New
Don’t Trust “Done”
The sequel. The first book was about not trusting the model when it says it worked; this one is about not trusting it when it says it’s finished. A practical book about the gap between “it works” and it working — where most of what goes wrong, and most of the interesting engineering, actually lives.
Read more & free audiobook →
Don’t Trust the Green Light
The first field report: what running a fleet of twenty-five AI agents actually takes — not intelligence, but management, memory, judgement, and a stubborn refusal to believe the happy path. A green light is not proof. It is a hypothesis with good PR.
Read more & free audiobook →