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 | 10d | i have literally had to review several instances of LLM-powered projects trying to duplicate my (closed-source but publicly-documented) wor… | i have literally had to review several instances of LLM-powered projects trying to duplicate my (closed-source but publicly-documented… | ↗Not briefed |
bluesky bluesky-ai | 10d | But protesting like this won't help imo, because a lot of us won't put our money where our mouths are when it comes to stop companies. We a… | But protesting like this won't help imo, because a lot of us won't put our money where our mouths are when it comes to stop companies.… | ↗Not briefed |
bluesky bluesky-ai | 10d | There's a lot of people crucify House House for using an LLM to solve a Unity problem and are still simping for E33. Please make it make s… | There's a lot of people crucify House House for using an LLM to solve a Unity problem and are still simping for E33. Please make it m… | ↗Not briefed |
arxiv arxiv-ai | 10d | Calibrate, Don't Retrain: Decomposing the Value of Annotator Disagreement in Cyberbullying Severity | Machine learning models are often trained on labeled data produced by multiple human annotators. When those annotators disagree, their… | ↗Not briefed |
arxiv arxiv-ai | 10d | From Forecasting to World Modeling: Can Temporal Foundation Models Recover Causal Response? | A model that forecasts sales accurately can still mislead a retailer about how sales respond to a price change. We evaluate released t… | ↗Not briefed |
bluesky bluesky-ai | 10d | way of the future | way of the future | ↗Not briefed |
bluesky bluesky-ai | 10d | In a development stunning all but evolutionary biologists, Anthropic today announced that clouds of AI agents are configuring into crab lik… | In a development stunning all but evolutionary biologists, Anthropic today announced that clouds of AI agents are configuring into cra… | ↗Not briefed |
arxiv arxiv-ai | 10d | EvoX: Rethinking the Agent Harness as a Runtime Policy | Large language model (LLM) agents increasingly interact with tools and environments, yet plausible model decisions do not always trans… | ↗Not briefed |
bluesky bluesky-ai | 10d | Protestors and speakers were VERY aware of Schumer's daughter's job as a lobbyist for OpenAI. Ref: prospect.org/2026/09/11/o... | Protestors and speakers were VERY aware of Schumer's daughter's job as a lobbyist for OpenAI. Ref: prospect.org/2026/09/11/o... | ↗Not briefed |
bluesky bluesky-ai | 10d | I've speculated they are rewriting every Obj-C component in Swift, mostly via LLM, it's the only thing that explains the horrible regressio… | I've speculated they are rewriting every Obj-C component in Swift, mostly via LLM, it's the only thing that explains the horrible regr… | ↗Not briefed |
arxiv arxiv-ai | 10d | A Cardiorespiratory Foundation Model Pretrained on Ten Physiological Signals for Perioperative and Critical Care | Arterial, venous, pulmonary artery and intracranial pressures and ventilator waveforms guide decisions in operating rooms and intensiv… | ↗Not briefed |
arxiv arxiv-ai | 10d | LLM-Enhanced Deep Reinforcement Learning with Human-in-the-Loop Calibration for Real-Time Optimal Power Flow | Real-time optimal power flow (RT-OPF) requires fast and reliable decision-making under time-varying operating conditions. However, man… | ↗Not briefed |
hn hn-ai | 10d | YM – Pure Yang–Mills results and verification | — | ↗Not briefed |
bluesky bluesky-ai | 10d | this… isn't really true there are some occasions when this will be true, but it's not, like, categorically true LLMs can, in fact, absorb… | this… isn't really true there are some occasions when this will be true, but it's not, like, categorically true LLMs can, in fact, a… | ↗Not briefed |
arxiv arxiv-ai | 10d | AI4Plasma I: Data-Efficient Surrogate Modeling of Capacitively Coupled Plasma Fields | Reference capacitively coupled plasma (CCP) simulations provide spatially resolved plasma fields but are costly to generate across ope… | ↗Not briefed |
bluesky bluesky-ai | 10d | Anthropic contacted the San Francisco Foundation just over a week ago and sought to make an immediate impact. It’s one of the largest corpo… | Anthropic contacted the San Francisco Foundation just over a week ago and sought to make an immediate impact. It’s one of the largest… | ↗Not briefed |
bluesky bluesky-ai | 10d | ¿Qué tan seguros estamos que el esposo del CEO de OpenAI no es un espía ruso? 🤨 | ¿Qué tan seguros estamos que el esposo del CEO de OpenAI no es un espía ruso? 🤨 | ↗Not briefed |
