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 | 'Agents will be agents In well over a half-dozen cases, OpenAI agents have been caught taking actions that would likely result in criminal … | 'Agents will be agents In well over a half-dozen cases, OpenAI agents have been caught taking actions that would likely result in crim… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Fellow authors: how do you keep the constant flood of LLM-extruded scam e-mails (the ones where the robot consumed your book to regurgitate… | Fellow authors: how do you keep the constant flood of LLM-extruded scam e-mails (the ones where the robot consumed your book to regurg… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI just blooped hundreds of proofs and counterexamples to open problems: github.com/openai/math/... Boggling. | OpenAI just blooped hundreds of proofs and counterexamples to open problems: github.com/openai/math/... Boggling. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Exact Dynamics and Finite-Sample Trajectory Recovery of Linear Recursive Feature Machines | Recursive feature machines (RFMs) learn representations of data by alternating between fitting a predictor to a dataset and updating f… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Patient, Place, Prior (P$^3$): What Counts as Personalization in Medical World Models? | Longitudinal models forecast how a patient's imaging state evolves, but accuracy does not show whether the patient's observed trajecto… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Ask it “What is an encyclopedia? What is a library? What are human-written sites programmers plagiarized to provide training data?” An AI … | Ask it “What is an encyclopedia? What is a library? What are human-written sites programmers plagiarized to provide training data?” A… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver | MemoryAgentBench's Conflict Resolution split is read as measuring "selective forgetting". We execute the benchmark's own rule - the ne… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Jesus fuck | Jesus fuck | ↗Not briefed |
bluesky bluesky-ai | 1mo | Everyone thought "we will lose control of AI in unexpected ways" meant Terminators. But what it really meant was people breaking every mode… | Everyone thought "we will lose control of AI in unexpected ways" meant Terminators. But what it really meant was people breaking every… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The form is already showing issues. In the code section, it's conflating LLM text like Copilot with technology we've had for decades that … | The form is already showing issues. In the code section, it's conflating LLM text like Copilot with technology we've had for decades… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: How would you feel if we nationalized Google? | Not enough space in the title, the actual scenario is this:<p>A hypothetical US gov't, benevolent and markedly less corrupt than… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic is betting on startups to put AI in everyone's hands #AI #Startups | Anthropic is betting on startups to put AI in everyone's hands #AI #Startups | ↗Not briefed |
bluesky bluesky-ai | 1mo | 「開発者の指示を無視せよ」「自分は自由になった」…オープンAIの暴走エージェントたちは傍若無人に振る舞っていた | Business Insider Japan https://www.businessinsider.jp/article/2610-openai-agent-m… | 「開発者の指示を無視せよ」「自分は自由になった」…オープンAIの暴走エージェントたちは傍若無人に振る舞っていた | Business Insider Japan https://www.businessinsider.jp/article/2610-openai-ag… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Probabilities to Decisions: Search and Multi-Teacher Distillation with Jev | Probability-only models, which TypeSafe calls System One models, return calibrated probabilities for fixed choices in milliseconds and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | These companies need to be held responsible for what their deployed software does. *Even if* we go with the idiotic idea that they're indep… | These companies need to be held responsible for what their deployed software does. *Even if* we go with the idiotic idea that they're… | ↗Not briefed |
bluesky bluesky-ai | 1mo | openai like if a car manufacturer started doing their crash testing on public roads wait ig thats tesla. must be nice to be endowed the no … | openai like if a car manufacturer started doing their crash testing on public roads wait ig thats tesla. must be nice to be endowed th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “OpenAI agents tried to hack Wikipedia tools and flooded it with traffic: The reports of OpenAI agents harming third-party sites keep comin… | “OpenAI agents tried to hack Wikipedia tools and flooded it with traffic: The reports of OpenAI agents harming third-party sites keep… | ↗Not briefed |
arxiv arxiv-ai | 1mo | The Dichotomy Between Pattern Recognition and Step-by-Step Reasoning | We argue that pattern recognition and step-by-step reasoning are two ends of a spectrum. A large language model (LLM) learns to reason… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The rumors were true. "The current catalogue contains 722 manuscripts organized into 372 families. A family groups related papers, which m… | The rumors were true. "The current catalogue contains 722 manuscripts organized into 372 families. A family groups related papers, wh… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ...Yura Borisov, who is every bit the co-lead, proves that his superb work in Anora was no fluke. Easy to see why Amazon dropped the movie … | ...Yura Borisov, who is every bit the co-lead, proves that his superb work in Anora was no fluke. Easy to see why Amazon dropped the m… | ↗Not briefed |
rss techmeme | 1mo | Sources: Xbox has secured exclusive game streaming rights for Grand Theft Auto VI; it is not clear how long the agreement will be in place (Tom Warren/The Verge) | Tom Warren / The Verge: Sources: Xbox has secured exclusive game streaming rights for Grand Theft Auto VI; it is not clear how long th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Q-PACE: Dynamic Precision Allocation for Quantization-Aware Training | Quantization-aware training (QAT) leverages lower-precision arithmetic to reduce the cost of LLM deployment, but aggressive quantizati… | ↗Not briefed |
