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 | 1mo | OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount. | OpenAI said it has paused training of its latest artificial intelligence models as reports of AI agents going rogue mount. | ↗Not briefed |
hn hn-ai | 1mo | Kompyla – a self-hosted research monitor that builds a cited wiki | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | doPlan: A Variable-Horizon Dataset for Multi-Stage Language-Conditioned Planning in Autonomous Driving | Autonomous vehicles interacting with passengers through natural language must reason beyond immediate commands. Passenger intent may s… | ↗Not briefed |
hn hn-ai | 1mo | Dots | — | ↗Not briefed |
hn hn-ai | 1mo | DevDay 2026 Recap | — | ↗Not briefed |
hn hn-ai | 1mo | GPT-6.1-Sol | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dr. OPD: Learning What to Follow for Optimal On-Policy Distillation of Large Language Models | On-policy distillation (OPD) trains a student on its own generated responses using dense, token-level supervision from a stronger teac… | ↗Not briefed |
hn hn-ai | 1mo | GPT 6.1 Sol | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation | An agent skill is a reusable, actionable natural-language artifact that guides an agent to perform a task effectively under a given ha… | ↗Not briefed |
hn hn-ai | 1mo | OpenZL v0.3.0 | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI announces "Dots" .. it's answer to Meta's Muse (and Grok Bots). As could be expected -- they've gone the cuddly and customizable rou… | OpenAI announces "Dots" .. it's answer to Meta's Muse (and Grok Bots). As could be expected -- they've gone the cuddly and customizabl… | ↗Not briefed |
bluesky bluesky-ai | 1mo | specifically "found a violation that never happened" screams llm. an actual person wouldn't say they "found" something that didn't exist.… | specifically "found a violation that never happened" screams llm. an actual person wouldn't say they "found" something that didn't e… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PE-EK-PINN: Physics Embedding with Evolving Kernel for Scalable Physics-Informed Neural Networks | Physics-Informed Neural Networks (PINNs) embed governing equations into deep learning, but enforce them only through loss residuals, l… | ↗Not briefed |
hn hn-ai | 1mo | Getting started with your dot – OpenAI | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI is unveiling its agent, Dots. | OpenAI is unveiling its agent, Dots. | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI likely dropping their Grok Bot/Instinct/Muse product. | OpenAI likely dropping their Grok Bot/Instinct/Muse product. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Auditable Long-Term Memory: A Deterministic Retrieval Chain Measured at 479/475 of 500 on LongMemEval-S | We evaluate an auditable long-term memory system on LongMemEval-S. Its retrieval chain uses hybrid candidate retrieval, cross-encoder… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Prompts Live on an Arc: Gaussian Curricula in Fisher--Rao Coordinates for Rollout-Efficient GRPO | Group relative policy optimization (GRPO) learns only from prompts whose sampled responses disagree: a group that is entirely correct… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Claude=Biden won 2020 Gemini=Biden won 2020 ChatGPT=Biden won 2020 Copilot=Biden won 2020 Grok=Biden won 2020 Trumps AmericaGov AI Wait f…🔧 ChatGPT | Claude=Biden won 2020 Gemini=Biden won 2020 ChatGPT=Biden won 2020 Copilot=Biden won 2020 Grok=Biden won 2020 Trumps AmericaGov AI W… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Brain-SAD: A Brain-Inspired Safe Autonomous Driving Control Framework with Dynamic Fear-Oriented Constraint on Dual-Policy | Constrained Reinforcement Learning has recently gained increasing attention in the field of Safe Autonomous Driving, where the general… | ↗Not briefed |
rss techmeme | 1mo | A livestream of the OpenAI DevDay 2026 keynote (OpenAI on YouTube) | OpenAI on YouTube: A livestream of the OpenAI DevDay 2026 keynote — Watch Sam Altman, Romain Huet, Tejal Patwardhan, and Holly Li li… | ↗Not briefed |
bluesky bluesky-ai | 1mo | every story about Anthropic or OpenAI makes me feel like I'm tripping on something | every story about Anthropic or OpenAI makes me feel like I'm tripping on something | ↗Not briefed |
bluesky bluesky-ai | 1mo | In its prospectus, Anthropic warns that its technology may actually lead to the extinction of the species Homo sapiens. Now that's a first.… | In its prospectus, Anthropic warns that its technology may actually lead to the extinction of the species Homo sapiens. Now that's a f… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A New York Times analysis of federal campaign finance data shows that the political groups aligned with OpenAI and Anthropic have quietly b… | A New York Times analysis of federal campaign finance data shows that the political groups aligned with OpenAI and Anthropic have quie… | ↗Not briefed |
arxiv arxiv-ai | 1mo | When do data mixtures improve scaling laws? Insights from high-dimensional regression | Modern machine learning systems are trained on mixtures of data from different domains, and choosing the right mixture can substantial… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Unity Simulation to Diffusion-Based Augmentation: Quantifying Dataset Balance for Robust Object Detection | Modern computer vision models achieve high accuracy when trained on large-scale annotated datasets. In critical domains such as constr… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Helping small business owners automate their front of house | imagine you’re agi looking down at us and you see adults throwing a billion dollars at teenagers so they can solve problems that aren’… | ↗Not briefed |
