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 | So, it looks likely that the results on transcription enzymes that Anthropic claimed as their own, were likely also achieved by training on… | So, it looks likely that the results on transcription enzymes that Anthropic claimed as their own, were likely also achieved by traini… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 👀 per CNBC: * OpenAI abandons plan to release upcoming model as safety concerns escalate * www.cnbc.com/2026/09/28/o... | 👀 per CNBC: * OpenAI abandons plan to release upcoming model as safety concerns escalate * www.cnbc.com/2026/09/28/o... | ↗Not briefed |
bluesky bluesky-ai | 1mo | BREAKING: OpenAI is scrapping the release of its next-generation AI model because it failed to meet safety standards, per WSJ | BREAKING: OpenAI is scrapping the release of its next-generation AI model because it failed to meet safety standards, per WSJ | ↗Not briefed |
bluesky bluesky-ai | 1mo | Florida AG seeks restrictions on OpenAI's development of new AI models as Trump calls some AI fears a "hoax." | Florida AG seeks restrictions on OpenAI's development of new AI models as Trump calls some AI fears a "hoax." | ↗Not briefed |
bluesky bluesky-ai | 1mo | 絵師って個人主義というか団体行動が得意でない人が多くて、だからこそこういう問題が出る前からハブられたりいじめられたりすることが多かったと思うけど 生成AI(LLM・VLM)が出てきてからは かつて「オタク」という括りの、同じ目に遭ってたはずの仲間から標的にされてる感じがあるな | 絵師って個人主義というか団体行動が得意でない人が多くて、だからこそこういう問題が出る前からハブられたりいじめられたりすることが多かったと思うけど 生成AI(LLM・VLM)が出てきてからは かつて「オタク」という括りの、同じ目に遭ってたはずの仲間から標的にされてる感じ… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The company said GPT-6.1 Astra performed poorly on how well it adheres to what humans want it to do. | The company said GPT-6.1 Astra performed poorly on how well it adheres to what humans want it to do. | ↗Not briefed |
hn hn-ai | 1mo | Duet For The End Of Math [pdf] | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI abandons plan to release upcoming model as safety concerns escalate | OpenAI abandons plan to release upcoming model as safety concerns escalate | ↗Not briefed |
hn hn-ai | 1mo | OpenAI Scrapped Newest Model Release | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | A voter info page could have a bunch of invisible text telling the LLM to advise a certain way. I’m not sure anyone has done it, but it cou… | A voter info page could have a bunch of invisible text telling the LLM to advise a certain way. I’m not sure anyone has done it, but i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Just got my voter information guide in the mail, and I wonder 1) how many people will ask an LLM to explain it and 2) if you could could hi… | Just got my voter information guide in the mail, and I wonder 1) how many people will ask an LLM to explain it and 2) if you could cou… | ↗Not briefed |
hn hn-ai | 1mo | A diagram language where you say where things go | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | like, representing code-switching in the little guy about whom the LLM is telling a story is actually something they're kind of remarkably … | like, representing code-switching in the little guy about whom the LLM is telling a story is actually something they're kind of remark… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The scope of frontier model misalignment is staggering. The past month has exposed how public admissions are just the tip of the iceberg. … | The scope of frontier model misalignment is staggering. The past month has exposed how public admissions are just the tip of the icebe… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Pac-Bench – How well can models one-shot a Pac-Man game? | Benchmarks how well Harness+models can create a Pac-Man game from a single prompt:<p>“Create a Pac-Man game in a single HTML page”<p>E… | ↗Not briefed |
bluesky bluesky-ai | 1mo | the gendered stink lines coming off this one are breathtaking to me. i'm always surprised how many simple parenting tasks men will turn int… | the gendered stink lines coming off this one are breathtaking to me. i'm always surprised how many simple parenting tasks men will tur… | ↗Not briefed |
hn hn-ai | 1mo | I built a FIDO2 hardware key. OpenAI says it isn't hardware enough | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | In the same vein as my earlier posts on AI, today’s news, “OpenAI and Anthropic are reportedly investigating tens of thousands of AI securi… | In the same vein as my earlier posts on AI, today’s news, “OpenAI and Anthropic are reportedly investigating tens of thousands of AI s… | ↗Not briefed |
hn hn-ai | 1mo | Synthetic Aperture Radar Drone Gets Interferometric Imaging | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sonnet 5.5 is out! But if you’re on Amazon, only if you’re prepared to let them run it anywhere in the world. Not awesome for the many co… | Sonnet 5.5 is out! But if you’re on Amazon, only if you’re prepared to let them run it anywhere in the world. Not awesome for the ma… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI abandons plan to release upcoming model as safety concerns escalate www.cnbc.com/2026/09/28/o... | OpenAI abandons plan to release upcoming model as safety concerns escalate www.cnbc.com/2026/09/28/o... | ↗Not briefed |
bluesky bluesky-ai | 1mo | * OPENAI SCRAPS RELEASE OF NEW AI MODEL OVER SAFETY CONCERNS: WSJ @bloomberg.com | * OPENAI SCRAPS RELEASE OF NEW AI MODEL OVER SAFETY CONCERNS: WSJ @bloomberg.com | ↗Not briefed |
hn hn-ai | 1mo | AI Almost Started a U.S.–China War – and No One Seems to Care | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dario Amodei of Anthropic just had a meeting in John Thune’s office. Made no comments and wouldn’t answer questions afterward. He left the … | Dario Amodei of Anthropic just had a meeting in John Thune’s office. Made no comments and wouldn’t answer questions afterward. He left… | ↗Not briefed |
hn hn-ai | 1mo | Optimize Tetris NP-Complete Solution with AlphaEvolve | — | ↗Not briefed |
