Scanning
How to Think When the Tools Can Do Anything
The tools got free. The thinking didn't.
Aluminum was once worth more than silver. In 1884 the United States capped the Washington Monument with a hundred-ounce aluminum pyramid because it was the most precious metal the nation could put at its highest point. Two years later the Hall-Heroult process arrived, and within a decade the price had fallen by more than ninety-nine percent. The metal did not change. What changed was that making it stopped being hard.
The same collapse has happened to the cost of building software. Writing code — the skill that defined a profession for fifty years — is no longer the scarce part. What remains scarce is knowing which problem is worth solving, recognizing when an answer is wrong, framing a question precisely enough to be answerable, and refusing to stop before the thing actually works. This book is about those four capacities: curiosity, analysis, creativity, and persistence.
It is not a book about prompts, models, or tools — none are named anywhere in it. It is a book about the habits of mind that make someone dangerous with whatever tool they happen to be holding, drawn from thirty years of shipping things and from the uncomfortable discovery that the mindset which once produced code now has to produce judgment instead.
What happens to a skill when the cost of exercising it falls to near zero, told through aluminum, artificial light, and the forty years of software work that just repriced.
Why capable tools widen the gap between practitioners rather than closing it, and what gets amplified when you point one at unclear thinking.
Curiosity as a working method rather than a personality trait: how to generate the question nobody in the room has asked yet, and why most people stop one question too early.
Reading a system until you can predict it. How to tell a real explanation from a plausible one, and how to find the load-bearing assumption inside your own reasoning.
Constraint as the generator, not the obstacle. Why the useful kind of creativity is structural recombination rather than inspiration.
Persistence as a scheduling problem rather than a character virtue — and the stopping rule that separates it from stubbornness.
The four capacities as one cycle: how to shorten the distance between a question and evidence, and why loop speed beats loop quality.
The difference between what a model can know and what only you can know. Knowing what a problem is worth to the business is permanently yours; knowing the reality a solution must survive is yours for now.
Turning all of it into a standing practice, with the finishing move: knowing when the thing is done.
Twelve usable assessments and exercises across the four capacities, built to be run on real work rather than scored on a scale.
Tim de Vallée
Tim de Vallée built the US operation for IMDb in the mid-1990s — from a hard disk shipped to a volunteer in Wisconsin to the first internet ad campaign a Hollywood studio ever bought. He now builds intelligence tools at INTLO Labs, the product house of Digital Boutique AI, drawing on thirty years of shipping things.
More about the author →