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 | These sonsabitches are gonna try and cast "Tilly Norwood," the fake non-person bullshit dead-eyed LLM creation thing, as Wonder Woman or so… | These sonsabitches are gonna try and cast "Tilly Norwood," the fake non-person bullshit dead-eyed LLM creation thing, as Wonder Woman… | ↗Not briefed |
hn hn-ai | 1mo | OpenAI just dropped 700 preprints of mathematical proofs and counterexamples | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Bees – Run agentic teams on your own computer (open source) | Author here. Bees is an agentic harness that runs on your own machine — macOS, Windows and Linux, MIT/Apache 2.0.<p>The problem w… | ↗Not briefed |
bluesky bluesky-ai | 1mo | https://techcrunch.com/2026/10/06/anthropic-gives-startups-a-free-year-of-enterprise-service-and-1000-in-token-credits/ | https://techcrunch.com/2026/10/06/anthropic-gives-startups-a-free-year-of-enterprise-service-and-1000-in-token-credits/ | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Jrs – A Cargo-like build tool for Java, written in Rust | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | l.smartnews.com/p-8CXmZxEA/T... | l.smartnews.com/p-8CXmZxEA/T... | ↗Not briefed |
bluesky bluesky-ai | 1mo | As the use of artificial intelligence continues to expand, so do concerns over its influence in our digital systems. Attorney General Raúl… | As the use of artificial intelligence continues to expand, so do concerns over its influence in our digital systems. Attorney General… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I am not in a position to be able to evaluate these, but at least some of the results appear to be significant. Below is a Claude summary … | I am not in a position to be able to evaluate these, but at least some of the results appear to be significant. Below is a Claude sum… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Consistent Distribution Matching for Data-Free Diffusion Distillation | Flow and diffusion models suffer from slow inference due to computationally expensive numerical integration. Distillation provides a p… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They literally don’t care about human lives or the damage they do. www.theverge.com/ai-artificia... | They literally don’t care about human lives or the damage they do. www.theverge.com/ai-artificia... | ↗Not briefed |
arxiv arxiv-ai | 1mo | CM-DPO: Constraint-Margin Direct Preference Optimization for LLM Planning | Direct Preference Optimization (DPO) treats all constraint violations equally: a $1 budget overshoot and a $1,000 overshoot induce the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic is expanding access to its most advanced AI models, allowing a select group of organizations to test the startup’s cutting-edge c… | Anthropic is expanding access to its most advanced AI models, allowing a select group of organizations to test the startup’s cutting-e… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Solutions to ‘hundreds’ of open questions are said to be included in a batch of manuscripts released by OpenAI. www.theverge.com/ai-artific… | Solutions to ‘hundreds’ of open questions are said to be included in a batch of manuscripts released by OpenAI. www.theverge.com/ai-ar… | ↗Not briefed |
bluesky bluesky-ai | 1mo | https://wikimediafoundation.org/news/2026/10/05/openai-rogue-agent-activities-found-on-wikimedia-projects/ | https://wikimediafoundation.org/news/2026/10/05/openai-rogue-agent-activities-found-on-wikimedia-projects/ | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI has published what it describes as “a broad range of new mathematical results produced by an internal frontier model.” They claim t… | OpenAI has published what it describes as “a broad range of new mathematical results produced by an internal frontier model.” They cl… | ↗Not briefed |
arxiv arxiv-ai | 1mo | RLDISCOVER: LLM-driven co-evolution of reinforcement learning algorithms | LLM-guided program evolution has enabled discoveries in mathematics and computational optimization, raising the prospect of reinforcem… | ↗Not briefed |
rss theverge-ai | 1mo | OpenAI drops another batch of mathematical breakthroughs | OpenAI has revealed solutions to a number of long-standing mathematics problems produced by an unreleased frontier model in a batch of… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Deterministic Evidence Layer for Vision-Language Autism Screening from Naturalistic Home Video | Autism spectrum disorder (ASD) is diagnosed through specialist observation of a child's social behavior, and access to that expertise… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AGAR: a reinforcement learning substrate for LLM program evolution | Given a task and an evaluator, a language model can rewrite a candidate program while a search loop decides which rewrites survive, of… | ↗Not briefed |
rss techmeme | 1mo | Google releases Nano Banana 2.1, based on Gemini 3.6 Flash, saying it improves on previous versions "across the board"; pricing is ~50% lower vs. Nano Banana 2 (Matthias Bastian/The Decoder) | Matthias Bastian / The Decoder: Google releases Nano Banana 2.1, based on Gemini 3.6 Flash, saying it improves on previous versions “a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | it's funny to think that openai is showing off how cheaply they can do math compared to their peer | it's funny to think that openai is showing off how cheaply they can do math compared to their peer | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: What do you run on a $5 VPS that's worth keeping online 24/7? | DHH's "AI shed" idea got me thinking about getting a small $5–10/month VPS.<p>I'm a Linux guy and have on man… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Show HN: InterPoll, signed posts replicated over Gun relays/browsers | Live instance: <a href="https://endless.sbs" rel="nofollow">https://endless.sbs</a> Youtube explainer: <a href="ht… | ↗Not briefed |
hn hn-ai | 1mo | Read the 722 new OpenAI math papers | — | ↗Not briefed |
