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 | 17d | Anthropic's S1 just dropped, thanks to my colleague Anissa Gardizy www.wsj.com/tech/ai/anth... | Anthropic's S1 just dropped, thanks to my colleague Anissa Gardizy www.wsj.com/tech/ai/anth... | ↗Not briefed |
hn hn-ai | 17d | You can now enjoy TUI Radio with all the best features | — | ↗Not briefed |
bluesky bluesky-ai | 17d | Get powered by Z. Humans are cooked 😎 | Get powered by Z. Humans are cooked 😎 | ↗Not briefed |
bluesky bluesky-ai | 17d | Anthropic outlines its approach to "responsible data centre development", and........explicitly promises to carpet bomb communities with ma… | Anthropic outlines its approach to "responsible data centre development", and........explicitly promises to carpet bomb communities wi… | ↗Not briefed |
bluesky bluesky-ai | 17d | I've been thinking a lot about this demo of a discriminative model (read: not an LLM, not generative AI) that can rapidly make enough decis… | I've been thinking a lot about this demo of a discriminative model (read: not an LLM, not generative AI) that can rapidly make enough… | ↗Not briefed |
bluesky bluesky-ai | 17d | "Anthropic... said around one-fifth of its $16.3 billion in revenue in the first half of 2026 came from just two customers. The company exp… | "Anthropic... said around one-fifth of its $16.3 billion in revenue in the first half of 2026 came from just two customers. The compan… | ↗Not briefed |
bluesky bluesky-ai | 17d | Narrator: Anthropic hoped to turn the corner, but they weren't able to | Narrator: Anthropic hoped to turn the corner, but they weren't able to | ↗Not briefed |
bluesky bluesky-ai | 17d | Anthropic is on the hook to pay around $450 billion in cloud computing agreements over the next seven to ten years - WSJ | Anthropic is on the hook to pay around $450 billion in cloud computing agreements over the next seven to ten years - WSJ | ↗Not briefed |
bluesky bluesky-ai | 17d | *ANTHROPIC BOOKS $80.7B NET LOSS IN 1H ON ACCOUNTING CHARGES:WSJ | *ANTHROPIC BOOKS $80.7B NET LOSS IN 1H ON ACCOUNTING CHARGES:WSJ | ↗Not briefed |
bluesky bluesky-ai | 17d | A local channel is using AI to promote shows. Stores are using AI to advertize their products. Our work is using AI in its presentations. E… | A local channel is using AI to promote shows. Stores are using AI to advertize their products. Our work is using AI in its presentatio… | ↗Not briefed |
arxiv arxiv-ai | 17d | A General Topological Symmetry Index for Critical Orbits In Neural Network Loss Landscapes | A systematic understanding of the global structure of neural-network loss landscapes remains challenging due to their high dimensional… | ↗Not briefed |
bluesky bluesky-ai | 17d | In a new draft prospectus viewed by the Wall Street Journal, Anthropic discloses an $80.7 billion net loss for the first half of 2026 and a… | In a new draft prospectus viewed by the Wall Street Journal, Anthropic discloses an $80.7 billion net loss for the first half of 2026… | ↗Not briefed |
bluesky bluesky-ai | 17d | Anthropic S-1 leaks from the WSJ - $16.3bn in 1H revenue (vs $1.1bn in 1H '25) - Nearly $9bn in compute spend - $450bn of future compute o… | Anthropic S-1 leaks from the WSJ - $16.3bn in 1H revenue (vs $1.1bn in 1H '25) - Nearly $9bn in compute spend - $450bn of future comp… | ↗Not briefed |
bluesky bluesky-ai | 17d | OpenAI Reportedly Looking to Hurry and Get Valued at $1.4 Trillion Before Anthropic IPO https://gizmodo.com/openai-reportedly-looking-to-hu… | OpenAI Reportedly Looking to Hurry and Get Valued at $1.4 Trillion Before Anthropic IPO https://gizmodo.com/openai-reportedly-looking-… | ↗Not briefed |
hn hn-ai | 17d | Digital Freedom Across the World | — | ↗Not briefed |
bluesky bluesky-ai | 17d | I had this debate with my kids over Itch's AI policy. They require generative AI disclosure. If you build your own game engine with LLM as… | I had this debate with my kids over Itch's AI policy. They require generative AI disclosure. If you build your own game engine with L… | ↗Not briefed |
