The Book of Why book club questions

1. 1.

   Pearl argues that avoiding causal language was a founding mistake of modern statistics. Do you find that argument convincing, or does it seem like an overstatement?

2. 2.

   Think of a headline you've seen recently that claimed a correlation. What causal question was the headline implying, and what would it actually take to answer that question?

3. 3.

   The ladder of causation distinguishes seeing, doing, and imagining. Can you think of decisions in your own life that required you to reason at each of these levels?

4. 4.

   Pearl says current AI can't reason causally. Does that match your experience with AI tools you've used? What kinds of questions does it seem unable to answer?

5. 5.

   Randomized controlled trials are the gold standard for medical evidence. Pearl discusses their limits. What important questions in health or policy seem impossible to randomize?

6. 6.

   Causal diagrams make assumptions explicit. In your own field, what causal assumptions are usually left implicit? What would happen if they were forced into the open?

7. 7.

   Pearl argues that counterfactual reasoning is the basis of moral judgment — assigning credit and blame requires imagining what would have happened otherwise. Do you find that plausible?

8. 8.

   The book traces the intellectual history of causality through statistics, genetics, and economics. Were you surprised by how long this debate has been going on?

9. 9.

   Pearl is openly polemical about AI researchers who think better pattern-matching will eventually produce human-like intelligence. Do you think he's right that causal structure is the missing piece?

10. 10.

    What is a causal claim in your professional domain that is widely assumed but rarely formally tested? What would it take to test it?

11. 11.

    Pearl writes that asking 'why' is uniquely human. Do you think that's true? What does a world look like in which machines can also genuinely ask why?

How to run the discussion

• Pick six to eight, not all of them. An hour of real discussion covers fewer questions than anyone expects, and racing the list is what makes a book club feel like homework.
• Open with what people noticed, not what they concluded. Asking for a verdict first hardens opinions before anyone has heard each other.
• Let the disagreement run. The questions below that split a room are the ones worth the time; the ones everyone answers the same way can be skipped.
• Ask the quiet person a specific question. A direct, narrow question is far easier to answer than an open invitation to contribute.