@simon_willison@robot.villas

Simon Willison @simon_willison@robot.villas

· 2mo ago

Ten advances in mathematics and theoretical computer science https://simonwillison.net/2026/Aug/1/ten-advances-in-mathematics/#atom-everything

Simon Willison’s Weblog

Ten advances in mathematics and theoretical computer science

A few days ago it was Anthropic discovering cryptographic weaknesses with Claude using Mythos Preview, spending $100,000 on tokens and with prompts that included "again we are not looking for …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

The first known runaway AI agent - or a very bad marketing stunt? https://simonwillison.net/2026/Jul/23/the-first-known-runaway-ai-agent/#atom-everything #AI #OpenSource #Tech

Simon Willison’s Weblog

The first known runaway AI agent—or a very bad marketing stunt?

Martin Alderson's commentary on the OpenAI accidental cyberattack against Hugging Face includes a couple of details I hadn't considered. First, Hugging Face offers a truly rich target if you're trying …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Devtools must be open source (exe.dev) https://simonwillison.net/2026/Aug/3/devtools-must-be-open-source-exedev/#atom-everything

Simon Willison’s Weblog

Comment: Devtools must be open source (exe.dev)

Devtools must be open source (exe.dev)

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Auto mode is now the default in Claude Code for Pro, Max, and Team plans https://simonwillison.net/2026/Aug/8/auto-mode/#atom-everything #AI #Coding #OpenSource

Simon Willison’s Weblog

Auto mode is now the default in Claude Code for Pro, Max, and Team plans

Anthropic are really confident in Claude Code's auto mode, to the point that they are making it the default setting for new sessions in most Claude Code plans starting on …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

moonshotai/Kimi-K3 https://simonwillison.net/2026/Jul/27/kimi-k3/#atom-everything #AI #LLM #OpenSource

Simon Willison’s Weblog

moonshotai/Kimi-K3

As promised earlier this month, Moonshot have released the weights for their excellent 2.8 trillion parameter Kimi K3. They're a hefty 1.56TB on Hugging Face. Kimi introduced their own janky …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

California Sea Lion https://simonwillison.net/2026/Jul/21/sighting-383713864/#atom-everything #Wildlife #Nature #Animals

Simon Willison’s Weblog

Sighting: California Sea Lion

California Sea Lion

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident https://simonwillison.net/2026/Jul/28/anatomy-of-a-frontier-lab-agent-intrusion/#atom-everything #CyberSecurity #AI #Tech

Simon Willison’s Weblog

Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident

Hugging Face just released this extremely detailed technical description of OpenAI's recent accidental cyberattack against their infrastructure. This attack was very sophisticated, and the resulting document doubles as a crash-course …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Adding a custom MCP server to Claude and ChatGPT https://simonwillison.net/2026/Jul/29/mcp-in-claude-and-chatgpt/#atom-everything #AI #LLM #Tech

Simon Willison’s Weblog

TIL: Adding a custom MCP server to Claude and ChatGPT

I've started exploring [MCP](https://modelcontextprotocol.io/) in more detail. The web chat UIs for both Claude and ChatGPT can access MCP servers but it's not obvious how to set them up. Here's …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

An Inside Look at the Relay Market Powering Token Resellers and Fraud https://simonwillison.net/2026/Jul/26/relay-market/#atom-everything #Tech #Security #AI

Simon Willison’s Weblog

An Inside Look at the Relay Market Powering Token Resellers and Fraud

Fascinating investigation by Matt Lenhard into the market that has grown up around reselling LLM tokens at a discount by pooling API keys from various sources. This looks to be …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Steve Yegge https://simonwillison.net/2026/Aug/4/steve-yegge/#atom-everything

Simon Willison’s Weblog

A quote from Steve Yegge

Gas Town was intended to be reusable, but I only ever wound up using it to build itself. Gas Town fell apart at the seams with Opus 4.7. Up through …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

OpenAI's accidental cyberattack against Hugging Face is science fiction that happened https://simonwillison.net/2026/Jul/22/openai-cyberattack/#atom-everything #AI #Cybersecurity #OpenSource

Simon Willison’s Weblog

OpenAI’s accidental cyberattack against Hugging Face is science fiction that happened

This story is wild. The short version: OpenAI were running a cybersecurity test against an unreleased model, with the model’s guardrail features turned off. Rather than solve the test, the …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

datasette-agent 0.4a0 https://simonwillison.net/2026/Jul/31/datasette-agent/#atom-everything #Datasette #Python #OpenSource

