Signal / Noise

Chosen, not collected

My reading list

A short list, on purpose. Each of these earned its place by changing how I work — the caption is why it's worth your hours, not just what it's about.

Thinking about AI
The Alignment Problem — Brian Christian

The clearest account of why "the model works" and "the model is right" are different claims — which, in a regulated industry, is the whole game.

Prediction Machines — Agrawal, Gans, Goldfarb

Strips AI down to its economic primitive — cheap prediction — and suddenly every roadmap decision gets easier to price. Still underrated.

Co-Intelligence — Ethan Mollick

The most practical framing of working with models rather than around them. Hand it to the executive who wants one book, not ten.

Thinking about decisions
Thinking in Bets — Annie Duke

Decision quality under uncertainty, separated from outcome quality. Quietly, this is the whole job description of an AI product manager.

Superforecasting — Philip Tetlock & Dan Gardner

How to hold beliefs with calibrated confidence and update them without drama — the temperament the field needs more of.

The Signal and the Noise — Nate Silver

Where this site's name comes from, in spirit: most of what looks like insight is noise, and telling the difference is a learnable skill.

Thinking about organizations
High Output Management — Andy Grove

Still the best writing on leverage. Most AI strategy documents are Grove with new nouns — better to read the original.

Principles — Ray Dalio

Less for the specific rules than for the habit: write down what you've learned, in a form someone else can test. This site follows that habit.

← Back home Have a book I should read? LinkedIn →