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 | ik there are artists who program their own AI models and input their own images which i think is totally fine. thats whatever, the model wa… | ik there are artists who program their own AI models and input their own images which i think is totally fine. thats whatever, the mod… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ARC: A Reasoning Recipe for Robot Foundation Models | The prevailing approach to improving robot foundation models (RFMs) relies on larger models, more robot demonstrations, and costly tra… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Does that mean they've gone so far that "OpenAI" now has to be "OpenSI?" | Does that mean they've gone so far that "OpenAI" now has to be "OpenSI?" | ↗Not briefed |
bluesky bluesky-ai | 1mo | They’re good for translating human lingo messes into structured data or summaries I guess. Great for blind people too. Shame no one is sell… | They’re good for translating human lingo messes into structured data or summaries I guess. Great for blind people too. Shame no one is… | ↗Not briefed |
bluesky bluesky-ai | 1mo | When someone asks you if you are a LLM, you say YES! | When someone asks you if you are a LLM, you say YES! | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yeah, it's weird to me to see Anthropic themselves do this for their blog posts. Like, c'mon you're a trillion-dollar company?? claude.dev… | Yeah, it's weird to me to see Anthropic themselves do this for their blog posts. Like, c'mon you're a trillion-dollar company?? claud… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Major tech companies, including OpenAI, NVIDIA, Anthropic, Amazon and Google, are contributing $2.4 billion worth of computing resources in… | Major tech companies, including OpenAI, NVIDIA, Anthropic, Amazon and Google, are contributing $2.4 billion worth of computing resourc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | De la même manière qu'il était épaté par Macron... www.20minutes.fr/high-tech/sc... | De la même manière qu'il était épaté par Macron... www.20minutes.fr/high-tech/sc... | ↗Not briefed |
arxiv arxiv-ai | 1mo | Marformer: A Transformer for Predicting Missing Data Distributions | Real decisions are made under incomplete information. If we observe only some of the random variables we need, we can predict the othe… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Latent Core Tokenizer: Compress, but Meaningfully | Tokenizers are commonly optimized for compression, but a compact vocabulary does not necessarily distribute its capacity evenly across… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is brilliant …. His SI is an LLM (low language model) 😂👏🏻🤣 | This is brilliant …. His SI is an LLM (low language model) 😂👏🏻🤣 | ↗Not briefed |
arxiv arxiv-ai | 1mo | OnTrack: Real-Time Monitoring and Intervention in LLM Agent Trajectories via Streaming Structure-Aware Optimal Transport | Agents are deployed in applications from trip planners and stock trading to IT incident triage. In most cases, LLM agents work autonom… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I feel like I am losing my mind every time I see that a university has purchased another AI product while *gestures* ALL OF THIS is still n… | I feel like I am losing my mind every time I see that a university has purchased another AI product while *gestures* ALL OF THIS is st… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The Financial Times is reporting that OpenAI’s previously leaked ARR of $70 billion is actually $50 billion. I’ve been wondering when an A… | The Financial Times is reporting that OpenAI’s previously leaked ARR of $70 billion is actually $50 billion. I’ve been wondering when… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is so much Slopulism I can’t. OpenAI should speedrun math now out of spite @eleanor.lockhart.contact Information wants to be free | This is so much Slopulism I can’t. OpenAI should speedrun math now out of spite @eleanor.lockhart.contact Information wants to be f… | ↗Not briefed |
bluesky bluesky-ai | 1mo | otherwise you're gonna have to train an AI model to pretend to be your kid because they're not gonna talk to you | otherwise you're gonna have to train an AI model to pretend to be your kid because they're not gonna talk to you | ↗Not briefed |
hn hn-ai | 1mo | We reverse engineered ChatGPT Intelligent UI🔧 ChatGPT | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Bilevel optimization for data-driven learning of Koopman embeddings using kernel-based autoencoders | Koopman operator theory provides a linear framework for analyzing nonlinear dynamical systems and has become a major tool for data-dri… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ‘The metric, along with the equivalent figure for OpenAI rival Anthropic, is the most important indicator of overall demand for AI and unde… | ‘The metric, along with the equivalent figure for OpenAI rival Anthropic, is the most important indicator of overall demand for AI and… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Which Skill to Distill? SGUID: Selecting a Compact Skill Bank for Model-Skill Co-Evolution | Skills, reusable procedural guidance added at inference, can substantially improve LLM downstream performance (Li et al., 2026). Prior… | ↗Not briefed |
arxiv arxiv-ai | 1mo | HANS: A Handwritten Answer Sheet Dataset for Noisy Hybrid Document Parsing | Intelligent grading and automated scoring technologies constitute critical infrastructure for smart education. However, existing docum… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Closing the Horizon Gap in Policy Optimization for Adversarial MDPs | We consider policy optimization for online episodic tabular Markov decision processes (MDPs) with adversarial losses and bandit feedba… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I suggest that the right attitude to take toward the OpenAI math release is to imagine we have discovered a notebook full of results from a… | I suggest that the right attitude to take toward the OpenAI math release is to imagine we have discovered a notebook full of results f… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Cited but Not Consulted: A Counterfactual Audit of Legal Chain-of-Thought Faithfulness🔧 Replicate | Large language models increasingly justify legal decisions by naming the statute or precedent behind a verdict, treated as evidence th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Accurate but Not Humble: Evaluating Epistemic Humility in LLM Agents under Knowledge Conflict | When retrieved evidence contradicts an agent's prior beliefs, does it revise its answer, acknowledge uncertainty, or persist with an i… | ↗Not briefed |
