I’ve heard some buzz around the new glm 5.2 open-weights model. They say it’s very capable! I won’t run a full comparison benchmark, but I have some credits sloshing around on OpenRouter so I figured I might compare glm 5.2 to the similarly-priced Gemini 3 Flash11 The market currently infers with the glm 5.2 model at $4.4 per million output tokens, whereas Google charges $3 per million output tokens for their model. I expect the price of the glm model to go down somewhat when people figure ou... Read more ›
Sword juggling is simple. I’ll prove it to you by explaining it. First, acquire three – or, if you feel ambitious, four swords. Divide the swords between your hands. Starting with one hand, the one holding two swords if juggling with three, toss a sword into the air. As this sword is about to reach your other hand, toss the sword in that hand and catch the falling sword by, and this is important, its handle, not the blade. Then repeat this process as the airborne sword reaches the hand you st... Read more ›
I’ve written a couple posts lately on getting an LLM to generate code to solve chess problems. The first used Claude to generate Prolog and the second used ChatGPT to generate Prolog. This post will use Claude to generate Z3/Python code. The puzzle is one I’ve written about before: Place all the pieces—king, queen, two […] The post first appeared on . Read more ›
AI made you faster. But you’re not more productive; you’re outsourcing the slow part to everyone else Read more ›
Today we launched a new plugin for Datasette, on the Datasette project blog. That post has the what, but I'm going to expand on that a little bit here to provide the why. The TL;DR Datasette Apps are self-contained HTML+JavaScript applications that run in a tightly constrained sandbox hosted on your Datasette application. They can use JavaScript to run read-only SQL queries against data in Datasette, and can run write queries too if you configure them . Here's a and a - the latter lo... Read more ›
There are fewer ways to leave your package than to kill it. Read more ›
So it’s been a big week for me after I published an exclusive covering OpenAI’s audited financials from 2024 and 2025, with reactions ranging from “oh my god, OpenAI spent $34 billion to make $13.07 billion in revenue!” to “actually, it’s good the company lost $21 Read more ›
How KV cache compression - from MQA and GQA to MLA and linear-attention hybrids - quietly unlocked the long context windows that make modern agentic LLMs possible. Read more ›
Remember that scene in the Matrix where Neo is strapped into the chair and Link uploads all sorts of martial arts into his mind? When Neo wakes up, he says "I know Kung-fu," then proceeds to demonstra Read more ›
I’ve been working since they came out, and finally, they’re surprisingly good now. I have a 2022 M2 Mac with 64 GB RAM and 1TB storage and I’ve used , as well as a number of other Qwen variants like across like raw llama.cpp with llama-cpp-python Ollama llamafiles and LM Studio Where are local models now? Early on, models were slow, hard to use, and just not that accurate for most programming tasks. The idea that local models were severely lagging behind was largely true until, for me, the re... Read more ›
We’re on a journey to advance and democratize artificial intelligence through open source and open science. Read more ›
Release: This release expands datasette-acl from table-only permissions toward a general resource-sharing system. Alex Garcia did most of the work for this release - we're fleshing out the plugin that will allow multi-user Datasette instances finely grained control over who can access which resources within Datasette. Tags: <a href=" <a href=" Read more ›
is a place to collect interesting things I’ve seen, read, or heard, along with some brief thoughts (often incomplete and/or inconclusive) that they provoked. – socket.dev Back in February this report landed in a security blog, and I found it fascinating. As is usual for me, it’s at the intersection of human behaviour with technological automation that the real interesting problems lie. The whole industry uses GitHub profiles as a proxy for a CV. Now agents can build a credible looking one in ... Read more ›
As James Carville might have said, “It’s the lack of a moat, stupid” Read more ›
"We see these AIs as a galaxy glittering with capabilities, but at their center, invisible to the naked eye, holding all the constellations together, is an unimaginably massive black hole of data." Read more ›
A few days ago I wrote about using Claude to solve a chess puzzle by writing Prolog code. This morning I tried a similar chess puzzle with ChatGPT. The task is to place a queen, king, rook, bishop, and knight on a 4 by 4 chessboard so no piece attacks another. Of course there’s not […] The post first appeared on . Read more ›
The real valuable capability MCP offers over skills/CLI is isolating the auth flow outside of the agent’s context window, and potentially out of the harness completely. [...] Maybe the idealized … Read more ›
The Washington Post editorial board yesterday \( News\+ link \), “Why Europe Won’t Have the New Siri: Brussels insists the decision is “ Apple’s and Apple’s only ” and that nothing in its flagship Digital Markets Act forbids the launch\. That’s technically true and wholly beside the point\. The law requires that the moment Siri AI ships in Europe, any rival AI agent must get the same sweeping access to a user’s messages, files and chat history\. Apple proposed putting in a software security l... Read more ›
If you liked this piece, you should subscribe to my premium newsletter. It’s $70 a year, or $7 a month, and in return you get a weekly newsletter that’s usually anywhere from 5,000 to 18,000 words, including vast, detailed analyses of NVIDIA, Anthropic and OpenAI’s Read more ›
A blog about making culture. Since 1999. Read more ›