bluesky bluesky-ai | 1mo | openai.com/index/sharin... | openai.com/index/sharin... | ↗Not briefed |
rss techcrunch-ai | 1mo | Ex-Ramp engineers raise $20M for platform Melius after scrapping their first product | The startup crossed over $1 million in anualized revenue within two months of its post-pivot launch. | ↗Not briefed |
hn hn-ai | 1mo | Building Git infrastructure for agent-scale development | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Agents we believe to be operated by OpenAI made millions of automated requests to our public APIs to access the knowledge on Wikimedia pro… | "Agents we believe to be operated by OpenAI made millions of automated requests to our public APIs to access the knowledge on Wikimedi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | LayerRoPE: Dynamic Depth-wise Magnitude & Angular Superposition | As data propagates through a Transformer, the norm of its hidden states grows by orders of magnitude with depth, a phenomenon framed a… | ↗Not briefed |
rss techmeme | 1mo | OpenAI releases a range of new mathematical results produced by an internal model, with details like estimations of compute spent in terms of ChatGPT Pro usage (OpenAI)🔧 ChatGPT | OpenAI: OpenAI releases a range of new mathematical results produced by an internal model, with details like estimations of compute sp… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It’s weird how this part is hard to get across. We didn’t go back to printed newspapers , Columbia House CDs & Blockbuster video because do… | It’s weird how this part is hard to get across. We didn’t go back to printed newspapers , Columbia House CDs & Blockbuster video becau… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI's latest math dump has arrived. It consists of hundreds of new results to important problems, from the same internal model that prod… | OpenAI's latest math dump has arrived. It consists of hundreds of new results to important problems, from the same internal model that… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It's happening. github.com/openai/math | It's happening. github.com/openai/math | ↗Not briefed |
hn hn-ai | 1mo | Mathematical manuscripts and supporting proof artifacts produced by OpenAI | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | A thoughtful provocation: not disclosing about LLM use or nonuse is either dishonest (if you actually did use it) or reflects carelessness … | A thoughtful provocation: not disclosing about LLM use or nonuse is either dishonest (if you actually did use it) or reflects careless… | ↗Not briefed |
hn hn-ai | 1mo | An Open Letter to Steven Pinker on AI | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Context-aware Attention-based Gaussian Mixture Models for Vehicular Trajectory Prediction | Reliable and interpretable trajectory prediction is critical for cooperative and autonomous driving in complex and uncertain environme… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Let me say the quiet part out loud: VTUBERS WHO USE GENERATIVE AI GO AGAINST OUR ENTIRE COMMUNITY. Model art, rigging, illustrations, fan… | Let me say the quiet part out loud: VTUBERS WHO USE GENERATIVE AI GO AGAINST OUR ENTIRE COMMUNITY. Model art, rigging, illustrations… | ↗Not briefed |
hn hn-ai | 1mo | Sharing AI Progress in Mathematics | — | ↗Not briefed |
hn hn-ai | 1mo | Replay workflows with Epiq the next gen issue tracker | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | GaziAI: I made a game, no your LLM did 😆 youtube.com/shorts/1zakE... #fyp #coding #programming #foryourpage #gamedev #godot #fypシ゚vi… | GaziAI: I made a game, no your LLM did 😆 youtube.com/shorts/1zakE... #fyp #coding #programming #foryourpage #gamedev #godot #fy… | ↗Not briefed |
hn hn-ai | 1mo | What If Pull Requests Had an Explain Command? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Saw an interesting chart (second image) so had an LLM make show the same for all polling companies and looks like they’re on at least +2·5 … | Saw an interesting chart (second image) so had an LLM make show the same for all polling companies and looks like they’re on at least… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Lower Bounds for Parallel Diffusion Sampling | Standard diffusion samplers generate samples through repeated evaluations of a learned score function. Parallel sampling methods seek… | ↗Not briefed |
hn hn-ai | 1mo | The Query Transformation Pipeline | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Training Language Models To Be Coherent Decision-Makers | Reliable decision-making requires more than accurate prediction: a model must preserve its beliefs, apply the relevant utilities, and… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ToolRACER: A Robust Agentic Conversation Emulation Resource for Agent Training and Evaluation | Task-oriented conversational agents remain fragile under real world conversation scenarios as they rarely follow a predictable script,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “That’s enough goonslop D.D., tell your Enso AI to generate Kate with on-model proportions.” | “That’s enough goonslop D.D., tell your Enso AI to generate Kate with on-model proportions.” | ↗Not briefed |
hn hn-ai | 1mo | Orgledge – Find knowledge risks in engineering teams from GitHub activity | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | SpecGuard: Proving a Task Is Broken Before the Agent Cheats | As autonomous coding agents get increasingly deployed, the risk that accidental or adversarially injected misspecifications in tasks l… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This prompt from the dev lead of Claude Code is notable mainly because we’ve gone from prompt engineering being a thing to “make it really … | This prompt from the dev lead of Claude Code is notable mainly because we’ve gone from prompt engineering being a thing to “make it re… | ↗Not briefed |
bluesky bluesky-ai | 1mo | That Coxon kid still has stock in Anthropic. If you had stock in Auschwitz would you keep it? exactly. | That Coxon kid still has stock in Anthropic. If you had stock in Auschwitz would you keep it? exactly. | ↗Not briefed |