Article archive
Browse individual source reports, including coverage still awaiting the summary or community analysis needed for our story feed.
5,367 articles · page 24 of 108
Browse topics
Anthropic Is Building a Predictive Surveillance System to Monitor Activists
6 min read
Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitoring system to keep tabs on activists who oppose the rapid development of artificial intelligence. In addition to monitoring activists in the vicinity of Anthropic executives and keeping tabs on protests near physical Anthropic assets, the firm is also implementi...
Understanding the recent DDoS attack against Read the Docs
10 min read
Which site best fits your project? Free for open-source software. Sign upPlans starting at $50/month. Sign upWhich site best fits your project? Free for open-source software. Sign upPlans starting at $50/month. Sign upAnalytics from the June 2026 attack In mid-to-late June 2026, Read the Docs experienced the largest and most sophisticated distributed denial-of-service (DDoS) attack in our history....
GPT-6 Astra, looped transformers, and hidden reasoning
30 min read
A lot has happened in the last few weeks. I am sure that OpenAI’s GPT-6 Astra is top of mind for everyone right now. In particular, thoughts on its performance, the looped transformer/recurrent depth aspects, and rumors that Astra is “hiding” its reasoning trace (i.e., chain of thought). So, in this article, I want to start with some brief impressions of Astra and some thoughts on where all this i...
Playing whack-a-mole is losing
13 min read
CLOSURE I want to juxtapose two classes of people in the security industry on the defender side of security engineering for products. The first are the Security as Identity people. This goes back to the security-as-counterculture attitude, glamorized in Hackers and dating back to phone phreaking. It reflects Thompson’s “Workism”, where work is not just a way to earn a living, but a way to build a ...
What will our economic future look like?
13 min read
What will our economic future look like? We don’t know yet how AI will reshape the economy. Will it lead to unprecedented growth? Widespread unemployment? Neither, or something else? How can we tell? Anthropic’s Economics team built a model of how AI might affect jobs, growth, and unemployment in the US in coming years. Read about possible economic futures and make your own predictions about AI ca...
Tailwind Labs is joining Shopify
2 min read
Big one today — Tailwind is joining Shopify. When I started working on Tailwind over nine years ago, my only goal was to create something that would make it easier to build beautiful interfaces for my own projects. Fast-forward to today and the framework is installed over 110 million times per week and is trusted by many of the world's biggest companies to style products like ChatGPT, X, Cloudflar...
How I advertise malicious software on Google Ads
5 min read
How I advertise malicious software on Google Ads A caller phones Radio Erywan. “Is it true that in Moscow, on Red Square, they give out cars?” “True. Not in Moscow, however, but in St. Petersburg. Not on Red Square, but on Revolution Square. Not cars, but bikes. And they don’t give them out, but steal them”. Edit: Through the apparent magic of Hacker News, my Google Ads account has been reinstated...
Desert Ant Labs: local, fast models that run on device
6 min read
On-device intelligence for every product Today we're launching Desert Ant Labs, a European frontier AI lab building opinionated on-device intelligence. We believe the best path to efficient intelligence starts on-device. We're building small, specialized models for audio, vision, and text – each model answers in milliseconds, and costs nothing to run, so you can put intelligence in every product i...
How An AI math breakthrough ignited a controversy
9 min read
Nearly 200 years ago, Claude-Louis Navier and George Gabriel Stokes sought to describe the chaotic motion of fluids mathematically, writing down equations that are still routinely used by everyone from aircraft designers to climatologists. But the formulas they developed weren’t guaranteed to describe all possible conditions. Mathematicians have spent decades trying to either prove that the equati...
I resigned from Anthropic today (Jacob Coxon)
2 min read
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below. Sep 9, 2026 · 12:04 AM UTC 14,831 130,553 655,249 121,928,912 Do not underestimate the power of this technology. These will soon be...
Claude, change the "Add to Cart" button to blue
1 min read
Opusfived Make the “Add to Cart” button blue. Do not let Claude change anything else. Let’s go About & Terms
Gambling with our lives: AI researcher quits Anthropic with warning about safety
1 min read
An artificial intelligence researcher who worked at Anthropic — and previously OpenAI — has resigned, saying both tech giants are "gambling with our lives." Jacob Coxon shared his exit in a post on X, warning that the world should "not underestimate the power of this technology." He added: "These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire...
