Use · prompts
Prompts: forcing default → tail → critique
Reusable prompts that install the Less Typical pattern into AI workflows. Each prompt pairs a default-shaped request with a structural alternative and a critique step.
Prompts
These prompts exist to make the Less Typical move operational inside an AI workflow. They do not remove human judgment; they make human judgment the gate that decides.
1. Default → Tail → Critique
Use when the first answer feels too smooth.
List the 10 obvious answers first.
Identify the shared assumption hiding inside all 10.
Generate 3 structurally different alternatives that reject that assumption.
Argue against each alternative. Pick the one that survives.
This is the operational form of verbalized sampling and the AI as comparison surface move.
2. Diagnosis before explanation
Use when a student, founder, or teammate says “I don’t get it”.
Before you explain, ask:
1. What is the person currently doing wrong?
2. What specific line, decision, or moment shows that wrong move?
3. Which missing relation, if named, would make the explanation land?
Then explain only against that gap.
This installs the diagnosis before explanation move into teaching and advisory conversations.
3. Captured judgment interview
Use when a founder says “we just need better marketing copy”.
Ask the founder:
1. Which buyer looked serious but wasn't?
2. Which buyer looked hesitant but became valuable?
3. What phrase from a prospect makes you suspicious?
4. What kind of "yes" is fake?
5. What would you never put in the copy even if AI suggested it?
Turn the answers into a one-page lens the junior team can use.
This is the operational form of the captured judgment move.
Caution
Every prompt here assumes a human review gate. The prompt produces candidates; the human selects the one that survives contact with a real reader, student, founder, or buyer.
Related moves
ai as comparison surfacedifferent correct useful