Tyler Cowen wrote 42,000 words on humans losing their edge to machines. Page one invites you to read it with AI. Page one invites you to explore the text with AI. He calls it a generative book. It is free. Read it. I disagree with some of it, which is usually a sign that something is worth your time. His case: the move that built modern economics, change one variable and hold the rest constant, is losing ground. Where data runs deep, machines can see patterns no economist could find alone. Exhibit A: a finance model with 360,000 factors. It cut pricing errors 54.8% against the Fama-French benchmark, and it holds up out of sample. Fermi, 1953, to a young Freeman Dyson: with four parameters I can fit an elephant, and with five I can make him wiggle his trunk. Dyson went home and killed a program he had run for years. Four was already too many. This one has 360,000, it works, and it cannot tell you why. At some point, you are not discovering the animal. You are building one that walks. You have heard that water is cheap, diamonds are dear, and nobody could explain it for two thousand years. You are already ahead of me: scarcity. Water is everywhere; diamonds are rare. So was everybody before 1871 an idiot? Not by itself. They had the answer and kept setting it down. Galileo wrote it out in 1632 and spent it complaining that people are fools for prizing gold over dirt. Smith saw it in 1776, called it a paradox, and moved on because it did not fit the theory he was building. And scarcity is not quite the answer anyway. Scarcity is a fact about the world's supply. The marginal idea is a fact about your cupboard. There is the same water on Earth whether you are at your kitchen sink or eleven miles into the Mojave. What moved is not supply. It is how many glasses you already have. Scarcity cannot tell those two men apart. Your cupboard can. Finding the answer was not what took two centuries. Knowing it mattered was. Because the model eats data, and somebody still has to go make it. Machines win where the thing you measure holds still. Judgment earns its keep where the goal is arguable, the ground shifts, your own move changes the board, and being wrong costs you something you don't get back. The decisions that cost me the most money were all the second kind. So were the ones that saved me. A model can make the same mistake ten thousand times before lunch and come out smarter, because in the search every wrong answer is just data. Mine happen in the world, with money in them. So the question is not how many shots I get. It is how many I can take and still be standing.