How GPT‑5.6 Sol helps run quantum computing experiments
4 min read
How GPT‑5.6 Sol helps run quantum computing experiments Connecting GPT‑5.6 Sol to laboratory software to run and refine routine measurements on quantum chips freed Beatriz Yankelevich to focus on experiment design and data analysis. Quantum computing is an emerging technology that uses the unique properties of quantum mechanics to process information. It could one day better simulate complex mater...
On Really Trying (2009)
74 min read
On Really Trying What are the true limits to motivation? The Programmers’ Credo: we do these things not because they are easy, but because we thought they were going to be easy Student: How can one realize his Self-nature? I know so little about the subject. Yasutani: First of all, you must be convinced you can do so. The conviction creates determination, and the determination zeal. But if you lac...
Copyright does more harm than good and should be abolished
1 min read
To use the Mastodon web application, please enable JavaScript. Alternatively, try one of the native apps for Mastodon for your platform.
The Navier–Stokes Millennium Prize Problem
5 min read
8th September 2026 - Link Blog On the Navier–Stokes Millennium Prize Problem (via) Impressive result from OpenAI, who used an unreleased model to produce a resolution to the Navier–Stokes existence and smoothness problem, one of the seven Millennium Prize Problems that have been subject to a $1,000,000 prize since May 24th, 2000. The discovery is somewhat overshadowed by accusations of skulduggery...
AI Has a Discovery Problem
2 min read
2026-09-03 The Discovery Problem The biggest bottleneck to AI adoption is a simple question: what can this do for me? The hard part is that you don’t know what you don’t know. You don’t know what a button does until you press it. You don’t know what a prompt can produce until you write it, hit go, and watch it run. As long as capabilities stay locked behind a blank text box, the possibilities stay...
AI Responsibility – OpenAI and Anthropic
1 min read
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
Anthropic researcher quits over AI fears
1 min read
An Anthropic researcher is quitting the artificial-intelligence industry over fears that the lab and its competitors are racing to build systems they won’t be able to control, a sign of mounting safety concerns within top AI companies. Jacob Coxon, a researcher who specializes in training new AI models by having them consume vast amounts of data, said Tuesday that he is leaving the company because...
I resigned from Anthropic today
1 min read
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
Tao: Open math problems being non-renewably mined by AI
1 min read
To use the Mastodon web application, please enable JavaScript. Alternatively, try one of the native apps for Mastodon for your platform.
Mercury 2.5
4 min read
Today, weâre releasing Mercury 2.5, our most capable production model yet. It is a significant step up in quality over Mercury 2, with the same low-latency, low-cost serving profile. Since Mercury 2âs launch, thousands of developers have built with it, dozens of enterprises have put it into production, and usage has grown over an order of magnitude. It now serves latency-sensitive workloads across...
Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs
10 min read
TL;DR — Kimi K3 (2.8T-parameter MoE, 1.45 TB of expert weights) running on one M5 Max MacBook Pro with 128 GB, experts streamed from four SSDs. - 1.00 tok/s steady decode over a 512-token answer; 1.13 over 128; 0.96 on the public 17-token prompt (upstream reported 0.68). - The honest limit: a 512-token prompt takes ~6.3 minutes to its first token. Cause found (prefill re-reads each layer's experts...
OpenAI fought dirty on career-making math problem
4 min read
NYU mathematics professor Tristan Buckmaster announced three proofs on Tuesday with a preliminary finding on one of the major unsolved problems in theoretical mathematics. The findings, made in collaboration with Anthropic mathematician Levent Alpöge and using both Codex and Claude AI models, are significant in themselves — but they’re also accompanied by an unusual controversy surrounding OpenAI’...
Muse: Meta's personal AI agent, features and capabilities
1 min read
Meet Muse, Meta's personal AI agent. Learn what it can do across everyday tasks, how it works, and how it helps you get more done.
Real photos of young girls were in nudify-app ads on Facebook, Instagram
1 min read
Meta took days to remove ads containing AI-generated child sexual abuse material (CSAM) on Facebook and Instagram. Some ads featured photos of real kids, including a press photo of a young member of a European royal family and images swiped from a popular Instagram profile of a preteen girl deemed an influencer. In an investigation published Tuesday, the Tech Transparency Project (TTP) reported th...
