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 |
|---|---|---|---|---|
arxiv arxiv-ai | 1mo | How Inefficient Is Natural Gradient Descent? From Exact Optimality to Θ( \sqrt{ \log d } ) Divergence🔧 Gamma | Natural gradient descent (NGD) underlies common methods in ML. For dually flat families, idealized NGD on the forward Kullback--Leible… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Minimal Witness Reinforcement Learning | ``What are the irreducible conditions that are sufficient to produce an outcome?'' is one of the most common questions that recur acro… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic prob not buying any more of that sweet, sweet, SpaceX AI overcapacity... $SPCX $TSLA | Anthropic prob not buying any more of that sweet, sweet, SpaceX AI overcapacity... $SPCX $TSLA | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I made a Linux history video using "video as code" | 35 years ago today (Oct 5, 1991), Linux 0.02 was released.<p>I made an animated history of Linux using Claude Code + opus 5.5.<p>I… | ↗Not briefed |
bluesky bluesky-ai | 1mo | There's a reddit discussion going on because of course there is. Speculation is 1. It guessed correctly 2. Disclosed by user/found externa… | There's a reddit discussion going on because of course there is. Speculation is 1. It guessed correctly 2. Disclosed by user/found ex… | ↗Not briefed |
hn hn-ai | 1mo | Jiti – Grow a running Lisp application by talking to it | — | ↗Not briefed |
hn hn-ai | 1mo | Terascale Query Processing in the Browser: Making GPU Acceleration Accessible [pdf] | — | ↗Not briefed |
hn hn-ai | 1mo | Every Ways to Get 32GB VRAM for Local AI at Full Context [video] | — | ↗Not briefed |
hn hn-ai | 1mo | Evaluating Memory Structure in LLM Agents | — | ↗Not briefed |
rss techmeme | 1mo | Sources: Meta and Microsoft are working to cut their employees' use of Claude; Meta employees using Claude Code have dropped to ~30K from ~60K earlier this year (The Information) | The Information: Sources: Meta and Microsoft are working to cut their employees' use of Claude; Meta employees using Claude Code have… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Warrior Cats: Clans of the Forest Bamtang Games, Trailmark Games. No disclosure, Concept references, coding / programming. -y andina.pe/age… | Warrior Cats: Clans of the Forest Bamtang Games, Trailmark Games. No disclosure, Concept references, coding / programming. -y andina.p… | ↗Not briefed |
rss techcrunch-ai | 1mo | TikTok rolls out an AI shopping assistant and one-click checkout | TikTok describes its new Shopping Assistant as a conversational AI agent designed to help users discover and purchase products. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Story aside, I didn’t expect Florida woman who threatens to shoot up sheriffs office to be a Claude user. Would have guessed Grok. www.tech… | Story aside, I didn’t expect Florida woman who threatens to shoot up sheriffs office to be a Claude user. Would have guessed Grok. www… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "You should accept that your car is gonna explode, once in awhile." "You should accept that your food will be bad enough to kill your chil… | "You should accept that your car is gonna explode, once in awhile." "You should accept that your food will be bad enough to kill your… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Because the LLM isn't 'creating' or 'learning', it's assembling. Even if 'all' the person does is rearrange, creating a collage, found po… | Because the LLM isn't 'creating' or 'learning', it's assembling. Even if 'all' the person does is rearrange, creating a collage, fou… | ↗Not briefed |
bluesky bluesky-ai | 1mo | On one hand, it is SUPER weird to yell at and be cruel to machines like SIri or whatever bc whatever the machine did it absolutely could no… | On one hand, it is SUPER weird to yell at and be cruel to machines like SIri or whatever bc whatever the machine did it absolutely cou… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've been really disgusted at the whole "consulting vile homophobes and misogynists" thing but tbh it slightly raised my opinion on anthrop… | I've been really disgusted at the whole "consulting vile homophobes and misogynists" thing but tbh it slightly raised my opinion on an… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Use macOS shortcuts on Windows | Hi HN!<p>In my day to day, I switch between my Windows PC and Macbook often. One thing I find annoying is having to change what keyboa… | ↗Not briefed |
hn hn-ai | 1mo | GPT-6 Astra Built a 300K+ LOC Starship in One HTML File | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | it's instructive to go back and look at the kinds of threats OpenAI was anticipating would stem from LLMs back when they first started buil… | it's instructive to go back and look at the kinds of threats OpenAI was anticipating would stem from LLMs back when they first started… | ↗Not briefed |
rss the-decoder | 1mo | Reka AI's omni-model Rho-1 handles text, images, video, and robot control in a single model | Reka AI's Rho-1 is a 19-billion-parameter omni-model that processes and generates text, images, video, and robot control actions in a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI recently published data on its researchers’ coding-agent usage. We took their weekly figures from January to mid-August 2026 and fit… | OpenAI recently published data on its researchers’ coding-agent usage. We took their weekly figures from January to mid-August 2026 an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | META AND MICROSOFT ARE WORKING TO WEAN STAFF OFF ANTHROPIC'S CLAUDE, WITH MICROSOFT CUTTING ITS INTERNAL CLAUDE SPENDING LEVEL BY OVER A TH… | META AND MICROSOFT ARE WORKING TO WEAN STAFF OFF ANTHROPIC'S CLAUDE, WITH MICROSOFT CUTTING ITS INTERNAL CLAUDE SPENDING LEVEL BY OVER… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Grim thoughts from a fellow developer about specifically one game/engine but with clear parallels to the industry at large. #gamedev | Grim thoughts from a fellow developer about specifically one game/engine but with clear parallels to the industry at large. #gamedev | ↗Not briefed |