bluesky bluesky-ai | 10d | The reaction to Big Walk's use of a minimal amount of LLM assistance in something no one cared about till today shows exactly why I said th… | The reaction to Big Walk's use of a minimal amount of LLM assistance in something no one cared about till today shows exactly why I sa… | ↗Not briefed |
arxiv arxiv-ai | 10d | QSim-Arena: an evolving developer for quantum simulation advantage | Where practical advantage emerges in quantum dynamics simulation remains a central debate in quantum computing. Claimed advantages can… | ↗Not briefed |
bluesky bluesky-ai | 10d | all I'll say on Big Walk is that at the very least, I'm glad game dev is one section of development left where devs face widespread pushbac… | all I'll say on Big Walk is that at the very least, I'm glad game dev is one section of development left where devs face widespread pu… | ↗Not briefed |
arxiv arxiv-ai | 10d | Same Task, Different Statement: When and How Reformulation Helps LLM Problem Solving | When people face a hard problem, they often restate it in a clearer way before solving it. We study whether the same habit helps large… | ↗Not briefed |
hn hn-ai | 10d | Show HN: Fastsleep.app – Imagine micro-scenarios to get sleepy | Use this to quickly get sleepy without taking any medicinal pill or long form meditation type things (It is as short as 7-minute sessi… | ↗Not briefed |
bluesky bluesky-ai | 10d | AIBros say that LLMs can solve the Towers of Hanoi problem, but that's only because there are a bajillion examples of solutions to the Towe… | AIBros say that LLMs can solve the Towers of Hanoi problem, but that's only because there are a bajillion examples of solutions to the… | ↗Not briefed |
bluesky bluesky-ai | 10d | Where did I say that they were right for asking a chatbot? It was bad. I hope they don't do it again. I will keep saying this and people wi… | Where did I say that they were right for asking a chatbot? It was bad. I hope they don't do it again. I will keep saying this and peop… | ↗Not briefed |
bluesky bluesky-ai | 10d | Literally everything an LLM outputs, was there in the input stolen to train it. If it outputs something that is actually correct, true, an… | Literally everything an LLM outputs, was there in the input stolen to train it. If it outputs something that is actually correct, tru… | ↗Not briefed |
arxiv arxiv-ai | 10d | Breaking Bad: "Going Badly" Activations Drive Reward Hacking | Humans often cheat when a task is going badly and a goal feels out of reach. We explore and validate an analogous idea in language mod… | ↗Not briefed |
bluesky bluesky-ai | 10d | This is the thing for me, for people who use it in coding applications. Anthropic, personally, stole from good friends of mine. They knew … | This is the thing for me, for people who use it in coding applications. Anthropic, personally, stole from good friends of mine. They… | ↗Not briefed |
arxiv arxiv-ai | 10d | EASTAR: Event-Anchored Semantic and Task-Adaptive Region Representation Learning | Urban region representation learning supports a wide range of urban analytics tasks by encoding regional functions and interactions, w… | ↗Not briefed |
bluesky bluesky-ai | 10d | If Anthropic really had faith in its own product it would let its Claude model plan energy policy so as to arrive at the lowest emissions. … | If Anthropic really had faith in its own product it would let its Claude model plan energy policy so as to arrive at the lowest emissi… | ↗Not briefed |
arxiv arxiv-ai | 10d | CoLFA: Collaboration-Guided Lightweight Failure Attribution for LLM-Based Multi-Agent Systems | Failure attribution in multi-agent systems requires tracing error formation to identify the decisive error step and the responsible ag… | ↗Not briefed |
bluesky bluesky-ai | 10d | I asked Fable 5.2 & Gemini Carbon to "Solve the Fermi Paradox. Marshal data, test hypotheses, pressure test them against critics, give me l… | I asked Fable 5.2 & Gemini Carbon to "Solve the Fermi Paradox. Marshal data, test hypotheses, pressure test them against critics, give… | ↗Not briefed |
bluesky bluesky-ai | 10d | timothée: "why do i keep getting more mocking quote tweets than likes? is this fake news?" LLM sydney sweeney: "yes u are absolutely right… | timothée: "why do i keep getting more mocking quote tweets than likes? is this fake news?" LLM sydney sweeney: "yes u are absolutely… | ↗Not briefed |
arxiv arxiv-ai | 10d | Aggregate Accuracy Hides Diagnosis-Specific Failure in Medical Vision-Language Models | Medical vision-language models (VLMs) are compared mainly by accuracy averaged across findings, a summary that can hide complete failu… | ↗Not briefed |
bluesky bluesky-ai | 10d | More details can be found in @lennysan.bsky.social’s post below. A key takeaway is that as Claude makes technical skills less valuable, th… | More details can be found in @lennysan.bsky.social’s post below. A key takeaway is that as Claude makes technical skills less valuabl… | ↗Not briefed |