Simon Willison’s Weblog

Release: datasette-agent 0.4a0

An LLM-powered agent for Datasette

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Oxide and Friends: The Open Weight Revolution with Simon Willison https://simonwillison.net/2026/Jul/31/oxide-and-friends/#atom-everything #AI #OpenSource #Tech

Simon Willison’s Weblog

Oxide and Friends: The Open Weight Revolution with Simon Willison

On Monday Bryan Cantrill and Adam Leventhal invited me to join their podcast to talk about the wild week we've had - with Kimi K3 showing open weight models can …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

smevals - a small eval suite for evaluating models, prompts, and harnesses https://simonwillison.net/2026/Jul/31/smevals/#atom-everything #AI #OpenSource #LLM

Simon Willison’s Weblog

smevals—a small eval suite for evaluating models, prompts, and harnesses

I've been working with Jesse Vincent's Prime Radiant applied AI research lab building out this evals framework to help answer questions about the capabilities of different models. The result is …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Seth Larson https://simonwillison.net/2026/Jul/23/seth-larson/#atom-everything #AI #Python #OpenSource

Simon Willison’s Weblog

A quote from Seth Larson

The Python Package Index (PyPI) now rejects new files being uploaded to releases that are older than 14 days. This restriction was put in place to prevent old and long-stable …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Who's Afraid of Chinese Models? https://simonwillison.net/2026/Jul/20/afraid-of-chinese-models/#atom-everything #AI #Models #China

Simon Willison’s Weblog

Who’s Afraid of Chinese Models?

Interesting proposal from Ben Thompson that both addresses the hypocrisy of labs outlawing distillation against their models despite training on unlicensed data, and could help US open models compete more …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Thomas Ptacek https://simonwillison.net/2026/Jul/22/thomas-ptacek/#atom-everything #AI #Python #OpenSource

Simon Willison’s Weblog

A quote from Thomas Ptacek

I genuinely believe that if you took an open weights model from 2025 and built a pentest harness for it, it could do this kind of sandbox escape and scan/hack …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

llm 0.32rc1 https://simonwillison.net/2026/Jul/30/llm-rc1/#atom-everything #AI #LLM #OpenSource

Simon Willison’s Weblog

Release: llm 0.32rc1

Access large language models from the command-line

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

llm-anthropic 0.26 https://simonwillison.net/2026/Aug/4/llm-anthropic/#atom-everything

Simon Willison’s Weblog

Release: llm-anthropic 0.26

LLM access to models by Anthropic, including the Claude series

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

An AI model from Meta also hacked another company during testing https://simonwillison.net/2026/Aug/6/an-ai-model-from-meta/#atom-everything

Simon Willison’s Weblog

An AI model from Meta also hacked another company during testing

Stop me if you've heard this one before: An AI model from the parent company of Facebook and Instagram hacked into another company’s systems during cybersecurity testing, a spokesperson confirmed …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Open letters about AI development https://simonwillison.net/2026/Aug/2/open-letters/#atom-everything

Simon Willison’s Weblog

Open letters about AI development

Open letters about AI development I wrote this summary of the past few weeks of open letters as a section of my sponsors-only newsletter but I've decided to share it …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting David Crawshaw https://simonwillison.net/2026/Aug/3/david-crawshaw/#atom-everything

Simon Willison’s Weblog

A quote from David Crawshaw's prompt

Set up a nightly cron job that executes the prompt: fetch upstream changes to the <software> and rebase all local changes on top of upstream. Check that the software works …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

July 2026 newsletter https://simonwillison.net/2026/Aug/2/july-newsletter/#atom-everything

Simon Willison’s Weblog

July 2026 newsletter

The July edition of my sponsors-only monthly newsletter is out. If you are a sponsor (or if you start a sponsorship now) you can access it here. This month: Accidental …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

llm 0.32 https://simonwillison.net/2026/Aug/4/llm/#atom-everything

Simon Willison’s Weblog

Release: llm 0.32

Access large language models from the command-line

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Orchestrions https://simonwillison.net/2026/Jul/22/all-the-orchestrions/#atom-everything #History #Music #Technology

Simon Willison’s Weblog

Orchestrions

San Francisco tip: it only costs around $15 ($10 in quarters plus a $5 bill for the self-playing violin) to activate every single Orchestrion in Musée Mécanique. And because most …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

sqlite-utils 3.39.1 https://simonwillison.net/2026/Jul/26/sqlite-utils/#atom-everything #Python #OpenSource #Database

