Claude Clue #38: Read the Manual Before the Model
Some people read the manual first. Some people start assembling and find out.
Christopher Moravec
Episode 81 - How on earth do AI watermarks work?
An AI watermark works by letting a secret function, not a random number generator, determine which word is selected from the words the model was already willing to say next.
Claude Clue #37: Let Another Claude Keep Score
Claude is very good at telling you it's done. Now another Claude gets a say.
Episode 80 - Two Types of Writing
There are two ways to write with AI: it drafts and you edit, or you draft and it edits. Each path leaves different “watermarks” on the result, so pick one on purpose.
Claude Clue #36: Give Your Claudes a Bulletin Board
Your Claude just renamed an interface. Your coworker's Claude has no idea.
Episode 79 - AI-Generated Metadata Philosophy
Metadata matters to search as much as it does to the AI Agent, so make it count!
Claude Clue #35: Put Fable in the Corner Office
Anthropic's biggest model is also the easiest one to waste.
Claude Clue #34: Give Your Orchestrator Two Jobs
Put a heavy-thinking model at the top and give it exactly two jobs: hand out assignments and verify that each one came back complete.
Episode 78 - Secret Cyborg: Special Disclosure Unit
On August 2, 2026, part of the EU AI Act became active, and now you legally can't be a Secret Cyborg (kinda).
Episode 77 - The Incredible Shark
Can a Shark (I mean AI) build a physics-based Rube Goldberg machine video game?