AI for Decision-Making

• Use AI to explore options and trade-offs
• Get help with pros-and-cons analysis
• Use AI as a devil's advocate
• Know the limits of AI for important decisions
• Maintain human judgment in the loop

Decisions are hard. Not because we lack information, but because we have too much of it, we cannot see all the angles, and our own biases cloud our judgment.

AI cannot make decisions for you. But it can be a powerful thinking partner that helps you see options you missed, challenge assumptions you did not know you had, and structure your thinking more clearly.

This lesson shows you how to use AI to make better decisions while keeping human judgment where it belongs: in charge.

One of the most common decision-making errors is premature closure: settling on an option before exploring alternatives. AI can expand your option set.

Generating Alternatives

Basic prompt: "I'm trying to decide [decision]. What options am I not considering?"

Better prompt: "I'm deciding whether to [option A] or [option B]. But I might be missing other options entirely. What are 5 alternative approaches I haven't considered?"

Challenging Assumptions

Prompt: "I'm assuming [assumption] in my decision about [topic]. Is that assumption valid? What would change if it weren't true?"

Prompt: "I've narrowed my job search to [industry]. What assumptions am I making about what kind of work I need? Are there other paths that might meet my actual needs?"

The "Third Option" Technique

Prompt: "I'm stuck between [option A] and [option B]. Neither feels right. Is there a third option that combines the best of both? Or a completely different approach?"

Many binary decisions have hidden alternatives once you look for them.

The classic pros-and-cons list is more powerful with AI assistance.

Building the List

Prompt: "Help me create a pros and cons list for [decision]. Include factors I might not have thought of."

Better prompt: "I'm considering [decision]. Create a comprehensive pros and cons list. For each point, rate its importance (high/medium/low) and explain why."

Weighted Analysis

Prompt: "Here are the factors I'm considering: [list]. Help me weight them by importance. Which 3 factors should matter most and why?"

Prompt: "I've listed the pros and cons of [decision]. But I'm not sure I'm weighting them correctly. Challenge my priorities."

Comparing Options

Revealing Hidden Trade-offs

Prompt: "I'm leaning toward [option]. What would I be giving up that I might not realize? What might I regret later?"

When you are leaning toward a decision, confirmation bias makes you seek evidence that supports it. AI can deliberately argue the other side.

The Devil's Advocate Prompt

Prompt: "I'm planning to [decision]. Play devil's advocate: Give me 5 strong reasons this might be a mistake. Be aggressive. Assume you're trying to talk me out of it."

Prompt: "I've decided to [decision]. What would a skeptic say? What are the strongest arguments against my choice?"

Challenging Specific Beliefs

Prompt: "I believe [belief related to decision]. Argue against this. What evidence might suggest I'm wrong?"

Prompt: "I'm confident that [prediction]. What are the ways this could turn out differently than I expect?"

Steelmanning the Other Side

Prompt: "I'm rejecting [option] because [reasons]. Make the strongest possible case FOR that option. What am I missing?"

This is called "steelmanning": creating the strongest version of the opposing argument, not a strawman.

Stress-Testing Your Reasoning

Prompt: "Here's my reasoning for [decision]: [explain your logic]. What are the weak points in this reasoning? Where might I be fooling myself?"

Sometimes you need a structured approach to a complex decision. AI can suggest and apply decision frameworks.

Getting Framework Suggestions

Prompt: "I'm facing a complex decision about [topic]. What decision-making frameworks might be useful? Explain how I'd apply each one."

Common Frameworks AI Can Help With

10-10-10 Analysis: "For [decision], help me think through: How will I feel about this in 10 minutes? 10 months? 10 years?"

Pre-Mortem: "Imagine I made [decision] and it turned out badly. What went wrong? What caused the failure?"

Regret Minimization: "When I'm 80 years old looking back, which choice will I regret more: doing [option A] or not doing it?"

Reversibility Check: "How reversible is [decision]? If it doesn't work out, what are my options? How hard is it to change course?"

Building Your Own Framework

Prompt: "I need to evaluate [type of decision] regularly. Help me create a simple framework or checklist I can use each time."

AI is a thinking tool, not a decision-maker. Here is why the final call must always be yours.

What AI Cannot Know

• Your values: What matters most to you (career vs. family, security vs. adventure, etc.)
• Your context: The full complexity of your situation, relationships, and history
• Your gut feeling: The intuition built from all your experiences
• The weight you give factors: Which considerations truly matter most to you

The Information Problem

AI can only work with what you tell it. You know things about your situation you did not think to mention. You have tacit knowledge that cannot be easily articulated.

Value Judgments

Many decisions hinge on questions AI cannot answer:

• Is it worth the risk?
• How much does money matter vs. fulfillment?
• What can you live with?
• What do you actually want?

When to Trust AI Analysis

Use AI insights when:

• You need to generate options or see blind spots
• The decision involves information processing you might miss
• You want to stress-test your reasoning
• You need structure for complex analysis

When to Trust Yourself

Trust yourself when:

• The decision involves your core values
• You have relevant experience AI cannot capture
• Your gut strongly disagrees with the analysis
• The stakes are high and very personal

• Expand your options: AI helps you see alternatives you missed
• Structure your analysis: Pros/cons, frameworks, weighted factors
• Argue against yourself: Use AI as devil's advocate to challenge assumptions
• Apply decision frameworks: 10-10-10, pre-mortem, regret minimization
• Keep humans in charge: AI informs; you decide

Think of a decision you are currently facing. Work through this exercise:

1. Ask AI: "I'm trying to decide [your decision]. Ask me 5 questions that will help you understand my situation better."
2. Answer the questions honestly and thoroughly.
3. Ask AI: "Based on what I've told you, what options am I considering? What options might I be missing?"
4. For your leading option, ask: "Play devil's advocate. Give me 5 strong reasons this might be a mistake."
5. Finally, ask: "What's the most important question I should ask myself before deciding?"

Notice that AI helped you think through the decision more thoroughly, but the final choice is still yours.

• Research on AI-assisted decision making: https://www.nature.com/articles/s41562-023-01648-z
• Cognitive biases in decision making: https://www.annualreviews.org/doi/10.1146/annurev-psych-081920-024932
• Pre-mortem technique effectiveness: https://hbr.org/2007/09/performing-a-project-premortem