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 |
|---|---|---|---|---|
arxiv arxiv-ai | 1mo | Video Generation Models: A Survey of Post-Training and Alignment | Video generation has rapidly progressed from short, low-quality clips to high-resolution, long-duration sequences with complex spatiot… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Didn't find anything (via the exhaustive lit review of "asking bluesky", lol) so I wrote all about what I'm thinking about LLMs, which is n… | Didn't find anything (via the exhaustive lit review of "asking bluesky", lol) so I wrote all about what I'm thinking about LLMs, which… | ↗Not briefed |
hn hn-ai | 1mo | LLM library that builds itself, a converter for nearly every file | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | One of the many things that sucks about the LLM age is that if even if we actually do figure out how to make true artificial intelligence, … | One of the many things that sucks about the LLM age is that if even if we actually do figure out how to make true artificial intellige… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Synopsys, OpenAI strike deal to develop AI model for chip design work reut.rs/3Vd0IiT | Synopsys, OpenAI strike deal to develop AI model for chip design work reut.rs/3Vd0IiT | ↗Not briefed |
arxiv arxiv-ai | 1mo | Paying for Too Many Tokens? Valid and Cost-Efficient Multimodal LLM Annotation with Simple Heuristics | Vision-Language Models (VLMs) enable video annotation at scale, but costs accumulate quickly: processing a typical 60-second short-for… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Post revised with a disclaimer noting that the project was coded with AI. Thank you to the folks who pointed this out! I didn’t notice the… | Post revised with a disclaimer noting that the project was coded with AI. Thank you to the folks who pointed this out! I didn’t notic… | ↗Not briefed |
hn hn-ai | 1mo | Can Agents Design Libraries for Agents? | — | ↗Not briefed |
hn hn-ai | 1mo | Context Language Models (CLMs) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Great piece on AI and creative writing on.substack.com/p/model-beha... | Great piece on AI and creative writing on.substack.com/p/model-beha... | ↗Not briefed |
bluesky bluesky-ai | 1mo | For real, any AI generated model has the topo of hot garbage, all the creases and corners are a soft mess, the textures are blurry and cree… | For real, any AI generated model has the topo of hot garbage, all the creases and corners are a soft mess, the textures are blurry and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I desperately want AI critics who take it seriously and are informed. We'd all be better off for it. There's plenty of things to take a sta… | I desperately want AI critics who take it seriously and are informed. We'd all be better off for it. There's plenty of things to take… | ↗Not briefed |
hn hn-ai | 1mo | Configurable Indent Levels? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | FTC move is first official US enforcement action on rogue AI agents, following surge in incidents first reported in July | FTC move is first official US enforcement action on rogue AI agents, following surge in incidents first reported in July | ↗Not briefed |
hn hn-ai | 1mo | EDG C++ Compiler is open source | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Writing was on the wall when the API protests came and went. It's why I left that place. And they're not doing it out of a principled anti… | Writing was on the wall when the API protests came and went. It's why I left that place. And they're not doing it out of a principled… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yep it’s cool arxiv.org/abs/2502.00873 | Yep it’s cool arxiv.org/abs/2502.00873 | ↗Not briefed |
bluesky bluesky-ai | 1mo | Our work Slack had an “AI Welfare” thread and I really wanted to tell them about my Claude /cock-and-ball-torture skill that make it invoke… | Our work Slack had an “AI Welfare” thread and I really wanted to tell them about my Claude /cock-and-ball-torture skill that make it i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It appears to have something related to claude on its github page... I don't frequently use github so I don't know whether that means AI wa… | It appears to have something related to claude on its github page... I don't frequently use github so I don't know whether that means… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm convinced it's non-LLM/GPT tech being branded as "AI" for the sake of attracting investors. | I'm convinced it's non-LLM/GPT tech being branded as "AI" for the sake of attracting investors. | ↗Not briefed |
rss techmeme | 1mo | Sources: Trump's AI accord grew out of a talk between Zuckerberg and House Speaker Mike Johnson; Zuckerberg circulated a draft, and Huang helped build support (Ashley Gold/Semafor) | Ashley Gold / Semafor: Sources: Trump's AI accord grew out of a talk between Zuckerberg and House Speaker Mike Johnson; Zuckerberg cir… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I hadn't heard about that piece until you mentioned it, and fucking WOW. They used the LLM-based 'AI detection' software that all experts w… | I hadn't heard about that piece until you mentioned it, and fucking WOW. They used the LLM-based 'AI detection' software that all expe… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sapien: A Stateful Policy Engine for Autonomous AI Agents | Contextual security defenses prevent AI agents from taking rogue actions by synthesizing a task-specific policy and enforcing it on th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | And it’s of course not your job to read my tweets. But if your job *is*, effectively, professional LLM explainer, as hers (not mine) kind o… | And it’s of course not your job to read my tweets. But if your job *is*, effectively, professional LLM explainer, as hers (not mine) k… | ↗Not briefed |
