Reputation interrogation

The reputation interrogation is a fixed set of questions asked of every major model on the same schedule, logged verbatim, so that change in your AI resume is measured rather than guessed at.

The book's own term for the monthly discipline it closes on.

Nearest equivalents are AI visibility monitoring and share of voice tracking, usually sold as software.

The audit is the first read. The interrogation is the same read, repeated, so movement becomes visible.

It is read for movement rather than mood: what changed since last month, not whether this month's answer feels good.

The book puts it at an hour a month and calls it the steering wheel rather than a report card, because findings feed straight back into the correction, evidence and response work. Its closing warning is that most readers will skip it, and will later call the consequences bad luck.

Read about the book's argument.