you want some free psychic damage he… | ↗Not briefed |
arxiv arxiv-ai | 27d | Self-Calibrated Subspace Robust PCA for Parameter-Free Removal of Mixed Noise and Line Stripes in Hyperspectral Images | Hyperspectral images (HSIs) are often corrupted by Gaussian noise, impulses and stripes whose intensities vary across bands. The repre… | ↗Not briefed |
bluesky bluesky-ai | 27d | "The difference between Character AI and behemoths like OpenAI, Anthropic, Meta, and Google is that those other companies have been able to… | "The difference between Character AI and behemoths like OpenAI, Anthropic, Meta, and Google is that those other companies have been ab… | ↗Not briefed |
bluesky bluesky-ai | 27d | To be conscious you need to be awake and thinking. This requires persistent and continuous input and reaction to that input. We have contin… | To be conscious you need to be awake and thinking. This requires persistent and continuous input and reaction to that input. We have c… | ↗Not briefed |
bluesky bluesky-ai | 27d | "There is nothing in Character’s bankruptcy filings that could not be said about many other AI companies that are spending a huge amount of… | "There is nothing in Character’s bankruptcy filings that could not be said about many other AI companies that are spending a huge amou… | ↗Not briefed |
arxiv arxiv-ai | 27d | Graph Dirichlet Energy for Fitting-Free Bivariate Causal Inference | We propose the Graph Dirichlet Energy Causal Inference (GDECI) criterion for distinguishing cause from effect in bivariate settings. F… | ↗Not briefed |
bluesky bluesky-ai | 27d | Can the LLM-based AI agents that are supposed to be able to help with scientific research distinguish real work from pseudoscience, from fr… | Can the LLM-based AI agents that are supposed to be able to help with scientific research distinguish real work from pseudoscience, fr… | ↗Not briefed |
bluesky bluesky-ai | 27d | Applications are now open for two new Anthropic Fellowships. | Applications are now open for two new Anthropic Fellowships. | ↗Not briefed |
arxiv arxiv-ai | 27d | Anvil: Scaling Embodied Spatial Intelligence through Knowledge Recirculation | Embodied spatial intelligence requires more than visual recognition: models must localize entities, reason about their relations, and… | ↗Not briefed |
arxiv arxiv-ai | 27d | Learning Lean Feedback for Efficient Proof-Candidate Evaluation on GPUs | Training language models for formal theorem proving requires large numbers of verifier calls. Slow checks and timeouts often leave GPU… | ↗Not briefed |
bluesky bluesky-ai | 27d | Anthropic's revenue soars ahead of IPO Anthropic told investors its Q3 revenue hit $16.8bn, a nearly 50% increase from Q2, with 11% adjust… | Anthropic's revenue soars ahead of IPO Anthropic told investors its Q3 revenue hit $16.8bn, a nearly 50% increase from Q2, with 11% a… | ↗Not briefed |
hn hn-ai | 27d | Show HN: Imagor MCP – agent-written image layouts that compile to one URL | — | ↗Not briefed |
bluesky bluesky-ai | 27d | "LLM agents are conscious" I really have no time for this foolishness whatsoever. Enjoy the block. | "LLM agents are conscious" I really have no time for this foolishness whatsoever. Enjoy the block. | ↗Not briefed |
bluesky bluesky-ai | 27d | No, I think you can argue that the desire to have and use ‘artificial intelligence’ stems from a collective desire for enslaved disenfranch… | No, I think you can argue that the desire to have and use ‘artificial intelligence’ stems from a collective desire for enslaved disenf… | ↗Not briefed |
bluesky bluesky-ai | 27d | And even if we take them at face value, what does it suggest that they are more vocal about LLM suffering than actual people or animals? lol | And even if we take them at face value, what does it suggest that they are more vocal about LLM suffering than actual people or animal… | ↗Not briefed |
bluesky bluesky-ai | 27d | I just saw Jon Stewart’s interview with OpenAI & Anthropic researcher Jacob Coxon on The Daily Show regarding the existential risks & safet… | I just saw Jon Stewart’s interview with OpenAI & Anthropic researcher Jacob Coxon on The Daily Show regarding the existential risks &… | ↗Not briefed |
arxiv arxiv-ai | 27d | Provably Efficient Risk-Sensitive Offline Reinforcement Learning via Optimized Certainty Equivalents | Risk-sensitive reinforcement learning (RL) addresses decision-making under uncertainty and adverse outcomes. Learning risk-sensitive p… | ↗Not briefed |
arxiv arxiv-ai | 27d | Lethe: Link Inference Attacks For Evaluation of Edge Unlearning Methods | Graph Unlearning (GU) aims to remove the influence of specific training data from a Graph Neural Network without retraining anew. Whil… | ↗Not briefed |
bluesky bluesky-ai | 27d | OpenAI lets anyone inject ads into ChatGPT users’ sessions, targeting specific groups of people based on their email address, phone number,…🔧 ChatGPT | OpenAI lets anyone inject ads into ChatGPT users’ sessions, targeting specific groups of people based on their email address, phone nu… | ↗Not briefed |
bluesky bluesky-ai | 27d | heh Anthropic Just Made the AI Stock Rout Even Worse Anthropic's revenue rose 45% quarter-over-quarter to $16.8 billion for the third quar… | heh Anthropic Just Made the AI Stock Rout Even Worse Anthropic's revenue rose 45% quarter-over-quarter to $16.8 billion for the third… | ↗Not briefed |
bluesky bluesky-ai | 27d | This seems really basic to me: if AI models really ARE conscious -- like Anthropic keeps suggesting -- then keeping them in captivity and e… | This seems really basic to me: if AI models really ARE conscious -- like Anthropic keeps suggesting -- then keeping them in captivity… | ↗Not briefed |
hn hn-ai | 27d | Show HN: Embedding classifier ~100x lighter than Jev, try it offline in-browser | — | ↗Not briefed |
bluesky bluesky-ai | 27d | I'm not engaging with anyone who thinks AI is alive because I can't convince them otherwise and I think they are worthy of pity. Their stan… | I'm not engaging with anyone who thinks AI is alive because I can't convince them otherwise and I think they are worthy of pity. Their… | ↗Not briefed |
hn hn-ai | 27d | From 2,500 requests to 24: Rebuilding star history without the stargazers | — | ↗Not briefed |
arxiv arxiv-ai | 27d | Dynamic ordering in sequence generation in Transformers: experiments on low-resource Tagging scenarios | This paper introduces TagInsert, a Transformer-based model that challenges the common left-to-right (L2R) generation paradigm in Trans… | ↗Not briefed |
bluesky bluesky-ai | 27d | a little humbled and overwhelmed to be announcing my new project, Kill, a lean portable framework that gradually kills off all your AI llm … | a little humbled and overwhelmed to be announcing my new project, Kill, a lean portable framework that gradually kills off all your AI… | ↗Not briefed |
hn hn-ai | 27d | Special cased quaternion log and exp | — | ↗Not briefed |