hn hn-ai | 17d | Anthropic S-1 Shows Red-Hot Growth, Risks and Computing Costs | — | ↗Not briefed |
rss simon-willison | 17d | Battle Drones - a vibe coded RTS with audio-based multiplayer | — | ↗Not briefed |
bluesky bluesky-ai | 17d | Wall Street Journal says Anthropic had a net loss of more than $80 billion in the first half of this year. | Wall Street Journal says Anthropic had a net loss of more than $80 billion in the first half of this year. | ↗Not briefed |
bluesky bluesky-ai | 17d | Exclusive: Anthropic's draft IPO filing reveals red-hot growth, massive computing costs and an $80 billion net loss. Here are the highlight… | Exclusive: Anthropic's draft IPO filing reveals red-hot growth, massive computing costs and an $80 billion net loss. Here are the high… | ↗Not briefed |
bluesky bluesky-ai | 17d | Anthropic posts net loss exceeding $80 billion in first half of this year: WSJ | Anthropic posts net loss exceeding $80 billion in first half of this year: WSJ | ↗Not briefed |
arxiv arxiv-ai | 17d | Agent AI Zarządzający Siecią Open RAN Zorientowana na Użytkownika | Jednym z najważniejszych wyzwań związanych z implementacją sieci Open RAN o architekturze zorientowanej na użytkownika jest odpowiedni… | ↗Not briefed |
rss latent-space | 17d | The Programmable Data Center — Evan Conrad, SF Compute | From GPU trading to physical settlement, independent markets, and programmable compute | ↗Not briefed |
bluesky bluesky-ai | 17d | Oh, what could POSSIBLY go wrong. www.washingtonpost.com/health/2026/... | Oh, what could POSSIBLY go wrong. www.washingtonpost.com/health/2026/... | ↗Not briefed |
bluesky bluesky-ai | 17d | The running theme is that the tech is garbage, but they still keep releasing it to the public and letting it cause damage. Once again, an L… | The running theme is that the tech is garbage, but they still keep releasing it to the public and letting it cause damage. Once again,… | ↗Not briefed |
arxiv arxiv-ai | 17d | DualHeads: Reward-tilted few-step diffusion made simple | Distilled diffusion models generate high-quality images in a few sampling steps, but aligning them with task requirements and human pr… | ↗Not briefed |
bluesky bluesky-ai | 17d | A new AI tool that identifies genetic risk factors for stroke could open the door to more personalised prevention and treatment. | A new AI tool that identifies genetic risk factors for stroke could open the door to more personalised prevention and treatment. | ↗Not briefed |
arxiv arxiv-ai | 17d | Speedrunning Relational Foundation Model Pretraining | Progress on foundation model pretraining is gated by who can afford a trial, and relational pretraining is no exception. Recipes are d… | ↗Not briefed |
hn hn-ai | 17d | GTA 6 Digital Pre-Orders Climb to 6.45M, PlayStation Holds 74% Share | — | ↗Not briefed |
bluesky bluesky-ai | 17d | I don't advocate people using LLMs for anything they're doing but I do recommend the "but LLMs are good for programming crowd" go to chatgp…🔧 ChatGPT | I don't advocate people using LLMs for anything they're doing but I do recommend the "but LLMs are good for programming crowd" go to c… | ↗Not briefed |
hn hn-ai | 17d | Blanking Beavers: Results and Open Problems | — | ↗Not briefed |
bluesky bluesky-ai | 17d | An online friend who is not in Bluesky and thus can be forgiven for not knowing my stance on LLM usage breathlessly told me that I could us…🔧 ChatGPT | An online friend who is not in Bluesky and thus can be forgiven for not knowing my stance on LLM usage breathlessly told me that I cou… | ↗Not briefed |
hn hn-ai | 17d | GitHub Actions Feedback Survey | — | ↗Not briefed |