ChatGPT Images 2.5
6 min read
Introducing ChatGPT Images 2.5 Sharper details, faster generation, more precise editing, and better tools for creating and sharing. Every week, people create more than 3 billion images across ChatGPT Images and the GPT‑Image models in the API. Today, we’re expanding what you can create with ChatGPT Images 2.5—our new state-of-the-art image model that brings sharper details, more precise editing, a...
GrapheneOS on AI Usage
1 min read
To use the Mastodon web application, please enable JavaScript. Alternatively, try one of the native apps for Mastodon for your platform.
On the Navier–Stokes Millennium Prize Problem
7 min read
On the Navier–Stokes Millennium Prize Problem We’re sharing a solution to the Navier–Stokes existence and smoothness problem, one of the Millennium Prize Problems. This proof, produced by an internal OpenAI system, shows that the dynamics of the Navier-Stokes equations for fluid motion can develop a singularity in finite time. We’re sharing both a writeup of the proof and a formalization in Lean. ...
OUI-1: world's first model for Generative UI
8 min read
OUI-1 is a finetuned DiffusionGemma model that writes user interfaces in openui-lang. It is 26BA4B model that can run on consumer grade GPU (RTX 5090, at FP8), and the weights are on Hugging Face under the Gemma Terms of Use. Why we built OUI-1 Agent-driven interfaces are the future of software. But getting there comes down to three constraints. Interfaces must be generated in under a second. They...
Y Combinator Early Access Network
1 min read
The Early Access Network is an invite-only group of senior technology leaders who want to see what's coming before it's on anyone's shortlist. Four times a year, we put a hand-picked group of YC's...
Do you think it happened? Research stolen from their Codex private chats
1 min read
Posted by KingOfMyRoom. Score: 60 points. Comments: 21.
US police fear Meta smart glasses could be used to secretly record them
6 min read
In January, the New York police department’s counter-terrorism unit sent out a memo warning police officers about a new potential “security and counterintelligence” threat: Ray-Ban Meta glasses. The memo directed officers to “conduct thorough inspections of all eyewear permitted within inmate cells” and cautioned that people who have been arrested could use these glasses, equipped with small camer...
Disappointed Optimists
12 min read
Disappointed Optimists Every time I give a talk on design, someone finds me afterward. They wait out the handshakes and the business cards, and then they ask a version of the same question. How do I make THEM do things the RIGHT WAY? (this happens with other disciplines too, but it ALWAYS happens when I talk to designers.) I never have a good answer. Sometimes I talk about persuasion, sometimes so...
Google DeepMind Releases AlphaGenome Atlas
2 min read
AlphaGenome Atlas: a high-resolution map of human DNA The human genome is made of about 3 billion base pairs of DNA — but much of it remains a mystery. Scientists understand the 2% of the human genome that codes for proteins relatively well, but have only limited knowledge of the remaining 98%. Our AlphaGenome model has already shown how single changes in these non-coding DNA regions can disrupt m...
Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses
6 min read
How much GPU RAM do you actually need to run Qwen3.8 27B without sacrificing quality? The full BF16 model weighs 55 GB, putting it beyond most consumer hardware. Yet the 17 GB Q4_K_M matches the full model on a popular agentic coding benchmark, Terminal-Bench 2.1. It fits on a 24 GB card such as RTX 4090, still leaving room for about 64k tokens of context. Compression eventually hits a cliff. At 1...
I-have-ADHD: A skill to stop coding agents from burying the answer
1 min read
ADHD-friendly outputs. No ADHD diagnosis needed! 🇬🇧 · 🇨🇳 · 🇧🇷 · 🇯🇵 · 🇻🇳 · 🇰🇷 · 🇹🇭 Copy/paste into your CLI prompt: Install the i-have-adhd skill/plugin from https://github.com/ayghri/i-have-adhd, refer to the repo's AGENTS.md for instructions. Or 🔗 check the installation instructions. A skill for your coding assistant that stops it from burying the answer. Action first. Steps numbered. No "Hope th...
LibreOffice breaks download records after declaring it has no AI features
2 min read
LibreOffice breaks download records after declaring it has no AI features LibreOffice 26.8, released on August 26th, became the software’s most popular update. LibreOffice is a free alternative to Microsoft Office that’s been around for nearly two decades. In one week, the installer was downloaded more than 1 million times — not counting updates via Linux distribution repositories. Someone might s...
