Prologue — Inside the Machine
You sit in front of six screens.
Price feeds scroll. Volatility surfaces ripple. Your system — the one you spent three years building — has just identified a mispricing in the options chain. It will be gone in ninety seconds. The algorithm fires. Delta-hedged. Risk-neutral. Clean.
You built this. You extracted signal from noise, compressed months of market behaviour into a covariance matrix, and designed a system that exploits the gap between what the market prices and what the mathematics says it should price. The beauty of a clean objective function. The elegance of constraint reduction. The satisfaction of measurable output.
Optimisation works.
Markets are computational miracles — millions of agents, each locally optimising, producing price discovery that no central planner could replicate. Abstraction is power. A derivative is a derivative of a derivative of a claim on future cash flows of a company that employs people who grow food from soil. Four layers of abstraction. Each layer compresses. Each layer accelerates. Each layer distances you from the thing itself.
And you benefit. Handsomely.
But somewhere between the third espresso and the closing bell, a thought:
The sharper the objective function, the more invisible everything outside it becomes.
When you build systems that optimise a single dimension, you feel their efficiency. You do not feel their blind spots. You do not feel the soil. You do not feel the fish. You do not feel the cognitive bandwidth being compressed out of the humans downstream of your system.
You are not a villain. You are an engineer. And this is the confession that earns the right to diagnose the machine from the inside.
No moral superiority. Just architectural clarity.