Methods · Embodied, practical & performance skills · Short guide
N5
Post-performance forensic
After a real attempt, reconstruct the sequence and locate the first deviation, so improvement targets the branch point, not the final mess.
When to use it
- You just finished a real attempt and the ending felt embarrassing, but you are not sure where it actually went wrong.
- The same crash shows up every session and your notes only say “panicked” or “rushed.”
- You made failure-point predictions earlier and want to compare them within ten minutes.
What to ask AI
Soon after the attempt, narrate what you did in order, plus what the plan or cues were supposed to be.
Ask it to help you find the first deviation and the belief behind it, one corrective drill only, no full lecture.
Keep the hard part on your side: you reconstruct; it probes for the earliest branch point.
How it helps
People remember the embarrassing ending, not the first wrong micro-choice.
By the time the plate hits the floor, the serve nets, or the demo stalls, three earlier decisions have already branched.
Forensic reconstruction finds the earliest deviation, the same “first error” logic as strong tutoring feedback, so tomorrow’s drill targets the branch point, not the final mess.
Paired with earlier failure-point predictions, you also learn which watch-outs were real.
How to run it
- Within about ten minutes of the attempt, narrate what you did in order, raw, not polished.
- Mark where the sequence diverged from your plan or cues.
- Identify the first deviation and the belief that made it seem okay in the moment.
- Design one corrective drill aimed only at that branch point.
- If you predicted failures beforehand, compare: which predictions hit?
Example exchange
I just did a short oral demo of the pipette protocol. The plan was open stock, change tip, fill A then B then C, then recap. Help me find the first place I left that plan and why I thought it was okay, then one drill only.
Walk me beat by beat from picking up the stock. Stop when you feel yourself editing the story.
Pick up stock, open, aspirate, then I go A and B with the same tip because I was speeding for the audience, then C looks cloudy, I panic-recap, spill on B.
The spill is the ending. The first deviation is reusing the tip from A into B. What belief made that feel okay in the moment?
That the audience was waiting, and changing tips would look fussy.
One drill: three transfers with a forced tip change between each, timer off, no audience. Don’t practise “not spilling” , practise the branch that never opens if the tip changes.
Copyable prompt
After-action narration: [ ].
Plan/cues were: [ ].
Help me find the FIRST deviation and the belief behind it. One corrective drill
only; no full lecture.The Tell
Here is how you know this method has flipped: the same ending failure shows up every time, and you never name a first deviation, only “I panicked” or “it fell apart.”
In that moment the outcome did the thinking, and the branch point stayed invisible.
Force the reconstruction until you can point to one early choice and one drill that targets only that.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Reconstructing a sequence to the earliest deviation is reflective practice and error analysis for motor skill, the same first-error logic used in good tutoring. You target the branch point, not the embarrassing ending.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI forensic interview quality is speculative. It can push you past the ending into earlier steps; it cannot see the attempt. Your narration and the single corrective drill are the learning event.