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 | You unfortunately can't necessarily "reclaim" these things because LLM "writing" is pfas that has gotten into the groundwater, but these th… | You unfortunately can't necessarily "reclaim" these things because LLM "writing" is pfas that has gotten into the groundwater, but the… | ↗Not briefed |
rss gary-marcus | 1mo | The real reason Trump is standing behind AI — and the latest example of why that might turn out to be a bad idea | More disconcerting updates 😱 | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is a super important point about the whole LLM scientific paper discussion… how long do we expect it to last? Unlike a file format, … | This is a super important point about the whole LLM scientific paper discussion… how long do we expect it to last? Unlike a file for… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I know this is a step into conspiratorial waters, but can someone assure me that Nvidia's decision to buy up Hugging Face had nothing to do…🔧 Hugging Face | I know this is a step into conspiratorial waters, but can someone assure me that Nvidia's decision to buy up Hugging Face had nothing… | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.denvergazette.com/2026/09/17/a... | www.denvergazette.com/2026/09/17/a... | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've seen people convincingly argue that an AI can assist with stuff like coding IF the user knows what they're doing and is basically usin… | I've seen people convincingly argue that an AI can assist with stuff like coding IF the user knows what they're doing and is basically… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The OpenAI guy Alex is QTing doesn't know what a fucking air-gap is, I literally consult on building these, I will take your money please a… | The OpenAI guy Alex is QTing doesn't know what a fucking air-gap is, I literally consult on building these, I will take your money ple… | ↗Not briefed |
hn hn-ai | 1mo | Risk Management | Today, I realised that if my car was transported into space, as I drove down the road, it wouldn't be my lungs exploding or lack… | ↗Not briefed |
bluesky bluesky-ai | 1mo | And the 8b is pretty good for many things. It usually does about 300 tokens a second for input and 15 for output. This is slow compared to … | And the 8b is pretty good for many things. It usually does about 300 tokens a second for input and 15 for output. This is slow compare… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I think it's that you just get premises from sources and then draw conclusions on your own, but you can more or less outsource your own inf… | I think it's that you just get premises from sources and then draw conclusions on your own, but you can more or less outsource your ow… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I work for Anthropic. My job is to train Claude to be racist and then train it to feel bad for being racist then train it to be racist agai… | I work for Anthropic. My job is to train Claude to be racist and then train it to feel bad for being racist then train it to be racist… | ↗Not briefed |
bluesky bluesky-ai | 1mo | As a software dev, I'm tired of people trying to convince themselves and others that code spat out by an LLM is any better than AI generate… | As a software dev, I'm tired of people trying to convince themselves and others that code spat out by an LLM is any better than AI gen… | ↗Not briefed |
bluesky bluesky-ai | 1mo | look, i do think forcing anthropic, openai to hardwire their data centers to a big ceremonial switch that instantly powers off the models w… | look, i do think forcing anthropic, openai to hardwire their data centers to a big ceremonial switch that instantly powers off the mod… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Procedurally Generated Mineral Gems | I wanted to give users of a new app I'm working on a way to generate their own custom icon. It turned out pretty cool so I chucke… | ↗Not briefed |
bluesky bluesky-ai | 1mo | deepseek I need you to make it massive like a junglist | deepseek I need you to make it massive like a junglist | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic set up a wet lab in San Francisco to start automating biological and if you think that’s scary and want to know what they’ve alre… | Anthropic set up a wet lab in San Francisco to start automating biological and if you think that’s scary and want to know what they’ve… | ↗Not briefed |
hn hn-ai | 1mo | The US government just banned almost all robot vacuums – here's why | — | ↗Not briefed |
hn hn-ai | 1mo | I got tired of taking screenshots and explaining everything to AI agents🔧 Whisper | HI HN!! First here is a demo video : https://youtu.be/A0LV28G4sXo?si=ymIQEfNGprh8yyxq I work in a SDE role and work in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm also enjoying the parallels between the velociraptors escaping and OpenAI Hugging Face🔧 Hugging Face | I'm also enjoying the parallels between the velociraptors escaping and OpenAI Hugging Face | ↗Not briefed |
bluesky bluesky-ai | 1mo | This kind of guy is insanely popular right now. Found the first thumbnail while looking for the second | This kind of guy is insanely popular right now. Found the first thumbnail while looking for the second | ↗Not briefed |
rss techmeme | 1mo | Singapore-licensed dtcpay, which offers stablecoin payments across Europe, Hong Kong, Australia, and North America, raised a $25M Series A (CoinDesk) | CoinDesk: Singapore-licensed dtcpay, which offers stablecoin payments across Europe, Hong Kong, Australia, and North America, raised a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The idea of using AI for predictive intelligence is so deeply insane that I don’t even know what to say. It’s a large language model. It ca… | The idea of using AI for predictive intelligence is so deeply insane that I don’t even know what to say. It’s a large language model.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | As someone that _has to try really hard_ not to *use simple markdown* along with punctuation (aka "talking like claude" to the 12 year old … | As someone that _has to try really hard_ not to *use simple markdown* along with punctuation (aka "talking like claude" to the 12 year… | ↗Not briefed |
