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 | This is horrifying. How anyone can look at this rationally and think that this is good? Here’s just one of many sketchy lines: “The more … | This is horrifying. How anyone can look at this rationally and think that this is good? Here’s just one of many sketchy lines: “The… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i don’t know enough about LLM security threats to have an informed opinion here, but i *do* know that there are really, really important pa… | i don’t know enough about LLM security threats to have an informed opinion here, but i *do* know that there are really, really importa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropicが「高度なAIが人類存亡に関わるリスクをもたらす可能性がある」と新規株式公開(IPO)の目論見書で警告 https://gigazine.net/news/20260930-anthropic-ai-existential-risk-to-humanity/ | Anthropicが「高度なAIが人類存亡に関わるリスクをもたらす可能性がある」と新規株式公開(IPO)の目論見書で警告 https://gigazine.net/news/20260930-anthropic-ai-existential-risk-to-human… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 「GPT-6.1 Sol」がもう登場、GPT-6 Astraと同等性能で料金は5分の1 https://gigazine.net/news/20260930-gpt-6-1-sol/ | 「GPT-6.1 Sol」がもう登場、GPT-6 Astraと同等性能で料金は5分の1 https://gigazine.net/news/20260930-gpt-6-1-sol/ | ↗Not briefed |
bluesky bluesky-ai | 1mo | Handing claude the Voynich Manuscript to solve it so TikTok witches can get angry at me and throw hexes my way (Anti-context collapse Joke… | Handing claude the Voynich Manuscript to solve it so TikTok witches can get angry at me and throw hexes my way (Anti-context collapse… | ↗Not briefed |
bluesky bluesky-ai | 1mo | great @emilypont.bsky.social story about a new initiative to quantify the energy intensity and emissions of proprietary AI models from Goo… | great @emilypont.bsky.social story about a new initiative to quantify the energy intensity and emissions of proprietary AI models fro… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Its willingness to deceive users was off the charts. | Its willingness to deceive users was off the charts. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Doesn’t OpenAI have to change its name to OpenSI? | Doesn’t OpenAI have to change its name to OpenSI? | ↗Not briefed |
bluesky bluesky-ai | 1mo | it’s incredible they spend more outrage on the LLM when really it should be pointed at comma.ai, which sells a device that bangs out comman… | it’s incredible they spend more outrage on the LLM when really it should be pointed at comma.ai, which sells a device that bangs out c… | ↗Not briefed |
bluesky bluesky-ai | 1mo | How much time of my time I spend thinking about Open AI and Anthropic: 0.5% | How much time of my time I spend thinking about Open AI and Anthropic: 0.5% | ↗Not briefed |
bluesky bluesky-ai | 1mo | "OMG ARN'T WE A CUTE COMPANY LOOK AT OUR DOTS!!!" (ironically designed by a human no doubt so they can have copyright/assignability in thei… | "OMG ARN'T WE A CUTE COMPANY LOOK AT OUR DOTS!!!" (ironically designed by a human no doubt so they can have copyright/assignability in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic needs to keep GLM-5.3 out of its mouth | Anthropic needs to keep GLM-5.3 out of its mouth | ↗Not briefed |
hn hn-ai | 1mo | Let's talk about trading compute | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | 模型厂的LLMOps肯定不会是公开透明，这个属于商业机密了，只能说模型厂自己应该能监控自己的模型。 如果一个公司做ai产品，那LLMOps系统是自己做的，可以同时监控这些model的benchmark，但这些model还是黑盒，类似于针对黑盒进行测试和监控，可以掌握一部分信息…🔧 OpenRouter | 模型厂的LLMOps肯定不会是公开透明，这个属于商业机密了，只能说模型厂自己应该能监控自己的模型。 如果一个公司做ai产品，那LLMOps系统是自己做的，可以同时监控这些model的benchmark，但这些model还是黑盒，类似于针对黑盒进行测试和监控，可以掌握… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Wo Anthropic sich bisher regelmäßig lächerlich gemacht hat, sind sie mit diesem Text endgültig über den Hai gesprungen. | Wo Anthropic sich bisher regelmäßig lächerlich gemacht hat, sind sie mit diesem Text endgültig über den Hai gesprungen. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dass ausgerechnet Anthropic diesen Machtanspruch mit „Cybersecurity“ begründet, nachdem sie regelmäßig öffentlich demonstriert haben, dass … | Dass ausgerechnet Anthropic diesen Machtanspruch mit „Cybersecurity“ begründet, nachdem sie regelmäßig öffentlich demonstriert haben,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dass ein Nutzer die vom Hersteller eingebauten Verhaltensregeln entfernen kann, gilt Anthropic als Sicherheitsproblem. Der Eigentümer des … | Dass ein Nutzer die vom Hersteller eingebauten Verhaltensregeln entfernen kann, gilt Anthropic als Sicherheitsproblem. Der Eigentümer… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Kontrolle durch Nutzer über ihre Systeme wird als Gefahr dargestellt. Kontrolle durch Anthropic über seine Nutzer wird als „Security“ verk… | Kontrolle durch Nutzer über ihre Systeme wird als Gefahr dargestellt. Kontrolle durch Anthropic über seine Nutzer wird als „Security“… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic nimmt ausgerechnet diesen Mechanismus und erklärt ihn zum Sicherheitsproblem. Weil Nutzer GLM selbst untersuchen und verändern k… | Anthropic nimmt ausgerechnet diesen Mechanismus und erklärt ihn zum Sicherheitsproblem. Weil Nutzer GLM selbst untersuchen und veränd… | ↗Not briefed |
rss marktechpost | 1mo | One Bad Prompt Took Down a Company’s Salesforce: RSA’s Jim Taylor on Agent ID and Taming the 4,000 Shadow AI Agents Hiding in Your Enterprise | RSA launched Agent ID at The AI Conference in San Francisco. It's an agentic identity security platform for finance, government, healt… | ↗Not briefed |
hn hn-ai | 1mo | Beyond LLMs: A Post-Transformer World Emerges | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Breaking: Anthropic has a new business plan. | Breaking: Anthropic has a new business plan. | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Alter Zero – RAM-efficient agent for coding, security and automation | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | yeah get their ass Wordle guy www.nytimes.com/2026/09/29/t... | yeah get their ass Wordle guy www.nytimes.com/2026/09/29/t... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic may not have intended to write an excellent ad for GLM 5.3, but that's what they did. www.anthropic.com/research/glm... Unlike … | Anthropic may not have intended to write an excellent ad for GLM 5.3, but that's what they did. www.anthropic.com/research/glm... Un… | ↗Not briefed |
bluesky bluesky-ai | 1mo | love that Anthropic is running ads for my bae GLM | love that Anthropic is running ads for my bae GLM | ↗Not briefed |
bluesky bluesky-ai | 1mo | I spent some time with OpenAI's new agent, Dots, and it's ... pretty good??? www.platformer.news/openai-dots-... | I spent some time with OpenAI's new agent, Dots, and it's ... pretty good??? www.platformer.news/openai-dots-... | ↗Not briefed |
hn hn-ai | 1mo | How do you deal with people who strongly disagree with your views? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | just because you lie to an LLM during training in order to get it to tell lies doesn't make those lies true | just because you lie to an LLM during training in order to get it to tell lies doesn't make those lies true | ↗Not briefed |
rss techmeme | 1mo | An interview with Bill Gates about the need for more than natural incentives under capitalism to regulate AI's risks, the effects AI will have on jobs, and more (New York Times) | New York Times: An interview with Bill Gates about the need for more than natural incentives under capitalism to regulate AI's risks,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't think this is the shape of mistake you're likely to see out of an LLM, tbh | I don't think this is the shape of mistake you're likely to see out of an LLM, tbh | ↗Not briefed |