rss techmeme | 1mo | Sources: SpaceX is seeking to raise $40B, including ~$10B in bank loans and ~$30B in investment-grade debt, to purchase Nvidia chips, in a deal led by Apollo (Financial Times) | Financial Times: Sources: SpaceX is seeking to raise $40B, including ~$10B in bank loans and ~$30B in investment-grade debt, to purcha… | ↗Not briefed |
hn hn-ai | 1mo | Integer multiplication below n log n | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | CurveTQ: Rotation-Free Trellis Quantization of LLM Weights via Curvature-Weighted Search | The best two-bit weight quantizers for large language models, such as QTIP and Proteus, rotate each weight matrix by a random orthogon… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI released 700 math solutions today. I went through them so you don't have to. Here's what you need to know 👇 | OpenAI released 700 math solutions today. I went through them so you don't have to. Here's what you need to know 👇 | ↗Not briefed |
arxiv arxiv-ai | 1mo | Multi-Objective Aligned Small Language Model Framework for SUD Patient Dialogue Generation | Substance Use Disorder (SUD) counseling requires patient responses that reflect underlying cognitive states such as beliefs, coping st… | ↗Not briefed |
bluesky bluesky-ai | 1mo | yeah we all owe royalties to sam altman now for everything, as he has repeatedly stated is his intent for the future of openai | yeah we all owe royalties to sam altman now for everything, as he has repeatedly stated is his intent for the future of openai | ↗Not briefed |
arxiv arxiv-ai | 1mo | An Accuracy--Information Tradeoff for Loss-Difference Conditional Mutual Information | Loss-difference conditional mutual information (ld-CMI) uses the smallest of the standard observations in the supersample hierarchy of… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sam is going to get so mad if everyone doesn't say thank you with enough sincerity | Sam is going to get so mad if everyone doesn't say thank you with enough sincerity | ↗Not briefed |
arxiv arxiv-ai | 1mo | Do Vision Models Learn Physical Constraints or Rendering Shortcuts? A Counterfactual Benchmark for Grounded Physical Consistency | Modern image editing models can satisfy a text instruction while breaking the physics of the edited scene. A new object may cast no sh… | ↗Not briefed |
hn hn-ai | 1mo | Add your GitHub profile in your Apple Wallet | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | This isn't the work of a group trying to "improve how we share results with the math community." It's an attack on the math community. | This isn't the work of a group trying to "improve how we share results with the math community." It's an attack on the math community. | ↗Not briefed |
hn hn-ai | 1mo | The Quasi-Riemann Hypothesis [pdf] | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI fired 3 of its safety and alignment staff due to leaking proprietary safety data to an external third party. Ejaaz says we're in t… | OpenAI fired 3 of its safety and alignment staff due to leaking proprietary safety data to an external third party. Ejaaz says we're… | ↗Not briefed |
bluesky bluesky-ai | 1mo | #AI #ArtificialIntelligence #AIEngineering #GenerativeAI #LLM #AIAgents #AgenticAI #LLMAgents #MCP #ModelContextProtocol #ToolCalling #RAG … | #AI #ArtificialIntelligence #AIEngineering #GenerativeAI #LLM #AIAgents #AgenticAI #LLMAgents #MCP #ModelContextProtocol #ToolCalling… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 377 novel results clearly stolen from mathematicians' chat logs | 377 novel results clearly stolen from mathematicians' chat logs | ↗Not briefed |
bluesky bluesky-ai | 1mo | smart folks want the correct answers though, not hallucinated nonsense. the last place to look for new answers is in LLM output | smart folks want the correct answers though, not hallucinated nonsense. the last place to look for new answers is in LLM output | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI is working on at least 722 (so far) math manuscripts, all on GitHub in various stages of verificantion github.com/openai/math Why? … | OpenAI is working on at least 722 (so far) math manuscripts, all on GitHub in various stages of verificantion github.com/openai/math… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Few Bits, One Law: Toward W2A4KV2 | Extreme low-bit LLM compression is most challenging when weights, activations, and KV caches are quantized together: their distributio… | ↗Not briefed |
bluesky bluesky-ai | 1mo | that blows. openai knocked off something i used to care about | that blows. openai knocked off something i used to care about | ↗Not briefed |
bluesky bluesky-ai | 1mo | Props for the term “LLM-extruded” 😅 | Props for the term “LLM-extruded” 😅 | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic hopes you'll hire Claude to detect and discard them for you... | Anthropic hopes you'll hire Claude to detect and discard them for you... | ↗Not briefed |
bluesky bluesky-ai | 1mo | My thought has always been VRC should closely align with Steam's AI disclosure policy (both as a good model of practice and covering the co… | My thought has always been VRC should closely align with Steam's AI disclosure policy (both as a good model of practice and covering t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Lastly, the nature of an LLM is that the longer it simulates a “conversation”, the more noise starts to outweigh signal. Even if the origin… | Lastly, the nature of an LLM is that the longer it simulates a “conversation”, the more noise starts to outweigh signal. Even if the o… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It’s wild that in any other circumstance, going from zero to ~50 million paid subscribers would be an incredible business success. But whe… | It’s wild that in any other circumstance, going from zero to ~50 million paid subscribers would be an incredible business success. Bu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FreeEvolve: Learning to Evolve Beyond Fixed Loops | Agent evolvers automate the design of the prompts, skills and workflows around language model agents, yet the optimization process the… | ↗Not briefed |
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 |