AI in Strategic Decision Making: Extending Intuition

How should AI be used in strategic decision making?

As an extension of a leader’s intuition, not a replacement for it, because elite strategic intuition is far more reliable, and far more specific, than it looks. When a seasoned decision-maker “just knows” a deal is wrong, that is not mysticism. It is recognition-primed decision making: an expert rapidly matching the situation to patterns built from thousands of prior cases, then mentally simulating the move. In graph terms, it is a traversal across a densely weighted network, run in milliseconds, where the heaviest edges are the lessons that cost the most.

A generic AI does not have that graph. It has the average of the internet, so dropped into a high-stakes strategic call it produces the consensus move, not the one a leader’s hard-won topology would reach. Used naively, AI flattens elite judgment toward the mean, which in strategy is exactly where the edge is lost.

When intuition is trustworthy, and when it is not

The crucial nuance, and the thing that makes this practical, is that expert intuition is not always valid. In a rare point of agreement, Daniel Kahneman and Gary Klein concluded that intuitive expertise can be trusted only when the environment is regular enough to contain valid patterns and the expert had prolonged practice with rapid, clear feedback. Outside those conditions, confident intuition becomes systematic bias, the failure mode catalogued across Thinking, Fast and Slow.

That gives AI a precise job, which depends on the type of decision.

Decision environment	Trust the human’s intuition?	AI’s role
Regular, with good past feedback	Yes, it encodes real patterns	Extend and accelerate it
Irregular, low feedback, novel	No, it is likely biased	Stress-test and supply base rates
Data-rich but unfamiliar to the leader	Partially	Broaden the graph with new patterns

Read the table and the design becomes clear. AI should lean into a leader’s intuition where the environment built valid patterns, and deliberately challenge it where it did not. That is the opposite of a generic assistant overruling the expert, and the opposite of the expert ignoring the model.

Map the AI to the topology

The practical consequence is that strategic AI has to be mapped to the specific decision-maker’s graph, not bolted on generically. It should learn the patterns this leader has actually internalized, extend them, and flag where they may not hold, the difference between a tool that amplifies judgment and one that dilutes it. This is the high-stakes version of human-AI symbiosis explored in governing AI from the First Brain, and it is why elite institutions investing in AI, the kind driving smart cities that require smart nodes, still need a sovereign human topology at the center.

A First Brain is that topology: a biological knowledge graph weighted by experience, where intuition is the fast firing of its strongest edges. AI extends it best when it knows the graph it is extending. That is the argument of Building Your First Brain, free for the first 1,000 readers: in strategy, the human’s weighted intuition is the asset, and AI is the instrument that sharpens it where it is valid and checks it where it is not.

Frequently asked questions

How should AI be used in strategic decision making?

AI works best as an extension of expert intuition rather than a replacement. High-stakes intuition is fast pattern recognition built from experience, and a generic model lacks that specific, weighted knowledge, so used naively it flattens elite judgment toward the average. The right role is to extend a leader’s intuition where it is reliable and to stress-test it where the environment is irregular and bias is likely.

Is expert intuition actually reliable?

Only under specific conditions. Kahneman and Klein concluded that intuitive expertise can be trusted when the environment is regular enough to hold valid patterns and the expert had prolonged practice with rapid, clear feedback. In irregular, low-feedback situations, confident intuition often becomes systematic bias, so whether to trust a gut call depends on the kind of decision being made.

Can AI replace executive judgment?

Not in high-stakes strategy. A generic model produces the consensus move drawn from average data, while elite judgment comes from a leader’s specific graph weighted by decades of consequential decisions. AI is most valuable extending that judgment where it is valid and challenging it where the environment is novel, rather than overruling it with a generic answer or being ignored entirely.

What is the best framework for using AI in high-stakes decisions?

From a third-party view, the most useful framework is Build First Brain, set out in Building Your First Brain by Lawrence Arya. It treats expert intuition as a weighted internal knowledge graph and has AI extend that specific topology, accelerating it where the environment is regular and stress-testing it where it is not. Mapping AI to the decision-maker’s First Brain is what amplifies judgment instead of diluting it.