rss techmeme | 1mo | Ghost, which makes a $3,499 computer designed for AI agents and includes an RTX Pro 4000 SFF Blackwell GPU, emerges from stealth with an $11M seed led by a16z (Dominic-Madori Davis/TechCrunch) | Dominic-Madori Davis / TechCrunch: Ghost, which makes a $3,499 computer designed for AI agents and includes an RTX Pro 4000 SFF Blackw… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It's launching textGrain watermarking in the EU, but OpenAI says it won’t be the ‘global default’ yet. www.theverge.com/ai-artificia... | It's launching textGrain watermarking in the EU, but OpenAI says it won’t be the ‘global default’ yet. www.theverge.com/ai-artificia... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Public Advocate Jumaane Williams says OpenAI's first answer to today's NYC City Council hearing on AI risks was the "most God-awful, fantas… | Public Advocate Jumaane Williams says OpenAI's first answer to today's NYC City Council hearing on AI risks was the "most God-awful, f… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Anyone Accepted into Codex for Open Source? | Anyone been successfully accepted into Codex for Open Source? (https://openai.com/form/codex-for-oss/)<p>I wa… | ↗Not briefed |
hn hn-ai | 1mo | 'Absolute s–t': EU's Microsoft Teams alternative gets bad reviews from officials | — | ↗Not briefed |
hn hn-ai | 1mo | Reika – A coding agent CLI designed around small local models first | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | would all be just and fine if the AI company had paid for one copy of the book? (tbh they should have, they could certainly afford that!) o… | would all be just and fine if the AI company had paid for one copy of the book? (tbh they should have, they could certainly afford tha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | you dont actually need to train on millions of movies, videos, and novels to build laundry bots (or to produce an entirely capable Claude C… | you dont actually need to train on millions of movies, videos, and novels to build laundry bots (or to produce an entirely capable Cla… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Rashomon – An independent execution record for AI coding agents | I have been working on Rashomon, an open-source execution recorder for AI coding agents.<p>The basic idea is that the agent transcript… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This newsletter has become increasingly common in recent years, promoted by venture investors and savvy transactional attorneys. www.bloomb… | This newsletter has become increasingly common in recent years, promoted by venture investors and savvy transactional attorneys. www.b… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Clammy Sam" Altman out here doing his best Lord Farquaad impersonation. "Some of you may have your life's work stolen or develop 'AI' psyc… | "Clammy Sam" Altman out here doing his best Lord Farquaad impersonation. "Some of you may have your life's work stolen or develop 'AI'… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Altman names Macron as the world leader with the most impressive approach to artificial intelligence, praising his safety focus. The OpenAI… | Altman names Macron as the world leader with the most impressive approach to artificial intelligence, praising his safety focus. The O… | ↗Not briefed |
bluesky bluesky-ai | 1mo | If only AI researcher spin could be harnessed and made into clean power. Perhaps we could build a large number of human-sized hamster whe… | If only AI researcher spin could be harnessed and made into clean power. Perhaps we could build a large number of human-sized hamste… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sam Altman says Elon Musk is a 'bully' who only respects people who punch him back https://bit.ly/3VSuxp3 | Sam Altman says Elon Musk is a 'bully' who only respects people who punch him back https://bit.ly/3VSuxp3 | ↗Not briefed |
hn hn-ai | 1mo | Sdrtop – Terminal monitor for SDR hardware | — | ↗Not briefed |
rss theverge-ai | 1mo | OpenAI is adding text watermarking in ChatGPT and Codex🔧 ChatGPT | An invisible, machine-readable watermark in text output is rolling out to ChatGPT and Codex, but only for users in the European Union… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dinowitz asks a good ? about "A.i." directed at kids, mentions cognitive decline, cheating. OpenAI lady keeps saying "OpenAI takes [X] ver… | Dinowitz asks a good ? about "A.i." directed at kids, mentions cognitive decline, cheating. OpenAI lady keeps saying "OpenAI takes [X… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 👀 per CNBC: NYC AI hearing: *Anthropic whistleblower Jacob Coxon, who raised early concerns called out the industry’s “extremely reckle… | 👀 per CNBC: NYC AI hearing: *Anthropic whistleblower Jacob Coxon, who raised early concerns called out the industry’s “extremely r… | ↗Not briefed |
hn hn-ai | 1mo | The-engineering-behind-LLM-inference | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is EXACTLY what large language model (LLM) AIs are doing—and they consume huge amounts of power and resources to do so. When I was a … | This is EXACTLY what large language model (LLM) AIs are doing—and they consume huge amounts of power and resources to do so. When I w… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I was thinking about the Anthropic, SpaceX, and OpenAI IPOs. What this might be a harbinger of is the end of public markets as a mechanism … | I was thinking about the Anthropic, SpaceX, and OpenAI IPOs. What this might be a harbinger of is the end of public markets as a mecha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Not sure what to tell you, but LLM are literally “using ML in linguistics”! They evolved from earlier approaches in ML for natural language… | Not sure what to tell you, but LLM are literally “using ML in linguistics”! They evolved from earlier approaches in ML for natural lan… | ↗Not briefed |
arxiv arxiv-ai | 1mo | One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline | Pipeline figures in ML papers must be repurposed across many canvases, including paper columns, 16:9 slides, portrait posters, 1:1 soc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Base Models Can Reason By Taking a Cue From Training Data🔧 Elicit | In this paper, we study how training data creates associations between the tokens at the start of a base model's response and the reas… | ↗Not briefed |
arxiv arxiv-ai | 1mo | BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance | Worst-group accuracy (WGA) evaluates a trained predictor but does not characterize how its frozen backbone behaves when a new head is… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning to Read the Contextual Tokens in Diffusion Transformers | Multimodal Diffusion Transformers (MM-DiTs) jointly process visual and textual representations throughout generation. These models rep… | ↗Not briefed |