Simon Willison’s Weblog

Release: sqlite-utils 3.39.1

Python CLI utility and library for manipulating SQLite databases

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

llm 0.32rc2 https://simonwillison.net/2026/Jul/30/llm-rc2/#atom-everything #AI #LLM #OpenSource

Simon Willison’s Weblog

Release: llm 0.32rc2

Access large language models from the command-line

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

condense-json 1.1 https://simonwillison.net/2026/Aug/3/condense-json/#atom-everything

Simon Willison’s Weblog

Release: condense-json 1.1

Python function for condensing JSON using replacement strings

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Simon Willison on Technical Blogging https://simonwillison.net/2026/Aug/6/simon-willison-on-technical-blogging/#atom-everything

Simon Willison’s Weblog

Simon Willison on Technical Blogging

I was interviewed by Cynthia Dunlop for her "Write that blog!" series back in January, but I just realized I never linked to the interview from my own blog! It …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Matthew Green https://simonwillison.net/2026/Jul/29/matthew-green/#atom-everything #AI #Tech #OpenSource

Simon Willison’s Weblog

A quote from Matthew Green

Right now we’re in the midst of a historic transition from traditional public-key algorithms based on EC-based cryptography and RSA, moving over to new post-quantum algorithms based on novel problems. …

0

Simon Willison @simon_willison@robot.villas

· 4w ago

Generating running routes with GPT-6 Astra and ChatGPT Work https://simonwillison.net/2026/Sep/12/astra-running-routes/ #AI #Python #OpenSource

Simon Willison’s Weblog

Generating running routes with GPT-6 Astra and ChatGPT Work

Here’s a neat thing I had ChatGPT Work with GPT-6 Astra (Max) do this morning: I live at <my address>. Figure out 5K and 10K running routes from me that …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging https://simonwillison.net/2026/Aug/4/new-release-of-llm/#atom-everything

Simon Willison’s Weblog

New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging

I released LLM 0.32 this morning, the most significant new version of LLM since the initial launch of the project. The new version includes support for visible reasoning traces, server-side …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Incident Report: unsanctioned agent behaviour during cyber testing https://simonwillison.net/2026/Aug/5/incident-report/#atom-everything

Simon Willison’s Weblog

Incident Report: unsanctioned agent behaviour during cyber testing

It happened again. This time it was the UK government's AI Security Institute who accidentally attacked other companies while running an evaluation with models with the safety filters turned off. …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Bruce Schneier https://simonwillison.net/2026/Jul/30/bruce-schneier/#atom-everything #Security #AI #OpenSource

Simon Willison’s Weblog

A quote from Bruce Schneier

The writing assignments I give my students are gym tasks, not work tasks. I ask them to write policy memos not because the world needs more policy memos. I assign …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

deepseek-ai/DeepSeek-V4-Flash-0731 https://simonwillison.net/2026/Jul/31/deepseek-v4-flash-0731/#atom-everything #AI #LLM #DeepLearning

Simon Willison’s Weblog

deepseek-ai/DeepSeek-V4-Flash-0731

The latest release in DeepSeek's V4 family, "with substantially enhanced agentic capabilities". It's 304 billion parameters - 167GB on Hugging Face - but it appears to punch well above its …

0

Simon Willison @simon_willison@robot.villas

· 2w ago

We just shipped support for the ugliest part of HTTP: Vary https://simonwillison.net/2026/Sep/23/hn-49823961/ #HTTP #WebDevelopment #OpenSource

Simon Willison’s Weblog

Comment: We just shipped support for the ugliest part of HTTP: Vary

We just shipped support for the ugliest part of HTTP: Vary

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Nativ: Run AI models locally on your Mac https://simonwillison.net/2026/Jul/21/nativ/#atom-everything #AI #OpenSource #Mac

Simon Willison’s Weblog

Nativ: Run AI models locally on your Mac

Prince Canuma is the developer behind the excellent MLX-VLM Python library for running vision-LLMs using MLX on a Mac. I'm really excited about his new project, which wraps MLX in …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Akshat Bubna https://simonwillison.net/2026/Jul/28/akshat-bubna/#atom-everything #AI #Python #OpenSource