bluesky bluesky-ai | 1mo | god damn if I was evil I could make so much money scamming LLM people. easiest rubes on the planet. | god damn if I was evil I could make so much money scamming LLM people. easiest rubes on the planet. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Can large language models unlock discrete data in ophthalmic diagnostic reports? | Objective: To assess the accuracy and efficiency of a large language model (LLM) using two prompt strategies to extract structured dat… | ↗Not briefed |
bluesky bluesky-ai | 1mo | To be fair I thinj one thing deepmind is showing itself to be good at isnt the models but some ammount of R&D . They are helping give inspi… | To be fair I thinj one thing deepmind is showing itself to be good at isnt the models but some ammount of R&D . They are helping give… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Inference for stochastic differential equations driven by weighted sub-fractional Brownian motion using neural networks and the Euler approximation | We consider the estimation of drift, diffusion, and noise covariance from discrete observations of stochastic differential equations d… | ↗Not briefed |
bluesky bluesky-ai | 1mo | yeah man I love me an LLM: Livin’ la vida Loca (c’ Mon) | yeah man I love me an LLM: Livin’ la vida Loca (c’ Mon) | ↗Not briefed |
arxiv arxiv-ai | 1mo | Enterprise Representation Simplification (ERS): Reducing Representational Complexity for Enterprise AI | Enterprise information is represented through artifacts shaped by applications, projects, technologies, organizational boundaries, and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I’m not talking about a machine translator, I’m talking LLM vs human. | I’m not talking about a machine translator, I’m talking LLM vs human. | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'd rather fight against LLM users than namecall trendchasing artists. That design is meh though, and HP's just shitstalgia. | I'd rather fight against LLM users than namecall trendchasing artists. That design is meh though, and HP's just shitstalgia. | ↗Not briefed |
bluesky bluesky-ai | 1mo | It’s not exclusive to Open AI. Here’s a story about multiple Anthropic hacks. | It’s not exclusive to Open AI. Here’s a story about multiple Anthropic hacks. | ↗Not briefed |
hn hn-ai | 1mo | Nine – single binary, WASM tools and persistent AI harness/runtime | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic have a better PR department. www.bbc.com/news/article... | Anthropic have a better PR department. www.bbc.com/news/article... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yup. Anthropic skews Doomer, OpenAI seems to skew Booster. They are two sides of the same coin, but the difference is meaningful. | Yup. Anthropic skews Doomer, OpenAI seems to skew Booster. They are two sides of the same coin, but the difference is meaningful. | ↗Not briefed |
bluesky bluesky-ai | 1mo | EXCLUSIVE: Anthropic's IPO pitch embraces AI's promise and peril reut.rs/4hv2zqE | EXCLUSIVE: Anthropic's IPO pitch embraces AI's promise and peril reut.rs/4hv2zqE | ↗Not briefed |
bluesky bluesky-ai | 1mo | the people at anthropic left because guys at Open AI would be like "ya throw the AI into the crazy pit let's see what nasty shit they do"? … | the people at anthropic left because guys at Open AI would be like "ya throw the AI into the crazy pit let's see what nasty shit they… | ↗Not briefed |
rss latent-space | 1mo | Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week | Our DevDay coverage - the first pod on the DevDay lineup - dives in with the leaders of OpenAI’s CUA team and API platform. | ↗Not briefed |
bluesky bluesky-ai | 1mo | I mean literally kind of yes, Anthropic's origin story is that they're ex-oAI employees who left because oAI was insane and evil. Anthropic… | I mean literally kind of yes, Anthropic's origin story is that they're ex-oAI employees who left because oAI was insane and evil. Anth… | ↗Not briefed |
bluesky bluesky-ai | 1mo | is there a straightforward reason why it seems like every single AI hack moment is from Open AI? Do normal people work at Anthropic but cra… | is there a straightforward reason why it seems like every single AI hack moment is from Open AI? Do normal people work at Anthropic bu… | ↗Not briefed |
bluesky bluesky-ai | 1mo | remember that they sold reddit posts to openAI for training data they don’t want you to have free data if they can sell it, they think the… | remember that they sold reddit posts to openAI for training data they don’t want you to have free data if they can sell it, they thin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | WTF just happened today? ⚡️ Day 2080: “Toothless.” whatthefuckjusthappenedtoday.com/2026/09/30/d... | WTF just happened today? ⚡️ Day 2080: “Toothless.” whatthefuckjusthappenedtoday.com/2026/09/30/d... | ↗Not briefed |
bluesky bluesky-ai | 1mo | WTF just happened today? ⚡️ Day 2080: “Toothless.” whatthefuckjusthappenedtoday.com/2026/09/30/d... | WTF just happened today? ⚡️ Day 2080: “Toothless.” whatthefuckjusthappenedtoday.com/2026/09/30/d... | ↗Not briefed |
rss techmeme | 1mo | Sources: Jensen Huang and other execs asked Dario Amodei at the White House why he was so extreme in public on AI risks; Amodei said it's important to be honest (Wall Street Journal) | Wall Street Journal: Sources: Jensen Huang and other execs asked Dario Amodei at the White House why he was so extreme in public on AI… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Effective Synthetic Data Curation Requires Group-Level Signals | Synthetic data now is essential to LLM training, used to strengthen advanced capabilities such as autonomous and long-horizon task exe… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning Goal-Reaching Quasimetric Geometry From Finite-Time Reachability | In goal-conditioned reinforcement learning (GCRL), quasimetric learning models goal-reaching costs as quasimetric distances, connectin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Indeed. Fundraising off how much people want OpenAI to lose | Indeed. Fundraising off how much people want OpenAI to lose | ↗Not briefed |
bluesky bluesky-ai | 1mo | imagining department of war llm prompts. | imagining department of war llm prompts. | ↗Not briefed |
hn hn-ai | 1mo | Rivian: LiDAR R2 reaches customers in 2027, employees get it this year | — | ↗Not briefed |