Discussion Guide — Chapter 6

1. "Nobody in this pipeline has to lie. Does that make it better or worse?"

Open here. Listen for: students working out that it makes it worse — a problem caused by liars can be solved by catching liars, while a problem caused by aligned incentives and format constraints cannot. Push on what that implies about solutions: if nobody is doing anything wrong, what exactly do you regulate?

2. "Who benefits from 'we don't know yet'?"

Ask it after building the incentive diagram, and let the silence sit. Listen for: someone eventually naming the reader — and then someone noticing that the reader is the only party with no ability to change what gets produced. That is the structural insight, and it lands harder when the room reaches it than when you state it.

3. "§6.9 says intelligent, skeptical, well-read people are the vector. Is that flattering or insulting?"

The chapter's most important discussion. Listen for: discomfort, then for someone articulating that the traits are genuinely virtues and that the environment converts them. Watch for students who exempt themselves. The gentle way in: ask whether anyone can name a compound or intervention they believe in and specify what would change their mind. The ones who cannot are demonstrating the point, and they should not be identified as doing so.

4. "Contempt is not a communication strategy. Do you agree?"

Particularly valuable with clinical students. Listen for: the recognition that dismissing a category teaches a patient that you do not understand their situation, and that they will then discount your accurate information along with your inaccurate framing. Then the harder question: what do you say instead, in eleven minutes, to someone who arrived with printouts? Chapter 39 answers it; asking here makes them want the answer.

5. "Is the gray market net positive?"

Genuinely open and worth running properly. The case for: people get access to compounds that regulators are slow to approve, and the alternative is waiting while suffering; information asymmetry harms people who will self-administer regardless; harm reduction is a legitimate tradition. The case against: unverifiable quality, no monitoring, no reporting pathway, no recourse, and an information environment optimized to sell rather than to inform. Do not stack the deck. A student who can state the strongest version of the position they reject has learned something a student who agrees with you has not.

6. "This book says it has its own biases. What would you check?"

Listen for: procedural answers rather than attitudinal ones. "Check whether the six trial-quality questions get the same answers for compounds the author likes and dislikes" is a good answer. "Read it critically" is not. This question also rehearses Chapter 37's drift exercise, and students who take it seriously here do that one much better.