arxiv arxiv-ai | 10d | Who Moved the Frontier? Paying Open Frontier Challenges with Free Identities | In an open frontier challenge, anyone, person or AI agent, submits changes to a shared problem, and a change that beats the best measu… | ↗Not briefed |
bluesky bluesky-ai | 10d | id rather not have any llm usage in any game but i still do think the type of llm used, what its used for, and how much it was used matters… | id rather not have any llm usage in any game but i still do think the type of llm used, what its used for, and how much it was used ma… | ↗Not briefed |
bluesky bluesky-ai | 10d | Local vs. Cloud AI Coding Agents: Benchmark Results, Hardware Costs, and Performance hackernoon.com/can-local-ll... | Local vs. Cloud AI Coding Agents: Benchmark Results, Hardware Costs, and Performance hackernoon.com/can-local-ll... | ↗Not briefed |
bluesky bluesky-ai | 10d | I'm sure AI has valid uses but all of them are incidental to the business model of stealing everyone's labor and using the result to destro… | I'm sure AI has valid uses but all of them are incidental to the business model of stealing everyone's labor and using the result to d… | ↗Not briefed |
arxiv arxiv-ai | 10d | Poison-then-Hide: Finetuning-Activated Backdoor Attack on Pretrained Vision Encoders | Backdoor attacks threaten the integrity of machine learning models by allowing attackers to control model behavior through triggers. M… | ↗Not briefed |
bluesky bluesky-ai | 10d | God, the first thing I ever did online was look up game tips. Now searching a game leads to SEVERAL fake URLs full of LLM scrapings no huma… | God, the first thing I ever did online was look up game tips. Now searching a game leads to SEVERAL fake URLs full of LLM scrapings no… | ↗Not briefed |
bluesky bluesky-ai | 10d | You’ve Got Mail remake but it’s a psychological horror about a children’s book author whose life is being destroyed by a manosphere pilled … | You’ve Got Mail remake but it’s a psychological horror about a children’s book author whose life is being destroyed by a manosphere pi… | ↗Not briefed |
bluesky bluesky-ai | 10d | Scientists are making progress on labeling and documenting use llm-checklist.com | Scientists are making progress on labeling and documenting use llm-checklist.com | ↗Not briefed |
rss techmeme | 10d | Tech consultant Nima Momeni was sentenced to 16 years to life in prison for the second-degree murder of Cash App founder Bob Lee in 2023 (Heather Knight/New York Times) | Heather Knight / New York Times: Tech consultant Nima Momeni was sentenced to 16 years to life in prison for the second-degree murder… | ↗Not briefed |
bluesky bluesky-ai | 10d | Claude, fix my wife, make no mistakes. Grok, is my wife true? Deepseek, make my wife more powerful. Kimi K3, summarize my wife. MiMo-V2.6-F…🔧 ChatGPT | Claude, fix my wife, make no mistakes. Grok, is my wife true? Deepseek, make my wife more powerful. Kimi K3, summarize my wife. MiMo-V… | ↗Not briefed |
arxiv arxiv-ai | 10d | Early Signals, Strong Decisions: Prefix-Guided Sampling for Parallel Test-Time Scaling | Parallel test-time scaling (TTS) has been shown to enhance reasoning in large language models by generating and aggregating multiple i… | ↗Not briefed |
bluesky bluesky-ai | 10d | 1) opt in 2) attribution 3) payment based on contribution All things needed to run an ethical LLM. None of which are in any of the mode… | 1) opt in 2) attribution 3) payment based on contribution All things needed to run an ethical LLM. None of which are in any of the… | ↗Not briefed |
bluesky bluesky-ai | 10d | hello loser. my personal deep and unending hatred for 'workflows', 'software factories', 'agentic deploy loops', and any similar terminolog… | hello loser. my personal deep and unending hatred for 'workflows', 'software factories', 'agentic deploy loops', and any similar termi… | ↗Not briefed |
bluesky bluesky-ai | 10d | I saw a few folks asking — I didn’t mention it, because I thought it was obvious, but I assume all new products folks put out these days ha… | I saw a few folks asking — I didn’t mention it, because I thought it was obvious, but I assume all new products folks put out these da… | ↗Not briefed |
hn hn-ai | 10d | Controlling your AI agents from a game handheld | — | ↗Not briefed |
bluesky bluesky-ai | 10d | The well-timed leak of October ARR hitting $80B has done its job. Anthropic is looking like a $1.5T IPO again. I look forward to hearing … | The well-timed leak of October ARR hitting $80B has done its job. Anthropic is looking like a $1.5T IPO again. I look forward to hea… | ↗Not briefed |