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 |
|---|---|---|---|---|
rss techmeme | 1mo | Sources: US banking regulators are questioning banks about their deployment of agentic AI, use of Chinese-developed AI models, and autonomous system safeguards (Reuters) | Reuters: Sources: US banking regulators are questioning banks about their deployment of agentic AI, use of Chinese-developed AI models… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: 4B VLM trained on 600k questions to query images in one pass | Heyo!<p>I'm Adit, an undergrad in Finland. This is a preview of a side project I worked on last weekend. Essentially, Dragonfly i… | ↗Not briefed |
hn hn-ai | 1mo | Tilefinch – An experimental web browser for the Sony PSP | — | ↗Not briefed |
hn hn-ai | 1mo | An agent says "done." Which files did it test? Who reviewed the work? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | ai Anthropic warns AI models could soon copy themselves across home devices | ai Anthropic warns AI models could soon copy themselves across home devices | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hmm, yes, a multi-terabyte LLM should fit in an LG smart refrigerator just fine | Hmm, yes, a multi-terabyte LLM should fit in an LG smart refrigerator just fine | ↗Not briefed |
hn hn-ai | 1mo | Signed | — | ↗Not briefed |
hn hn-ai | 1mo | Clean room Adobe/Microsoft forks are rapidly improving | — | ↗Not briefed |
rss the-decoder | 1mo | Anthropic says it cut Claude reasoning token theft by up to 99% in fight against industrial-scale distillation | Anthropic says it has cut the extraction of Claude's internal reasoning tokens by 90 to 99 percent. The company links the large-scale… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI a(urait) exhibé un algorithme de calcul de la FFT galactique. asymptotiquement meilleur que l'existant, mais probablement totalement… | OpenAI a(urait) exhibé un algorithme de calcul de la FFT galactique. asymptotiquement meilleur que l'existant, mais probablement total… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Ret – An inbox and canvas for your Codex and Claude Code agents | Over the last few weeks I've really been using Codex and Claude to solve a lot of parallel tasks, across a lot of different proje… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Anthropic warns"???? The media is poison. | "Anthropic warns"???? The media is poison. | ↗Not briefed |
bluesky bluesky-ai | 1mo | By the way – ever wondered, why the Anthropic paper is 99% about open weight models? Probably because it's just a bullshit lobby paper, to… | By the way – ever wondered, why the Anthropic paper is 99% about open weight models? Probably because it's just a bullshit lobby pape… | ↗Not briefed |
rss techmeme | 1mo | The IMF lifts its 2026 global GDP growth forecast to 3.1%, up from 3% in July, citing an AI boom that offsets Middle East war disruptions and high energy costs (Anniek Bao/CNBC) | Anniek Bao / CNBC: The IMF lifts its 2026 global GDP growth forecast to 3.1%, up from 3% in July, citing an AI boom that offsets Middl… | ↗Not briefed |
rss wired-ai | 1mo | Wikipedians Believed AI Would Come and Go. They Were Wrong | In a world of AI Overviews, Wikipedia usage declines. But the real loss may be far more existential than website traffic. | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Anima – Interactive, cited biology figures using Claude Code | — | ↗Not briefed |
hn hn-ai | 1mo | Veda: The Agentic Hobby OS Added Video Player | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | 「LLMを乱用するのはよくない」という主張と「LLMに道具として有用性がある」という話は両立するので……。 | 「LLMを乱用するのはよくない」という主張と「LLMに道具として有用性がある」という話は両立するので……。 | ↗Not briefed |
bluesky bluesky-ai | 1mo | 「LLMは熟練者が使うと役には立つよ」という話はしていますけど「Googleで検索するとLLMが勝手に要約をまとめてくるのは、明らかにエネルギーや水資源の無駄だろう、全部コタツ記事の転載じゃん」とは思っている。 | 「LLMは熟練者が使うと役には立つよ」という話はしていますけど「Googleで検索するとLLMが勝手に要約をまとめてくるのは、明らかにエネルギーや水資源の無駄だろう、全部コタツ記事の転載じゃん」とは思っている。 | ↗Not briefed |
bluesky bluesky-ai | 1mo | Not in USA, they won't have much research left if this continues. Just Super Intelligence running in circles, screening LLM-written lecture… | Not in USA, they won't have much research left if this continues. Just Super Intelligence running in circles, screening LLM-written le… | ↗Not briefed |
rss techmeme | 1mo | How crypto billionaires Christopher Harborne, who owns ~12% of Tether, and Ben Delo donated ~$100M to Nigel Farage's Reform, the largest donations in UK history (David Segal/New York Times) | David Segal / New York Times: How crypto billionaires Christopher Harborne, who owns ~12% of Tether, and Ben Delo donated ~$100M to Ni… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Blobz – a Deterministic, Playful SVG avatars generated from a string | — | ↗Not briefed |
hn hn-ai | 1mo | LLM-inspired study sheds light on how brain cells encode the meaning of words | — | ↗Not briefed |
hn hn-ai | 1mo | Tian Keyu Model-Training Sabotage Incident: Victim Accounts and Evidence | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The speed with which the media caved to Trump and started calling it superintelligence is honestly a little astonishing considering that th… | The speed with which the media caved to Trump and started calling it superintelligence is honestly a little astonishing considering th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | So anthropic are creating sophisticated viruses? I remember when that sort of thing was illegal. | So anthropic are creating sophisticated viruses? I remember when that sort of thing was illegal. | ↗Not briefed |
rss techmeme | 1mo | Accenture's market cap rebounded from $76B in June to ~$130B ahead of its investor day, where it seeks to reassure investors that AI won't threaten its business (Stephen Foley/Financial Times) | Stephen Foley / Financial Times: Accenture's market cap rebounded from $76B in June to ~$130B ahead of its investor day, where it seek… | ↗Not briefed |