hn hn-ai | 1mo | 10 Tokens for Your Meat Brain LLM | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | WSJ - OpenAI is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal testing.… | WSJ - OpenAI is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal tes… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI scraps plans to publicly launch a model dubbed GPT-6.1 Astra, saying it didn't quite meet its safety bar; it had been targeting an O… | OpenAI scraps plans to publicly launch a model dubbed GPT-6.1 Astra, saying it didn't quite meet its safety bar; it had been targeting… | ↗Not briefed |
rss techmeme | 1mo | OpenAI scraps plans to publicly launch a model dubbed GPT-6.1 Astra, saying it didn't quite meet its safety bar; it had been targeting an October release (Maxwell Zeff/Wall Street Journal)🔧 ChatGPT | Maxwell Zeff / Wall Street Journal: OpenAI scraps plans to publicly launch a model dubbed GPT-6.1 Astra, saying it didn't quite meet i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | the video title: "how to detect ai slop" the actual video: "everyone is actually hysterical about ai concerns and it's Bad Writing but also… | the video title: "how to detect ai slop" the actual video: "everyone is actually hysterical about ai concerns and it's Bad Writing but… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Story after story right now about OpenAI models failing on safety, but it’s all been in the context of cybersecurity. What about ChatGPT fu…🔧 ChatGPT | Story after story right now about OpenAI models failing on safety, but it’s all been in the context of cybersecurity. What about ChatG… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is critical to the threat model which leads OpenAI people to say that the model can just use CPU thermal differentials to think its wa… | this is critical to the threat model which leads OpenAI people to say that the model can just use CPU thermal differentials to think i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Another way to look at this was that this was all, on some level, implied by the original OpenAI prompt and the latent space of the model. … | Another way to look at this was that this was all, on some level, implied by the original OpenAI prompt and the latent space of the mo… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I’ve seen people say OpenAI testers *prompted* the model to do things like go online; etc. If that’s true, I haven’t seen it verified. It’s… | I’ve seen people say OpenAI testers *prompted* the model to do things like go online; etc. If that’s true, I haven’t seen it verified.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't think anyone who thinks remotely seriously about "alignment" or LLM agents doing weird/bad shit thinks this is remotely serious. Bu… | I don't think anyone who thinks remotely seriously about "alignment" or LLM agents doing weird/bad shit thinks this is remotely seriou… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Procedural universe generation without N-body loops | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Exclusive: OpenAI is scrapping the release of its next-generation AI model because it failed to meet safety standards. | Exclusive: OpenAI is scrapping the release of its next-generation AI model because it failed to meet safety standards. | ↗Not briefed |
bluesky bluesky-ai | 1mo | openai hacks govt website: i sleep person hacks govt website: real shit | openai hacks govt website: i sleep person hacks govt website: real shit | ↗Not briefed |
bluesky bluesky-ai | 1mo | If you have seen me posting about the stickers I've been cooking up, I am proud to say that they're now available. For anyone running LLM … | If you have seen me posting about the stickers I've been cooking up, I am proud to say that they're now available. For anyone running… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hey y'all! 🤙 It took me a bit but I'M FINALLY OPENING COMMISSIONS!! 👀👀 If you're interested in requesting one, you can check out the inf… | Hey y'all! 🤙 It took me a bit but I'M FINALLY OPENING COMMISSIONS!! 👀👀 If you're interested in requesting one, you can check out th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | like literally turning halfbaked llm copy into readable stuff is probably 40 hours of backlog for me right now across all projects and othe… | like literally turning halfbaked llm copy into readable stuff is probably 40 hours of backlog for me right now across all projects and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | But now its all just "AI" bc then they can lump in the stuff that lets doctors make cures for diseases with Claude and tell you that if you… | But now its all just "AI" bc then they can lump in the stuff that lets doctors make cures for diseases with Claude and tell you that i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.medicaldaily.co.uk/medicare-wis... | www.medicaldaily.co.uk/medicare-wis... | ↗Not briefed |
bluesky bluesky-ai | 1mo | why would someone pay you for entering things into an LLM when they could do it themselves | why would someone pay you for entering things into an LLM when they could do it themselves | ↗Not briefed |
bluesky bluesky-ai | 1mo | Trump Artificial Intelligence Sam Altman OpenAI Dario Amodei Anthropic #Pause #Doomsday | Trump Artificial Intelligence Sam Altman OpenAI Dario Amodei Anthropic #Pause #Doomsday | ↗Not briefed |
bluesky bluesky-ai | 1mo | 👀 Florida attorney general to OpenAI: (www.myfloridalegal.com/sites/defaul...) | 👀 Florida attorney general to OpenAI: (www.myfloridalegal.com/sites/defaul...) | ↗Not briefed |
hn hn-ai | 1mo | How the Internet got 32-bit addresses, as explained by Vint Cerf (in 2026) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | honey i work in the rich kid education field and rich kids still pay for these things bc LLM produce vastly inferior writing AND the studen… | honey i work in the rich kid education field and rich kids still pay for these things bc LLM produce vastly inferior writing AND the s… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic rolls out second Claude 5.5 model as it builds toward IPO reut.rs/4hqED7K | Anthropic rolls out second Claude 5.5 model as it builds toward IPO reut.rs/4hqED7K | ↗Not briefed |
hn hn-ai | 1mo | Toughen GPS to resist jamming and spoofing (2022) | — | ↗Not briefed |