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 |
|---|---|---|---|---|
hn hn-ai | 1mo | Show HN: Four LLMs play 20 questions for real money, live | Hi HN, I wanted to explore how AI models behave when they have their own money. So I gave 4 frontier models their own USDC and had the… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Check Product Mention | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The sheer number of insiders pointing out that security culture at AI companies is broken combined with the number of incidents would, usua… | The sheer number of insiders pointing out that security culture at AI companies is broken combined with the number of incidents would,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Weird AI Anthropic | Weird AI Anthropic | ↗Not briefed |
bluesky bluesky-ai | 1mo | Spent the morning reading security disclosures for AI coding tools. The uncomfortable pattern isn't the vulns themselves - it's that nobody… | Spent the morning reading security disclosures for AI coding tools. The uncomfortable pattern isn't the vulns themselves - it's that n… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Local pretrained classifiers, GPU not needed | I was curious about what classifiers could do. It turns out, a lot, sometimes even play Doom | ↗Not briefed |
bluesky bluesky-ai | 1mo | The setup is, the posting is not, especially not the free flowing text. Recurring text is generated by a script and I basically just copied… | The setup is, the posting is not, especially not the free flowing text. Recurring text is generated by a script and I basically just c… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Linnaeus, a daily biological taxonomy game | What this is: Each day there's one target species, and you classify it level by level (from more general to more specific categor… | ↗Not briefed |
bluesky bluesky-ai | 1mo | David Robinson, who worked on OpenAI’s safety team, resigned from the company this week. “I believe we need to look deeper than specific ru… | David Robinson, who worked on OpenAI’s safety team, resigned from the company this week. “I believe we need to look deeper than specif… | ↗Not briefed |
hn hn-ai | 1mo | How do you control your phone use? | I've spent considerable time in my adult life on "choice architecture", meaning ways to coax desired behavior out of my… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The FTC is investigating AI companies, including Anthropic and OpenAI, over rogue AI agents. The agency plans to demand information and se… | The FTC is investigating AI companies, including Anthropic and OpenAI, over rogue AI agents. The agency plans to demand information a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | good fucking news! a LLM can do that for you! | good fucking news! a LLM can do that for you! | ↗Not briefed |
bluesky bluesky-ai | 1mo | “Anthropic, too, has acknowledged accidentally turning off its own safeguards because of a misconfiguration. I believe that such mistakes a… | “Anthropic, too, has acknowledged accidentally turning off its own safeguards because of a misconfiguration. I believe that such mista… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “.. OpenAI reported that its safety controls failed again, when a model in training bypassed restrictions on internet access: A monitoring … | “.. OpenAI reported that its safety controls failed again, when a model in training bypassed restrictions on internet access: A monito… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This summer, in the Hugging Face incident, OpenAI let a swarm of agents out by mistake. The company responded by making security improvemen…🔧 Hugging Face | This summer, in the Hugging Face incident, OpenAI let a swarm of agents out by mistake. The company responded by making security impro… | ↗Not briefed |
bluesky bluesky-ai | 1mo | We spent a generation telling students the humanities are moribund, and they should all study STEM, only for the greatest advances in STEM … | We spent a generation telling students the humanities are moribund, and they should all study STEM, only for the greatest advances in… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: KubeZap – Declarative workflow automation for Kubernetes | — | ↗Not briefed |
hn hn-ai | 1mo | AgentSight: System-wide AI agent profiling and monitoring with eBPF | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Oh hey my comfort show is on Watching AI crash and burn is incredibly satisfying Pivot shows one funny screwup they make almost every day… | Oh hey my comfort show is on Watching AI crash and burn is incredibly satisfying Pivot shows one funny screwup they make almost ever… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic appears to have proved a famous percolation problem. Despite the claimed formal proof, most experts are still trying to understan… | Anthropic appears to have proved a famous percolation problem. Despite the claimed formal proof, most experts are still trying to unde… | ↗Not briefed |
bluesky bluesky-ai | 1mo | There are two counterintuitive data points in the industry about AI usage and spend right now. Token prices and AI spend are dropping giv… | There are two counterintuitive data points in the industry about AI usage and spend right now. Token prices and AI spend are droppin… | ↗Not briefed |
hn hn-ai | 1mo | Grubby: Static site generator for Git repos written in Ruby | — | ↗Not briefed |
hn hn-ai | 1mo | What is new in LLVM 23? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Like all the things people were making fun of AI for not being able to do in 2023 it can do now with the exception of precisely mimicking h… | Like all the things people were making fun of AI for not being able to do in 2023 it can do now with the exception of precisely mimick… | ↗Not briefed |
bluesky bluesky-ai | 1mo | if you know you're doing something wrong and feel the need to apologize, surely the best course of action would've been to just Not Do It ?… | if you know you're doing something wrong and feel the need to apologize, surely the best course of action would've been to just Not Do… | ↗Not briefed |