rss the-decoder | 1mo | Anthropic study warns that small AI models could soon replicate across everyday devices🔧 Replicate | Small AI models may be just months away from spreading on their own across poorly secured consumer devices like smart TVs and robot va… | ↗Not briefed |
rss theverge-ai | 1mo | Character.AI files for bankruptcy🔧 Character.AI | Character.AI, one of the first breakout startups of the generative AI boom, has filed for bankruptcy and is selling some of its assets… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is a free app. This doesn't equate to ROI. Meta could charge *millions* of paying customers double digit dollars monthly and still not… | This is a free app. This doesn't equate to ROI. Meta could charge *millions* of paying customers double digit dollars monthly and stil… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sorry? Anthropic warn us, like it’s nothing to do with them? What the fuck do they expect US to do about it? | Sorry? Anthropic warn us, like it’s nothing to do with them? What the fuck do they expect US to do about it? | ↗Not briefed |
bluesky bluesky-ai | 1mo | We're pleased to share that Alex Gillespie, Tom Reader, Alex Goddard and Hannah Bunt have received a grant from Anthropic to use Claude Sci… | We're pleased to share that Alex Gillespie, Tom Reader, Alex Goddard and Hannah Bunt have received a grant from Anthropic to use Claud… | ↗Not briefed |
hn hn-ai | 1mo | 11 year old unsoundness bugs found in Rust borrow checker | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: shadowclone, skills learned from your coding agent corrections | shadowclone reads the coding-agent sessions already on your disk, finds where you corrected the agent and turns those corrections into… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic warns rogue AI agents may ‘self-replicate’ to hijack everyday devices ft.trib.al/iDXGjT7🔧 Replicate | Anthropic warns rogue AI agents may ‘self-replicate’ to hijack everyday devices ft.trib.al/iDXGjT7 | ↗Not briefed |
rss techmeme | 1mo | Meta confirms Muse lets users "engage in adult conversations, if they choose, using a private and secure VM", saying it doesn't block "legal adult behavior" (Reece Rogers/Wired) | Reece Rogers / Wired: Meta confirms Muse lets users “engage in adult conversations, if they choose, using a private and secure VM”, sa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Reminder: Anthropic sets the conditions for which its models operate. Instead of issuing warnings, Anthropic could also change its behavior… | Reminder: Anthropic sets the conditions for which its models operate. Instead of issuing warnings, Anthropic could also change its beh… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i fucking hate this fucking world fuck your ai games fuck claude, fuuuuuck your ai driven development you're not engineers you're managers … | i fucking hate this fucking world fuck your ai games fuck claude, fuuuuuck your ai driven development you're not engineers you're mana… | ↗Not briefed |
hn hn-ai | 1mo | Synthwave-Tinted Hyprland Desktop for Apple Silicon Macs and Fedora Asahi Remix | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic and OpenAI have sought significant control over what outside evaluators can publish, concerned that findings could reveal trade s… | Anthropic and OpenAI have sought significant control over what outside evaluators can publish, concerned that findings could reveal tr… | ↗Not briefed |
hn hn-ai | 1mo | Know Your Mapper | — | ↗Not briefed |
rss techmeme | 1mo | Alibaba unveils its next-gen Qwen Glasses N1, starting at ~$450, with 50MP cameras, sensors, and eye-tracking on the ~$520 Pro model, and Qwen Clip AI earbuds (Bloomberg) | Bloomberg: Alibaba unveils its next-gen Qwen Glasses N1, starting at ~$450, with 50MP cameras, sensors, and eye-tracking on the ~$520… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Use your OpenAI subscription to create audiobooks | — | ↗Not briefed |
hn hn-ai | 1mo | AI models may soon self-replicate without access to data centers🔧 Replicate | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI is racing ahead. Are the safeguards keeping up? Today, the Business, Innovation, Science and Trade Committee is putting OpenAI, Anthrop… | AI is racing ahead. Are the safeguards keeping up? Today, the Business, Innovation, Science and Trade Committee is putting OpenAI, An… | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLM Large Loveable Mick | LLM Large Loveable Mick | ↗Not briefed |
rss techmeme | 1mo | SpaceXSI launches a unified X Subscription bundling Grok, Cursor, and X Premium, with tiers at $8, $30, $100, and $200 per month; the $30 tier includes Grok Bot (Mariella Moon/Engadget) | Mariella Moon / Engadget: SpaceXSI launches a unified X Subscription bundling Grok, Cursor, and X Premium, with tiers at $8, $30, $100… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Several commenters flagged a trend of recent LLM-assisted "ports" that claim native recompilation while quietly bundling emulators. Recompi… | Several commenters flagged a trend of recent LLM-assisted "ports" that claim native recompilation while quietly bundling emulators. Re… | ↗Not briefed |
rss mit-tech-review-ai | 1mo | Open models have real benefits. Do they outweigh the risks? | MIT Technology Review Explains: Let our writers untangle the complex, messy world of science and technology to help you understand wha… | ↗Not briefed |
rss wired-ai | 1mo | The Biggest Pro-AI Super PAC Network Has Poured Millions Into Nevada Shell Companies | Backed by Silicon Valley titans, Leading the Future has raised huge sums to promote pro-AI candidates on both side of the aisle. | ↗Not briefed |