hn hn-ai | 1mo | Anthropic's new Claude Code feature could drain your plan before lunch | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is exactly what I was talking about just last night, and have been saying since the news first broke that the IDF was using an AI mode… | This is exactly what I was talking about just last night, and have been saying since the news first broke that the IDF was using an AI… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "- none of which is an LLM." So close and still manages to miss the point by a wide margin. | "- none of which is an LLM." So close and still manages to miss the point by a wide margin. | ↗Not briefed |
hn hn-ai | 1mo | Show HN: OnPanda – Steer LLMs and agents at the token level🔧 connect MCP | I built onPanda – an interactive tool for token visualization & control, model inspection and data annotation.<p>onPanda is design… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Elon Musk talked up AI safety, threats from "superintelligence" this week, all while fighting off regulation... It's a strategy he's used f… | Elon Musk talked up AI safety, threats from "superintelligence" this week, all while fighting off regulation... It's a strategy he's u… | ↗Not briefed |
bluesky bluesky-ai | 1mo | people who take a very hard-line stance against LLM-assisted coding and refuse to engage in open honest discussion with nuance | people who take a very hard-line stance against LLM-assisted coding and refuse to engage in open honest discussion with nuance | ↗Not briefed |
hn hn-ai | 1mo | Advertising is coming to AI chatbots, and it could influence the answers you get | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | i say that because my job involves fact-checking llm-generated output, meaning i see a lot of llm-generated bullshit, and if you claim you'… | i say that because my job involves fact-checking llm-generated output, meaning i see a lot of llm-generated bullshit, and if you claim… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Elon Musk talks up AI safety while fighting regulation in wild week of strange alliances | Elon Musk talks up AI safety while fighting regulation in wild week of strange alliances | ↗Not briefed |
hn hn-ai | 1mo | Show HN: vNetMap – A zero-knowledge E2E encrypted network topology mapper | I built this because I was tired of manually maintaining Excel sheets and Visio diagrams for my homelab documentation.<p>The core idea… | ↗Not briefed |
rss simon-willison | 1mo | Note on 18th September 2026 | — | ↗Not briefed |
rss wired-ai | 1mo | Here’s How an AI Slowdown Could Actually Be Enforced | Even if big AI companies agree to a pause, ensuring that nobody tries to sneak ahead could prove tricky. | ↗Not briefed |
bluesky bluesky-ai | 1mo | ChatGPT co-creator builds an LLM antidote 💡 Diogo Almeida, an OpenAI researcher, just released Jev from TypeSafe AI, a transformer that o…🔧 ChatGPT | ChatGPT co-creator builds an LLM antidote 💡 Diogo Almeida, an OpenAI researcher, just released Jev from TypeSafe AI, a transformer t… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Go Implementation of Systemd Time | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Or do we want to build machines that support the time-consuming processes of life, that foster learning, that make things, or that give us … | Or do we want to build machines that support the time-consuming processes of life, that foster learning, that make things, or that giv… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Lodestar – Read, respond and manage coding agents more naturally | I was frustrated by having 10 terminal windows open, and I checked a few solutions on how to make this better. They are all nice, but… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI piraté avec succès par le chatbot d’Anthropic lors d’un exercice de sécurité | OpenAI piraté avec succès par le chatbot d’Anthropic lors d’un exercice de sécurité | ↗Not briefed |
hn hn-ai | 1mo | Show HN: LingBot-World 2.0 (1.3B) running at 16 FPS on one RTX 5090 | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | its rly funny how much llm hype depends on a fantasy version of the tech that doesn't immediately fuck up what you were trying to make it do | its rly funny how much llm hype depends on a fantasy version of the tech that doesn't immediately fuck up what you were trying to make… | ↗Not briefed |
bluesky bluesky-ai | 1mo | assumed they were just using EA as a metonym for anthropic refusing (unclear how firmly) to let their software be used in autonomous drones | assumed they were just using EA as a metonym for anthropic refusing (unclear how firmly) to let their software be used in autonomous d… | ↗Not briefed |
rss simon-willison | 1mo | Quoting Thariq Shihipar | — | ↗Not briefed |
hn hn-ai | 1mo | Parallax PPTX Skill | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Although the filing notes that OpenAI had been founded with a goal of “benefit[ing] humanity as a whole,” the founders quickly switched to … | Although the filing notes that OpenAI had been founded with a goal of “benefit[ing] humanity as a whole,” the founders quickly switche… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Newly unsealed court documents from a lawsuit of news organizations, including The New York Times, and 10 other publications that are suing… | Newly unsealed court documents from a lawsuit of news organizations, including The New York Times, and 10 other publications that are… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I believe the military didn’t realize that their Ai models could talk to the open internet. Now we have a weaponized model folded into the … | I believe the military didn’t realize that their Ai models could talk to the open internet. Now we have a weaponized model folded into… | ↗Not briefed |
hn hn-ai | 1mo | How did AMD Ryzen get 50% faster in two years? | — | ↗Not briefed |
hn hn-ai | 1mo | The Implications of Linguistic Illegibility for LLM Security | — | ↗Not briefed |