We Must Return to the Office to Use AI in Person
6 min read
When I awoke on RTO Day (officially “Remain to Office,” since management maintained that I’d misremembered and we’d always been in office six days a week), I was not as excited as I should have been. I felt queasy at the thought of signing the Memorandum of Loving RTO as a condition of my continued employment. I was an Associate Slop Doula for Mondo Mayo, founded as a mayonnaise company, which now...
PISA 2025 Students' reading and mathematics performance declined across the OECD
1 min read
Posted by mazokum. Score: 109 points. Comments: 117.
Among European Companies That Use a CDN, Nearly 9 in 10 Use Cloudflare
7 min read
A content delivery network sits in front of a website. Requests hit the CDN's edge servers first, which serve cached content, terminate TLS, filter traffic, and pass the rest back to the origin. That means the CDN is a shared dependency: if it goes down, every site behind it goes down together, whether or not those sites have anything else in common. I went through CipherCue's tracked European com...
AI Burnout Hits the People Charged With Defending Hospitals and Banks From Hackers
1 min read
Technology AI Burnout Hits the People Charged With Defending Hospitals and Banks From Hackers “The humans are going to break before the AI does.” The thing Val Mairet always loved about working in cybersecurity was the clarity of purpose. “You know what you’re working for: You have to make sure the company doesn’t explode,” she says. Over the course of almost a decade, Mairet worked as an ethical ...
Navier-Stokes – Tristan Buckmaster [pdf]
1 min read
Posted by procedurecall. Score: 315 points. Comments: 175.
Multi-Agents LLM Financial Trading Framework
9 min read
- [2026-08] TradingAgents v0.4.0 released with look-ahead / point-in-time fixes across FRED macro, social sentiment, and the decision-log memory; clearer decision signals; working CLI checkpoint resume; Trader price grounding; and the GPT-5.6 and GLM-5.3 models. See CHANGELOG.md for the full list. - [2026-07] TradingAgents v0.3.1 released with correctness and stability fixes: Alpha Vantage look-ah...
Mistral raises €3B
3 min read
Company Mistral raises €3B to make sovereign, open-weight AI the technology frontier Mistral today announced that it has raised €3 billion in a Series D funding round at a post-money valuation of more than €21 billion, the largest equity fundraising round ever completed by a European technology company, three years after the company's launch. Samsung Electronics led the round, joined by co-leads S...
We have a year to fix security everywhere
14 min read
we have a year to fix security everywhere GLM 5.3-flash released last week, and that means Project Glasswing and Daybreak are running out of time. Cheap models capable of dangerous hacking are now available to anyone, without the normal safeguards for refusing malicious actions. We need to fix vulnerabilities across the industry so that we aren't caught unawares. And for one of the first times in ...
The VMs Powering Mobile Agents (Instinct, Claude Code)
11 min read
One awesome product evolution is that agents (Claude Code, Instinct, Poke, etc) are moving off our local computers so that we can use them on our phones. Ultimately this is great for the customer because that means the agent companies provide us with VMs for them to run on! Here's some notes on how the major platforms work based on looking around on ws-term. Claude Code on Your Phone Claude Code's...
Arm Mali G2-Ultra NX GPU: desktop-class mobile gameplay with AI-native graphics
6 min read
Inside the Arm Mali G2-Ultra NX GPU: Delivering desktop-class mobile gameplay with AI-native graphics Demands from mobile users continue to grow. They want richer visuals, smoother frame rates, and increasingly immersive experiences. For developers the challenge is complex, as they aim to meet these expectations while staying within the strict power, thermal, and bandwidth limits of mobile devices...
I tested 10 model/harness combinations on the same Three.js task
3 min read
Hangar Harness / Model Tests I've been testing a simple prompt with different model and harness combinations to work out which one produces best results. I do this in /goal mode. Prompt: Build a single-page Three.js sci-fi hangar with hovering drones, animated warning lights, emissive runway strips, and subtle volumetric-style fog planes. Include drone formation toggle and cinematic camera path. O...
How well do agents use test/verification techniques?
64 min read
We previously noted that, while it's easier than ever to hit a particular quality bar by having coding agents use effective test techniques, software quality seems to be getting worse, indicating that whatever defaults developers are using may not work very well. Here, we test if simple instructions to agents to use particular techniques or libraries improve implementation correctness, as a kind o...