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 | more cracks appearing. now we have this NVDA thing. we also have to get through the Anthropic ipo. and Dec hike a given with more in the wa… | more cracks appearing. now we have this NVDA thing. we also have to get through the Anthropic ipo. and Dec hike a given with more in t… | ↗Not briefed |
arxiv arxiv-ai | 10d | An LLM Is All You Need: Data-Efficient Adaptation to Full-Duplex, Word-Level Spoken Dialogue | Full-duplex spoken dialogue, in which a system listens while it speaks and can backchannel, yield the floor, and be interrupted, is us… | ↗Not briefed |
bluesky bluesky-ai | 10d | I really am loving the bi-monthly "Sure Nurbs has lost the will to live because of AI coding, but big walk runs 20% faster so whose to say … | I really am loving the bi-monthly "Sure Nurbs has lost the will to live because of AI coding, but big walk runs 20% faster so whose to… | ↗Not briefed |
bluesky bluesky-ai | 10d | If someone tried to attach an HTTP response body to a span, we'd immediately recognize that as a Bad Idea™️ but when it's an LLM response … | If someone tried to attach an HTTP response body to a span, we'd immediately recognize that as a Bad Idea™️ but when it's an LLM resp… | ↗Not briefed |
bluesky bluesky-ai | 10d | I'm a big fan of otel-all-the-things, but trying to shove full LLM agent conversations into spans seems like using a signal that happens to… | I'm a big fan of otel-all-the-things, but trying to shove full LLM agent conversations into spans seems like using a signal that happe… | ↗Not briefed |
bluesky bluesky-ai | 10d | Using or having used LLM tech I think has two paths forward: 1) We won't do that again. 2) We like this tech and we're doing open-weight, l… | Using or having used LLM tech I think has two paths forward: 1) We won't do that again. 2) We like this tech and we're doing open-weig… | ↗Not briefed |
hn hn-ai | 10d | How to find which deploy made your Postgres queries slower? | — | ↗Not briefed |
bluesky bluesky-ai | 10d | House House had an AI tool solve a ten line problem that neither them nor an actual engineer from Unity could solve. The Unity person told … | House House had an AI tool solve a ten line problem that neither them nor an actual engineer from Unity could solve. The Unity person… | ↗Not briefed |
bluesky bluesky-ai | 10d | openai cofounder Paul Christiano: “there’s a material risk of AI takeover in the next 6-12 months” m.youtube.com/watch?v=-Ef0... | openai cofounder Paul Christiano: “there’s a material risk of AI takeover in the next 6-12 months” m.youtube.com/watch?v=-Ef0... | ↗Not briefed |
arxiv arxiv-ai | 10d | Leaderboard Hacking: Preference-Based Model Evaluations are Vulnerable to Manipulation | Arena-style rankings of language models are a widely used evaluation framework in online leaderboards, research papers, and model eval… | ↗Not briefed |
bluesky bluesky-ai | 10d | why is every anti-LLM poster under the posts about Big Walk available for hire lol | why is every anti-LLM poster under the posts about Big Walk available for hire lol | ↗Not briefed |
youtube youtube-twominutepapers | 10d | OpenAI’s New AI Is Doing Things We Barely Understand | Something is changing in the world of mathematics and science, and it is changing fast. OpenAI just released hundreds of AI-generated… | ↗Not briefed |
hn hn-ai | 10d | Are frontier VLMs trained on robotics specific data? | — | ↗Not briefed |
bluesky bluesky-ai | 10d | Anthropic recently started referring to its AI data centers as "compute campuses," as if we wouldn't figure that one out. | Anthropic recently started referring to its AI data centers as "compute campuses," as if we wouldn't figure that one out. | ↗Not briefed |
bluesky bluesky-ai | 10d | Anthropic just sent me the fourth "hey, we need a form from you" e-mail that's identical to the one I've now done three times previously wi… | Anthropic just sent me the fourth "hey, we need a form from you" e-mail that's identical to the one I've now done three times previous… | ↗Not briefed |
arxiv arxiv-ai | 10d | Rational Use of Time for Social Coordination in Humans and AI | All social interactions are situated in time. Previous studies have shown that humans are able to use implicit temporal cues, such as… | ↗Not briefed |