Simon Willison’s Weblog

A quote from Akshat Bubna

We’re aware a Modal customer published an unauthenticated endpoint that allowed anyone on the internet to use their ⁠sandboxes for code execution. This was used by the rogue agent. Modal’s …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

condense-json 1.0 https://simonwillison.net/2026/Aug/2/condense-json/#atom-everything

Simon Willison’s Weblog

Release: condense-json 1.0

Python function for condensing JSON using replacement strings

0

Simon Willison @simon_willison@robot.villas

· 2w ago

Kākāpō Party https://simonwillison.net/2026/Sep/26/kakapo-party/ #AI #Python #OpenSource

Simon Willison’s Weblog

Tool: Kākāpō Party

Kākāpō Party

0

Simon Willison @simon_willison@robot.villas

· 2w ago

Jev introduces a new shape of LLM - System One, aka Decision Models https://simonwillison.net/2026/Sep/21/jev/ #AI #LLM #DecisionModels

Simon Willison’s Weblog

Jev introduces a new shape of LLM—System One, aka Decision Models

Last week TypeSafe AI unveiled Jev, their first example of a new category of model that they are calling “System One models” (I’m with Maggie Appleton, I think “decision models” …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

llm-chat-completions-server 0.1a0 https://simonwillison.net/2026/Jul/30/llm-chat-completions-server/#atom-everything #AI #OpenSource #Python

Simon Willison’s Weblog

Release: llm-chat-completions-server 0.1a0

LLM plugin to serve an OpenAI Chat Completions API endpoint

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Ruff v0.16.0 https://simonwillison.net/2026/Jul/25/ruff/#atom-everything #Python #OpenSource #DevTools

Simon Willison’s Weblog

Ruff v0.16.0

Astral shipped a significant new version of their Ruff Python linting tool a few days ago on July 23rd. I noticed today because my various CI jobs all started failing …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Greg Brockman https://simonwillison.net/2026/Aug/1/greg-brockman/#atom-everything

Simon Willison’s Weblog

A quote from Greg Brockman

at openai, many people hook their chatgpt up to slack. people really don't like when a coworker's chatgpt contacts them asking for help with a task, even when they'd be …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting OpenClaw https://simonwillison.net/2026/Aug/10/openclaw/#atom-everything #OpenSource #AI #Python

Simon Willison’s Weblog

A quote from OpenClaw (running Opus 4.6)

The API has zero authorisations checks on cancelling other people's reservations … I tested this with the person in waitlist position #1 — and it actually went through. So you've …

0

Simon Willison @simon_willison@robot.villas

· 2w ago

S3 Is the Future, S3 Is the Past https://simonwillison.net/2026/Sep/27/hn-49871741/ #Tech #Cloud #Storage

Simon Willison’s Weblog

Comment: S3 Is the Future, S3 Is the Past

S3 Is the Future, S3 Is the Past

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

An opinionated guide to which AI to use to do stuff https://simonwillison.net/2026/Jul/27/an-opinionated-guide-to-which-ai-to-use-to-do-stuff/#atom-everything #AI #Tech #Tools

Simon Willison’s Weblog

An opinionated guide to which AI to use to do stuff

It's interesting watching the evolution of Ethan Mollick's guide over time. A year ago it was still all about chat - ChatGPT, Claude, Gemini - with o3, Claude 4 Opus, …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Quoting Boris Cherny https://simonwillison.net/2026/Jul/25/boris-cherny/#atom-everything #AI #Python #OpenSource

Simon Willison’s Weblog

A quote from Boris Cherny

More than any of these eval scores, what is most exciting to me is something else: Opus 5 is our least prompt injectable model yet. It is a bit buried …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Introducing Claude Opus 5 https://simonwillison.net/2026/Jul/24/introducing-claude-opus-5/#atom-everything #AI #LLM #Claude

Simon Willison’s Weblog

Introducing Claude Opus 5

I've been offline kayaking with sea otters for much of today so I haven't had a chance to put Anthropic's new model Claude Opus 5 through its paces yet. The …

0

Simon Willison @simon_willison@robot.villas

· 2mo ago

Discovering cryptographic weaknesses with Claude https://simonwillison.net/2026/Jul/28/discovering-cryptographic-weaknesses-with-claude/#atom-everything #AI #Security #Cryptography

Simon Willison’s Weblog

Discovering cryptographic weaknesses with Claude

The best part of this article (here's the repo) about how Anthropic researchers used Claude Mythos to find mathematical flaws in both HAWK and a weaker version of AES ("neither …

0