Studies
Things I wanted to understand properly. Mostly AI, decision-making, and how things actually work. Every claim cited.
What AI Has Done, What's Wrong With It, and What's Just Noise
Sorting the real problems from the ones that get shouted about
August 22, 2026
AI Agent Architectures: Workflows, Subagents, and Multi-Agent Systems
What the words mean, and what the evidence says about using more agents
August 22, 2026
Grok Bot vs Claude Cowork vs ChatGPT
Five agent products and where each one runs
August 21, 2026
How Large Language Models Actually Work
The whole mechanism, from tokens to hallucinations
August 14, 2026
The Mental Models I Keep Coming Back To
The handful that changed how I make decisions
July 7, 2026
A History of AI, Told Through One Number
How the weight went from a 1943 idea to hundreds of billions of them
July 1, 2026
Why AI Gives You a Different Answer Every Time
It comes down to one number called temperature
June 6, 2026
How to Tell When AI Is Making Things Up
The failure modes, the benchmark problem, and what actually catches it
May 3, 2026
Seven Things I Learned from Building
Mostly by doing the wrong version first
February 26, 2026