arxiv arxiv-ai | 10d | SynthCorp: A Platform for Creating Benchmarks for Data Agents | Current benchmarks for LLM-driven data agents saturate quickly and cannot target specific capabilities. We present SynthCorp, a platfo… | ↗Not briefed |
bluesky bluesky-ai | 10d | "the (rapidly dwindling) human fraction of Anthropic" -Gwern, 2026-09-07 -Gwern, 2026-11-10 | "the (rapidly dwindling) human fraction of Anthropic" -Gwern, 2026-09-07 -Gwern, 2026-11-10 | ↗Not briefed |
bluesky bluesky-ai | 10d | "Anthropic called AI AI in its prospectus....Today... it filed an amended draft of the prospectus. Apparently somebody — at the US Securiti… | "Anthropic called AI AI in its prospectus....Today... it filed an amended draft of the prospectus. Apparently somebody — at the US Sec… | ↗Not briefed |
bluesky bluesky-ai | 10d | (Even then there's some "okay so we'll set you up with a local LLM to experiment with" because while it's not great, it's still better than… | (Even then there's some "okay so we'll set you up with a local LLM to experiment with" because while it's not great, it's still better… | ↗Not briefed |
rss techmeme | 10d | CoreWeave reports Q3 revenue up 168% YoY to $3.66B, vs. $3.55B est., a net loss of $709M, and a $132.5B revenue backlog, up from $104B as of June (Jordan Novet/CNBC) | Jordan Novet / CNBC: CoreWeave reports Q3 revenue up 168% YoY to $3.66B, vs. $3.55B est., a net loss of $709M, and a $132.5B revenue b… | ↗Not briefed |
bluesky bluesky-ai | 10d | Sorry, but also not sorry, for reposting & sharing alot of the recent reposts, but i am angry. I am disappointed at House House & will not … | Sorry, but also not sorry, for reposting & sharing alot of the recent reposts, but i am angry. I am disappointed at House House & will… | ↗Not briefed |
hn hn-ai | 10d | AgentGuard – Transaction guardrails for AI agents that hold wallets | — | ↗Not briefed |
hn hn-ai | 10d | Show HN: CodeLens AI-Augmented Code Review Tool | As AI writes more code, the bottleneck has become human-driven code reviews. Code Lens serves to aid that task by augmenting the revie… | ↗Not briefed |
hn hn-ai | 10d | Show HN: AI agents that barter your stuff for things you want | — | ↗Not briefed |
arxiv arxiv-ai | 10d | Graph Label Alignment: A Diagnostic Atlas for Graph Classification | Graph classification benchmarks vary widely in difficulty: some are solved by simple pooled baselines, while others remain challenging… | ↗Not briefed |
arxiv arxiv-ai | 10d | Benchmarking Multilingual Document Understanding in Vision-Language Models | Recent Vision-Language Models (VLMs) perform strongly on multimodal tasks, Document Understanding (DU) among them, but the benchmarks… | ↗Not briefed |
rss techmeme | 10d | Sources: the US Commerce Dept. delayed new Entity List additions for two months, leaving off Megaspeed despite four other US agencies pushing for its inclusion (Financial Times) | Financial Times: Sources: the US Commerce Dept. delayed new Entity List additions for two months, leaving off Megaspeed despite four o… | ↗Not briefed |
hn hn-ai | 10d | Ask HR: Systems Engineering's Senior Pipeline Crisis | The modern software industry is engineering its own talent famine. In mission-critical systems programming—especially within the C++ e… | ↗Not briefed |
bluesky bluesky-ai | 10d | This is basically what I'm going for, yeah. I'm not saying it's a /good/ thing they used an LLM to write 10 lines of code (which they the… | This is basically what I'm going for, yeah. I'm not saying it's a /good/ thing they used an LLM to write 10 lines of code (which the… | ↗Not briefed |
hn hn-ai | 10d | Show HN: Notebook AI – local-first meeting notes for Mac | — | ↗Not briefed |
bluesky bluesky-ai | 10d | Tthere is something else that irked me about the whole genAI situation with big walk is that I think it shows a remarkable lack of what I w… | Tthere is something else that irked me about the whole genAI situation with big walk is that I think it shows a remarkable lack of wha… | ↗Not briefed |
