Judgment Times AI.

Every superhero story has the same structure underneath the costume. An ordinary person receives a power. The power is not the story. The story is what the person does with it, and whether their character is strong enough to carry it. That is the most honest picture I know of what AI does for a leader.

I think of it as simple arithmetic. Judgment times AI. If your judgment is strong, the product is large. If your judgment is zero, the product is zero no matter how powerful the tool. If fear or ego is steering, the tool multiplies the damage.

Three values of judgment

Strong judgment

A large product. Experience applied at a tempo that used to be impossible.

Zero judgment

Zero. The tool cannot supply what is missing. It produces impressive output and poor decisions.

Negative judgment

Bias, fear, ego and a stale picture of the market, multiplied at scale.

The frontier is jagged

The research now shows the negative case in the field. Fabrizio Dell’Acqua, Ethan Mollick and their colleagues ran an experiment with consultants at Boston Consulting Group. On tasks inside what they call AI’s jagged frontier, the consultants with AI did better and faster. On a task chosen to sit just outside it, they were 19 percent less likely to reach a correct answer than consultants working alone. The tool did not know where its edge was. Only judgment could tell.

I learned the same thing with neural networks at WiseWire in the 1990s, and later with genetic algorithms at Affinnova. The machine could search a space of possibilities far larger than any human team. It could not decide what the search was for. In every system I built, the value came from the moment a person with judgment looked at what the machine produced and decided what to do with it.

A multiplier does not care what it multiplies.

That is the good news. Every inner practice, the debrief, the stillness, the gratitude, the clear intent, is now worth more than it was before. You have been sharpening the multiplicand all along.

Raise the multiplicand

1. Find the edgeBefore you trust an output, test the tool on problems where you already know the answer. Learn where it is strong and where it bluffs.
2. Check the signBefore a major AI-assisted decision, write down what fear, ego or an old picture of the market might be steering you. Those get multiplied too.
3. Keep the decision with a personLet the machine gather, draft and argue both sides. A named person decides and owns the result.
4. Debrief the judgmentWhen the decision plays out, review the judgment that went in, not only the quality of the output.