Understanding AI

What I learned trying seven coding agents

Computer-use agents seem like a dead end

The first copyright ruling on generative AI training is a win for AI labs

How AI agents got good at using tools

Reinforcement learning, explained with a minimum of math and jargon

Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book

Claude-powered coding tools are poised to transform programming

I spent time testing Claude 4, Gemini 2.5, and o3

I got fooled by AI-for-science hype—here's what it taught me

Three great articles on the modest pace of AI progress

There may only be two people who can stop OpenAI's for-profit pivot

Where frontier language models are today