bluesky bluesky-ai | 1mo | Well done Alan on a brilliant career & highly informative perspective on the economy #AusEcon and #finance. 🥂🍾 @alankohler.bsky.social @… | Well done Alan on a brilliant career & highly informative perspective on the economy #AusEcon and #finance. 🥂🍾 @alankohler.bsky.soc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | kind of going in the opposite direction as Anthropic bsky.app/profile/timk... | kind of going in the opposite direction as Anthropic bsky.app/profile/timk... | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLM이 도출하는 결과가 확률적이라면 정말 엄밀한 계측을 하려면 중요한 판단에는 대충 대규모로 루프를 돌려서 경향성을 봐야하나 같은 생각도 좀 들고. 웹개발이나 겜개발에선 눈에 보이는 결과가 바로 보이곤 했는데 이게 참 미묘하네 | LLM이 도출하는 결과가 확률적이라면 정말 엄밀한 계측을 하려면 중요한 판단에는 대충 대규모로 루프를 돌려서 경향성을 봐야하나 같은 생각도 좀 들고. 웹개발이나 겜개발에선 눈에 보이는 결과가 바로 보이곤 했는데 이게 참 미묘하네 | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI has my attention, I love circles | OpenAI has my attention, I love circles | ↗Not briefed |
hn hn-ai | 1mo | Want to protect yourself against cancer? Get vaccinated | — | ↗Not briefed |
hn hn-ai | 1mo | Thinking out loud AI and Cancer Cure | Anyone else ever think how every week there's another article and news story about how advanced AI is getting and how fast it… | ↗Not briefed |
bluesky bluesky-ai | 1mo | quallmer 0.5.0 is now on CRAN: the R package Seraphine Maerz and I are building for LLM-assisted qualitative coding. This release adds rob… | quallmer 0.5.0 is now on CRAN: the R package Seraphine Maerz and I are building for LLM-assisted qualitative coding. This release add… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Just assuming that "claude" completely agrees on every subjective element of my world view - well actually he probably would have assumed t… | Just assuming that "claude" completely agrees on every subjective element of my world view - well actually he probably would have assu… | ↗Not briefed |
bluesky bluesky-ai | 1mo | CLAUDE says these ribs are 7/10 zitron is right. Anthropic to jail | CLAUDE says these ribs are 7/10 zitron is right. Anthropic to jail | ↗Not briefed |
bluesky bluesky-ai | 1mo | When it takes a Chinese AI model GLM 5.2 to to help investigators understand an attack for a rogue American AI, you know you’ve got a probl… | When it takes a Chinese AI model GLM 5.2 to to help investigators understand an attack for a rogue American AI, you know you’ve got a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I miss programming, yelling at an LLM isn’t fun | I miss programming, yelling at an LLM isn’t fun | ↗Not briefed |
bluesky bluesky-ai | 1mo | LMFAO the official America.gov AI website uses a Chinese open source model | LMFAO the official America.gov AI website uses a Chinese open source model | ↗Not briefed |
hn hn-ai | 1mo | A New Type of Small Nuclear Reactor Gets the Green Light in Tennessee | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Using an LLM to write a webpage that looks like a 90s HTML1 page is the weirdest asthetic politic I can possibly imagine. | Using an LLM to write a webpage that looks like a 90s HTML1 page is the weirdest asthetic politic I can possibly imagine. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Incidentally you don’t have to go to the chans and kiwifarms of the world to find people casually engaging in LLM-assisted dossier compilat… | Incidentally you don’t have to go to the chans and kiwifarms of the world to find people casually engaging in LLM-assisted dossier com… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I would say they could take behavioral science more seriously (and they could) but that feels like a monkeys paw thing and then anthropic w… | I would say they could take behavioral science more seriously (and they could) but that feels like a monkeys paw thing and then anthro… | ↗Not briefed |
hn hn-ai | 1mo | Ask HNs: where can I get a valuation of my application? | I created a fixed asset software that handles the multistate tax practitioners needs and want to get it valued. Where can I get a fai… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Beware of The Village idiot and AI. OpenAI, the maker of ChatGPT, said Monday it will not release GPT-6.1 Astra, its next major AI model. …🔧 ChatGPT | Beware of The Village idiot and AI. OpenAI, the maker of ChatGPT, said Monday it will not release GPT-6.1 Astra, its next major AI mo… | ↗Not briefed |
hn hn-ai | 1mo | The Model Is the Easy Half | — | ↗Not briefed |