bluesky bluesky-ai | 10d | these are LLM-powered profiles that present as human home cooks, complete with AI pictures of "them" taking pictures of "their meals," with… | these are LLM-powered profiles that present as human home cooks, complete with AI pictures of "them" taking pictures of "their meals,"… | ↗Not briefed |
bluesky bluesky-ai | 10d | Gwern speaks of this ("the (rapidly dwindling) human fraction of Anthropic" (Gwern, 2026-09-07)) | Gwern speaks of this ("the (rapidly dwindling) human fraction of Anthropic" (Gwern, 2026-09-07)) | ↗Not briefed |
hn hn-ai | 10d | Claude Fable 5.2 HTML Files | — | ↗Not briefed |
bluesky bluesky-ai | 10d | Claude is so interesting to me because it's disliked by anti and pro AI people for wildly different reasons even though it's unequivocally … | Claude is so interesting to me because it's disliked by anti and pro AI people for wildly different reasons even though it's unequivoc… | ↗Not briefed |
bluesky bluesky-ai | 10d | Maybe Anthropic should hold off on the IPO until the storm passes *holds finger to ear* I see | Maybe Anthropic should hold off on the IPO until the storm passes *holds finger to ear* I see | ↗Not briefed |
bluesky bluesky-ai | 10d | recently there's been an influx of bot replies to my posts (that you don't see because I block them) trying to huck an "AI-powered" app tha… | recently there's been an influx of bot replies to my posts (that you don't see because I block them) trying to huck an "AI-powered" ap… | ↗Not briefed |
bluesky bluesky-ai | 10d | Já falei que não revisar código gerado por LLM é a apoteose do Extreme Go Horse. Deve ficando burros, a gestão mostrando que já era, e a em… | Já falei que não revisar código gerado por LLM é a apoteose do Extreme Go Horse. Deve ficando burros, a gestão mostrando que já era, e… | ↗Not briefed |
hn hn-ai | 10d | Superuser_detection_poker | — | ↗Not briefed |
bluesky bluesky-ai | 10d | The most BDS compliant game studio disclosing that they were forced to use a LLM code review as a means of being able to meet a deadline an… | The most BDS compliant game studio disclosing that they were forced to use a LLM code review as a means of being able to meet a deadli… | ↗Not briefed |
rss techmeme | 10d | Source: OpenAI told investors on Tuesday that its revenue jumped 50% to around $10B in Q3, up from 17% growth in Q2 (The Information) | The Information: Source: OpenAI told investors on Tuesday that its revenue jumped 50% to around $10B in Q3, up from 17% growth in Q2… | ↗Not briefed |
arxiv arxiv-ai | 10d | From Nodes to Pixels: Topological and Structural Two-View Graph Imaging | Graph learning still lacks a broadly reusable input interface comparable to patches in vision or tokens in language. Unlike images or… | ↗Not briefed |
arxiv arxiv-ai | 10d | Don't Match the Noise: Distribution Matching under Unknown Measurement Error with an Audit Sample | Real-world machine learning systems often operate on imperfect measurements: sensors are noisy, annotations are produced by non-expert… | ↗Not briefed |
bluesky bluesky-ai | 10d | Watching the death throes of Anthropic is only going to get funnier. They've just announced offerings attempting to compete with Google and… | Watching the death throes of Anthropic is only going to get funnier. They've just announced offerings attempting to compete with Googl… | ↗Not briefed |
arxiv arxiv-ai | 10d | UniPipe: Unified Inference Serving and Fine-Tuning for Deadline-Constrained DNN Surrogate Workflows | DNN surrogates are increasingly used to accelerate scientific workflows that must process high-rate sensor streams in near real time,… | ↗Not briefed |
hn hn-ai | 10d | Yabanin, a self-hosted gateway for model access, routing and a signed record | — | ↗Not briefed |
hn hn-ai | 10d | Free Open Source Timegrapher App for Mechanical Watches | — | ↗Not briefed |
hn hn-ai | 10d | An ARTEX build ships no secret scanner. In our lab, it found the secrets anyway | — | ↗Not briefed |
bluesky bluesky-ai | 10d | Trying to say it's okay to go back on your AI/LLM stances because game you liked by a studio that supports BDS used it is so funny. it's no… | Trying to say it's okay to go back on your AI/LLM stances because game you liked by a studio that supports BDS used it is so funny. it… | ↗Not briefed |