hn hn-ai | 1mo | Bootstrap 6 Alpha | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | ‘Anthropic’s strongest asset today is that customers specifically want Claude. As long as that remains true, Anthropic has leverage despite… | ‘Anthropic’s strongest asset today is that customers specifically want Claude. As long as that remains true, Anthropic has leverage de… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Subspace Uncertainty and Sharp Sampling Thresholds on the Boolean Cube | We study Gaussian regression under squared population $L_2$ loss in a known $m$-dimensional subspace of degree-at-most-$k$ functions o… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Another reason not to use unknown/unverified tools or products... #AI #Cybersecurity #DataPrivacy #AI #InfoSec | Another reason not to use unknown/unverified tools or products... #AI #Cybersecurity #DataPrivacy #AI #InfoSec | ↗Not briefed |
arxiv arxiv-ai | 1mo | Distilling Routed 3D Privilege for Spatial Reasoning in Vision-Language Models | Spatial reasoning remains a persistent weakness of vision-language models (VLMs), because RGB inputs do not directly provide geometric… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 孫正義「ギャーッ！！」 OpenAI、年間収益が予想よりも200億ドル少ないと判明 終わる https://774neet.com/228005/ | 孫正義「ギャーッ！！」 OpenAI、年間収益が予想よりも200億ドル少ないと判明 終わる https://774neet.com/228005/ | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yup. Pretty much where I’m at. Useful technology, overhyped, overpromised, and overscaled. I won’t not use “AI” because of this. I have to … | Yup. Pretty much where I’m at. Useful technology, overhyped, overpromised, and overscaled. I won’t not use “AI” because of this. I hav… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic updates its usage policy to ban "sustained and needless abusive or cruel behavior" toward Claude, ending chats as the primary enf… | Anthropic updates its usage policy to ban "sustained and needless abusive or cruel behavior" toward Claude, ending chats as the primar… | ↗Not briefed |
bluesky bluesky-ai | 1mo | One of the companies in the duopoly that you can't criticize is called OPENAI lmao how does it even make sense | One of the companies in the duopoly that you can't criticize is called OPENAI lmao how does it even make sense | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI shares a report, previewed exclusively by NBC News, documenting foreign influence campaigns that used ChatGPT to varying degrees.🔧 ChatGPT | OpenAI shares a report, previewed exclusively by NBC News, documenting foreign influence campaigns that used ChatGPT to varying degree… | ↗Not briefed |
rss techmeme | 1mo | Anthropic updates its usage policy to ban "sustained and needless abusive or cruel behavior" toward Claude, ending chats as the primary enforcement mechanism (Hayden Field/The Verge) | Hayden Field / The Verge: Anthropic updates its usage policy to ban “sustained and needless abusive or cruel behavior” toward Claude,… | ↗Not briefed |
hn hn-ai | 1mo | APX Forth: A Surprise | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | that they report the revenue differently was well understood already, and even though I can't find it, I'm pretty sure there was a news sto… | that they report the revenue differently was well understood already, and even though I can't find it, I'm pretty sure there was a new… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SplitJEPA: Learning Invariant and Variant Latent Worlds without Reconstruction | Understanding a dynamical world calls for more than a latent state that summarizes its observations: the state should also be organize… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Claude Code spinner verbs reflect the model (Haiku: Licking screen..) | I found it tricky to know which intelligence level (effort+model) my claude code is on so I used the claude mod system to make the spi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Many nasty replies to my piece yesterday on flatlining model revenue b/c they were highly confident OpenAI ARR surged to 70bln based based … | Many nasty replies to my piece yesterday on flatlining model revenue b/c they were highly confident OpenAI ARR surged to 70bln based b… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I mean, if you thought they went from $40B AR on Aug 13 to $70B on Sept 30 (the Epoch chart shows it! expect a correction) and that these t… | I mean, if you thought they went from $40B AR on Aug 13 to $70B on Sept 30 (the Epoch chart shows it! expect a correction) and that th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Overcoming Prior Barriers: Supervised Fine-Tuning under Long-Tail Distribution | Supervised fine-tuning (SFT) adapts pretrained large language models (LLMs) to downstream tasks, but the required concepts can receive… | ↗Not briefed |
hn hn-ai | 1mo | Designing a local model harness optimized for MCP use | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Over the past year, Anthropic has repeatedly flirted with the idea that its AI models could be conscious in some way. In February, Anthropi… | Over the past year, Anthropic has repeatedly flirted with the idea that its AI models could be conscious in some way. In February, Ant… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Can AI Agents Learn Their Way to the Top? Evaluating Heuristic Learning in a Long-Running Game Agent Competition | Adversarial games have driven advances from heuristic search to reinforcement learning, yet learning and adapting strategies from limi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Looks like the שבכל דור ודור עומדים עלינו לכלותינו Now includes OpenAI 🤬 | Looks like the שבכל דור ודור עומדים עלינו לכלותינו Now includes OpenAI 🤬 | ↗Not briefed |
bluesky bluesky-ai | 1mo | absolutely the fuck not | absolutely the fuck not | ↗Not briefed |
arxiv arxiv-ai | 1mo | Ambient Discrete Diffusion: Using the Wrong Data at the Right Time for Data Efficient Learning | We introduce RefineMix, a framework for training discrete diffusion models under severe data scarcity, a common constraint in scientif… | ↗Not briefed |
hn hn-ai | 1mo | 2026 Usage Policy Update | — | ↗Not briefed |