Methods · Perspective, simulation & context · Short guide
D19
Pre-mortem simulation
Imagine the project already failed and work backwards to why: catching plan holes while you can still change them.
When to use it
- A project, essay plan, or experiment is about to start and you only imagine success.
- You want failure modes before they are expensive.
- A post-mortem after the fact is too late for this deadline.
What to ask AI
Type something like: “Pre-mortem: it is two weeks after the deadline and my dissertation chapter failed. Interview me on why, one question at a time. Do not comfort me. Then help me turn the top three causes into prevention checks.”
You invent the failure narrative; AI structures it.
How it helps
A pre-mortem imagines the project already failed and works backward, which surfaces risks optimism hides.
It is prospective, unlike an exam post-mortem after the fact.
How to run it
- Set a vivid failure in future tense.
- Generate causes without censoring the embarrassing ones.
- Rank by likelihood times damage.
- Convert the top causes into concrete prevention checks.
- Schedule those checks on a calendar.
Example exchange
Pre-mortem: imagine my group presentation already failed publicly. Why did it fail?
What did you not rehearse together that you told yourselves you would?
The handoff between slides 4 and 5,we each practiced alone and assumed it would be fine.
Prevention check: one timed joint run that includes that handoff, or it isn’t a plan yet. What’s the next failure you’d invent if that handoff were perfect?
Copyable prompt
Project: [ ]. Success looks like: [ ].
Lead a pre-mortem. Force me to generate failure stories in first person past
tense ("It failed because I…"). Push for specificity. Then help me cluster and
choose top 3 plausible, not most dramatic. For each, I propose a mitigation;
critique whether it actually lowers odds.
The Tell
You know this method has flipped when the pre-mortem was cathartic and nothing on your calendar changed.
Unscheduled prevention is daydreaming.
Top three causes must become dated checks.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Pre-mortem (imagine failure, work backwards) is a known decision-quality technique (Klein, prospective hindsight ) applied here to projects/plans. Principle moderate; AI-generated failure modes speculative.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
The learning principle may be moderate, but that does not prove an AI session delivers it well. Treat AI delivery as speculative unless a study of tutoring with this method is named, and none is claimed here.