Chapter 7: Using AI to Think Better — A Practical Playbook

Part 8 of 9 in How to Think Better: Harnessing the Power of Counterintuition← Part 7Part 9 →

This chapter presents a practical system for using AI as a thinking partner without handing over judgment. It helps professionals, founders, parents, and other decision-makers identify biases, consider alternatives, and review decisions over time.

The previous chapters described the mental traps. This chapter turns them into a repeatable system that uses AI as a thinking partner. The goal is not to outsource judgment. The goal is to use AI to surface what your own mind is likely to miss, to generate options you would not have considered, and to create a record that allows you to improve over time.

The Decision Journal Powered by AI

Mauboussin strongly recommends keeping a decision journal. Write down the decision, the reasoning, the expected outcome, and how you feel. Review it later. This combats hindsight bias—the tendency to believe, after the fact, that the outcome was more predictable than it actually was.

AI makes this practical. After you write a short note, paste it into an AI and ask:

• “What biases am I most likely displaying in this reasoning?”
• “What is the relevant base rate I should have considered?”
• “What are three reasonable alternative explanations for the outcome I expect?”
• “What information would most change my mind?”
• “Summarize this decision in a way that will be useful when I review it in six months.”

A Simple Weekly Thinking Routine

Many Indian professionals and founders are too busy for elaborate processes. Here is a lightweight routine that takes 30–45 minutes per week:

1. List the two or three most important decisions you will face in the coming week.
2. For each, ask an AI for the outside view and a premortem.
3. Generate at least three genuine alternatives.
4. Write a one-paragraph decision note that includes the base rate and the key uncertainties.
5. Schedule a short review in four weeks to compare outcome with prediction.

Over a year, this routine produces a personal dataset of your own decisions. That dataset is far more valuable than any generic advice.

Prompts That Work Especially Well in the Indian Context

These prompts have been refined through real use with Indian founders, parents, and professionals:

• “I am evaluating a pitch from an Indian edtech founder. What are the historical base rates for Indian edtech companies that raised Series B between 2018 and 2022? What specific claims in a typical pitch should I treat with the most skepticism?”
• “My 16-year-old wants to drop out of regular school to prepare full-time for JEE. What does the evidence say about outcomes for students who take this path versus those who continue regular schooling with targeted preparation? What are the main risks?”
• “We are considering opening a second clinic in another Indian city. What are the most common reasons multi-location healthcare businesses in India fail to scale profitably?”
• “A relative is recommending a particular mutual fund based on recent performance. Help me evaluate this recommendation using the outside view and the distinction between luck and skill.”

Limits of AI

AI is not magic. It can hallucinate facts, reflect biases present in its training data, and lacks real-world accountability. Treat its output as a high-quality first draft of thinking, not as truth. Always verify important claims, especially numbers and specific historical outcomes. The human still owns the final decision. The best results come when AI expands the set of options and challenges assumptions, while the decision-maker remains responsible for the final choice and its consequences.