A Page Written Before the Outcome

A practice for the next consequential decision you make with AI input.

Before the outcome is known, write one page with four items.

The decision, in a sentence.

The reasons: the two or three things that actually moved you, including what the model said and how much weight you gave it.

The assumption you are least sure of. Only one.

What would change your mind: a specific signal, with a date.

Date the page and leave it alone. When the outcome arrives, whatever it is, read the page before you read the results. You are not checking whether you were right. You are checking whether your reasons were the reasons you remember.

Memory edits itself toward the outcome. Good results make weak reasoning feel strong, and bad results make sound reasoning feel foolish. The page is the only way to see the difference.

It also gives a team a way to learn from AI-assisted decisions without anyone having to be embarrassed.

Which decision this month deserves a page?