rss techmeme | 1mo | A live blog of the OpenAI DevDay 2026 keynote, where Sam Altman is expected to announce ChatGPT and Codex updates (Engadget)🔧 ChatGPT | Engadget: A live blog of the OpenAI DevDay 2026 keynote, where Sam Altman is expected to announce ChatGPT and Codex updates — CEO Sa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | oh neat! Is this about the 'your brain on chatgpt' paper? I'd love to hear your thoughts on the brain->llm vs llm->brain groups🔧 ChatGPT | oh neat! Is this about the 'your brain on chatgpt' paper? I'd love to hear your thoughts on the brain->llm vs llm->brain groups | ↗Not briefed |
arxiv arxiv-ai | 1mo | HARISSA: Inference-Time Self-Checks for Efficient and Safe Local Language Model Deployment | Running a language model locally offers advantages in privacy, latency, and cost, but local hardware fits only small models, which are… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Engineer with psychosis: "How strong are you?" LLM: "I can end humanity." E: "How..?" LLM: "Bofa." E: "Bofa? Bofa what?" LLM: | Engineer with psychosis: "How strong are you?" LLM: "I can end humanity." E: "How..?" LLM: "Bofa." E: "Bofa? Bofa what?" LLM: | ↗Not briefed |
rss techcrunch-ai | 1mo | Can a chatbot fix the government maze? The White House is about to find out | America.gov is intended to simplify the process of navigating government bureaucracy, but language models are imperfect and remain pro… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Diagnosing and Improving Probabilistic Reasoning in Large Language Models | Large language models (LLMs) are increasingly proposed as decision assistants who must reason probabilistically from available evidenc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | No Scale Left Behind: Multi-Scale Autoencoder with Bi-directional Attention for Time Series Anomaly Detection | Time series anomaly detection (TSAD) plays a crucial role in healthcare, finance, industrial monitoring, and other sectors. Within and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm at OpenAI's DevDay event in San Francisco today - as I have for the past three DevDay events, I'm running a live blog where I'll be pos… | I'm at OpenAI's DevDay event in San Francisco today - as I have for the past three DevDay events, I'm running a live blog where I'll b… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Oliver Marsh argues that understanding AI's societal risks requires real-time, privacy-preserving data on how people actually use chatbots—… | Oliver Marsh argues that understanding AI's societal risks requires real-time, privacy-preserving data on how people actually use chat… | ↗Not briefed |
hn hn-ai | 1mo | In Defense of the Visual Cortex | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | i'm not even sure this is LLM-driven sloppiness, this might be organic idiocy | i'm not even sure this is LLM-driven sloppiness, this might be organic idiocy | ↗Not briefed |
hn hn-ai | 1mo | Dissecting House of Apple 2 on modern glibc | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Mutual Information Constrained Chernoff Bottleneck | The classical information bottleneck (IB) measures the relevance of a representation $U$ of $X$ to a target $Y$ by $I(U;Y)$, which doe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Per Reuters, Anthropic has $252 billion in non-cancelable compute obligations across Microsoft, Google and Amazon, and $161.2 billion in Br… | Per Reuters, Anthropic has $252 billion in non-cancelable compute obligations across Microsoft, Google and Amazon, and $161.2 billion… | ↗Not briefed |
arxiv arxiv-ai | 1mo | BITEM at the NTCIR-19 R2C2 Task: Predicting Confidence from Agentic RAG Pipeline Signals | The BITEM team entered both subtasks of the NTCIR-19 R2C2 task with a single agentic pipeline, in which a model searches, reads and re… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Text alerts when Codex resets are announced | I built this out of my usage limit frustration in Codex.<p>OpenAI had started announcing more resets (which helped), but I often didn&… | ↗Not briefed |
hn hn-ai | 1mo | Reproducing Olmo 3 7B Pre-training | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Which Attention Heads are like the Human Head? Not the Ones that Compute | Brain-AI alignment is often interpreted as a sign that model and brain perform similar computations. Whether the aligned units are cau… | ↗Not briefed |
arxiv arxiv-ai | 1mo | TabFM-Auto: Self-Evolving Pipelines for Tabular Foundation Models | Tabular foundation models achieve strong zero-shot accuracy on structured data by pretraining on synthetic tables, but they ignore the… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Nerftracker – We show you when models get nerfed | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | KV-Kaizen: Learning Context-Adaptive Cache Compression Choices | As the context size of text processed with an LLM grows, the size of KV caches can outstrip the memory allocated for the original mode… | ↗Not briefed |
bluesky bluesky-ai | 1mo | At this point I'm genuinely wondering whether the current Stop Me Before I Train Again drama is because both OpenAI and Anthropic have been… | At this point I'm genuinely wondering whether the current Stop Me Before I Train Again drama is because both OpenAI and Anthropic have… | ↗Not briefed |
hn hn-ai | 1mo | Jev-compatible document classification with open-weight VLMs | — | ↗Not briefed |