rss techmeme | 17d | Amazon plans to hire 300K+ US workers for seasonal, part-time, and full-time roles for the upcoming holiday season, up 20% from 2025, paying $20/hour on average (Spencer Soper/Bloomberg) | Spencer Soper / Bloomberg: Amazon plans to hire 300K+ US workers for seasonal, part-time, and full-time roles for the upcoming holiday… | ↗Not briefed |
arxiv arxiv-ai | 17d | Self-Supervised Tuning of Low-Fidelity Optimization Models via High-Fidelity Feedback | Low-fidelity (LF) optimization models are widely used in large-scale operations because they can be solved efficiently and reliably, b… | ↗Not briefed |
arxiv arxiv-ai | 17d | When Do Gaussian Priors Fail? A Latent Geometry Investigation of RL Fine-Tuning | Latent diffusion models almost universally place a Gaussian prior on the VAE latent space. For base generation this default is unprobl… | ↗Not briefed |
bluesky bluesky-ai | 17d | Claude unveils crypto's seed secrets Anthropic's Claude AI model reconstructed the origins of numerical seeds for ECDSA security standards… | Claude unveils crypto's seed secrets Anthropic's Claude AI model reconstructed the origins of numerical seeds for ECDSA security stan… | ↗Not briefed |
hn hn-ai | 17d | Show HN: Apple Reminders MCP – native read/write via ReminderKit | Create a MacOS mcp bridge for agents to control app reminders, more functions than similar projects. see <a href="https://gi… | ↗Not briefed |
arxiv arxiv-ai | 17d | Beyond Correct Answers: How Compositional Cues Shape Visual Judgments | Visual question answering requires models to assign attributes and relations to the correct objects, a problem known as visual binding… | ↗Not briefed |
bluesky bluesky-ai | 17d | One of my pet gripes right now is that people have allowed themselves to be massively distracted by LLM nonsense when social media has been… | One of my pet gripes right now is that people have allowed themselves to be massively distracted by LLM nonsense when social media has… | ↗Not briefed |
bluesky bluesky-ai | 17d | From what I could gather it was "hey LLM, I'm doing a CTF, can you get the private keys" and not even having time to get up for a cuppa bef… | From what I could gather it was "hey LLM, I'm doing a CTF, can you get the private keys" and not even having time to get up for a cupp… | ↗Not briefed |
arxiv arxiv-ai | 17d | KIRA: A Full-Stack Reading Screening System with Dual-Pathway Analysis for Children | Early-grade reading screening is critical for identifying children at risk of reading difficulty. Existing AI-based screeners are ofte… | ↗Not briefed |
hn hn-ai | 17d | Cadence an open-source, on-device daily check-in for iOS | — | ↗Not briefed |
bluesky bluesky-ai | 17d | Actually did a short test, and seems its more or less universal (delve.town/profile/did:...) | Actually did a short test, and seems its more or less universal (delve.town/profile/did:...) | ↗Not briefed |
bluesky bluesky-ai | 17d | Mercury 3 makes diffusion LLMs competitive Inception Labs has released Mercury 3, a diffusion-based LLM that it claims outperforms GPT-6 L… | Mercury 3 makes diffusion LLMs competitive Inception Labs has released Mercury 3, a diffusion-based LLM that it claims outperforms GP… | ↗Not briefed |
hn hn-ai | 17d | Show HN: A daily game where the goal is to be unpredictable to an LLM | — | ↗Not briefed |
hn hn-ai | 17d | Anthropic redesigned source-control for AI agents | — | ↗Not briefed |
arxiv arxiv-ai | 17d | How to Use Parallel Data for Multilingual Pretraining | High-quality human-translated parallel corpora can improve cross-lingual transfer, but are typically too small for large-scale pretrai… | ↗Not briefed |
arxiv arxiv-ai | 17d | Sketch2Sim: an end-to-end modular framework for hand-drawn circuit analysis | This study proposes a sketch-to-simulation system to solve handwritten circuit diagrams. The system converts raw images into a Simulat… | ↗Not briefed |
hn hn-ai | 17d | Cold starting a 32B model in 1.1s with GPU checkpoint/restore | — | ↗Not briefed |