Making Better Decisions with AI

Separate objectives, alternatives, constraints, assumptions, and uncertainties.

Identify who owns the decision, what matters most, and what can actually be controlled.

Distinguish correlation from causation and spot weak or misleading claims.

Challenge sources, assumptions, selection bias, and AI-generated evidence before relying on it.

Express uncertainty with probabilities instead of vague terms like “likely” or “unlikely.”

Use base rates and new evidence to update your beliefs rather than defend your first view.

Use scenarios, sensitivity analysis, and downside risk to test whether a decision is robust.

Find the tipping point, the assumption or threshold that would change your recommendation.

Decide whether additional research, testing, or delay is worth the time and cost.

Identify which unknowns are worth resolving and which are unlikely to change the decision.

Turn your analysis into a concise decision memo with clear reasoning, risks, and next steps.

Use AI to challenge your recommendation, then make and defend the final call yourself.