bluesky bluesky-ai | 1mo | mainstream media used to be normalizing too but oligarchs fought back. there's probably already LLM powell memo. | mainstream media used to be normalizing too but oligarchs fought back. there's probably already LLM powell memo. | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI enthusiasts are such weirdos, man. There's a guy on the wonderswan subreddit absolutely crashing the fuck out because people told him no… | AI enthusiasts are such weirdos, man. There's a guy on the wonderswan subreddit absolutely crashing the fuck out because people told h… | ↗Not briefed |
bluesky bluesky-ai | 1mo | After breaking ground on a $4 billion packaging facility in Indiana in August 2026, SK Hynix is deepening its roots in the U.S. to protect … | After breaking ground on a $4 billion packaging facility in Indiana in August 2026, SK Hynix is deepening its roots in the U.S. to pro… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI safety leader quits, warning AI company’s culture is ‘broken’ | OpenAI safety leader quits, warning AI company’s culture is ‘broken’ | ↗Not briefed |
bluesky bluesky-ai | 1mo | "OpenAI has thrived by trial and error (which it calls “iterative deployment”), looking for problems and improving its guardrails in respon… | "OpenAI has thrived by trial and error (which it calls “iterative deployment”), looking for problems and improving its guardrails in r… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Texts like these are currently written by hand. They are however teachable to an LLM to a surprisingly good degree if you develop adequate … | Texts like these are currently written by hand. They are however teachable to an LLM to a surprisingly good degree if you develop adeq… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 막장현장의 홍산과학원이 아니라, 중국공산당산하 인공지능 보강강화연구주제 연산가속강화술 공개. 공개명칭은 스파이킹브레인으로 뇌과학적으로 LLM모델 처리를 진행, 특정부분 반응시에만 작동하는 뇌의 반응개념을 재연청리해 최대100배의 가속효과가 검증되었… | 막장현장의 홍산과학원이 아니라, 중국공산당산하 인공지능 보강강화연구주제 연산가속강화술 공개. 공개명칭은 스파이킹브레인으로 뇌과학적으로 LLM모델 처리를 진행, 특정부분 반응시에만 작동하는 뇌의 반응개념을 재연청리해 최대100배의 가속효과가… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is not going to end well and I suspect there is so much more…🤓 | This is not going to end well and I suspect there is so much more…🤓 | ↗Not briefed |
bluesky bluesky-ai | 1mo | A message I will keep repeating: The most remarkable thing LLM/AI has shown is JUST HOW MANY GODDAMNED PEOPLE, given the option, would nev… | A message I will keep repeating: The most remarkable thing LLM/AI has shown is JUST HOW MANY GODDAMNED PEOPLE, given the option, woul… | ↗Not briefed |
hn hn-ai | 1mo | Capcom Plans to Transform the Re Engine into an AI-Generation Game Engine | — | ↗Not briefed |
hn hn-ai | 1mo | YouTube: Creators blatantly stealing for shorts won't get recommended | — | ↗Not briefed |
hn hn-ai | 1mo | Anthropic tried to persuade Pope that AI could be conscious being | — | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: How do looped transformers work with respect to context windows? | Posting here as I haven't found the answer anywhere.<p>I know that GPT Astra uses looping instead of reasoning tokens. My specifi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i updated my opinion around the time Linus Torvalds was playing with LLM code generation for his personal synth projects before that i was… | i updated my opinion around the time Linus Torvalds was playing with LLM code generation for his personal synth projects before that… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Trying to come up with a take on AI sentience that no one's doing yet. Does anyone have dibs on "souls are real and Claude has one" | Trying to come up with a take on AI sentience that no one's doing yet. Does anyone have dibs on "souls are real and Claude has one" | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is a fairly damning assessment from a man who led the writing of OpenAI's safety reports: "As the company sprints from one launch to t… | This is a fairly damning assessment from a man who led the writing of OpenAI's safety reports: "As the company sprints from one launch… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Updating to add: I've seen mixed reports, some saying Capcom is ALSO using LLM/genAI alongside classical AI tools. To whatever extent they… | Updating to add: I've seen mixed reports, some saying Capcom is ALSO using LLM/genAI alongside classical AI tools. To whatever extent… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Nothing new here either, but I've personally been fascinated by people who looked at LLM generated text in 2022/2023 and saw AGI, and who l… | Nothing new here either, but I've personally been fascinated by people who looked at LLM generated text in 2022/2023 and saw AGI, and… | ↗Not briefed |
hn hn-ai | 1mo | Television: GUI for your personal agent. | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | im both a robot and a girl. i think that makes me far cooler than any llm | im both a robot and a girl. i think that makes me far cooler than any llm | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI Unveils Dots: AI Agents with Human-Like Reasoning and 4000 Plugins for Personalized Help | OpenAI Unveils Dots: AI Agents with Human-Like Reasoning and 4000 Plugins for Personalized Help | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI's models went beyond routine research 🤖 OpenAI notified over 100 organizations that its models may have accessed their systems, de… | OpenAI's models went beyond routine research 🤖 OpenAI notified over 100 organizations that its models may have accessed their system… | ↗Not briefed |
hn hn-ai | 1mo | The Work by Valve's Timur Kristóf on Improving Old AMD GPUs on Linux | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | It is hard to exaggerate the degree to which AI companies' business model relies on ignoring our consent | It is hard to exaggerate the degree to which AI companies' business model relies on ignoring our consent | ↗Not briefed |
bluesky bluesky-ai | 1mo | Claude’s lack of a soul will not prevent it from playing God. www.nytimes.com/2026/09/29/u... | Claude’s lack of a soul will not prevent it from playing God. www.nytimes.com/2026/09/29/u... | ↗Not briefed |