patrickz.aiLet’s talk ↗

Everyday AI / Guide

Ask AI to spot patterns in your unfinished projects

Turn a small record of project decisions into a tentative pattern and a testable rule.

About 20 minutesBeginnerRead free · No sign-up

Before you start

A chat assistant and three fictional or carefully selected project summaries.

Why this lesson exists

When Patrick asked an AI to look for patterns across his conversations with it, one thing it named was a small “graveyard” of demos that never became repeatable products. This guide keeps the useful part (a pattern you can check against evidence) and drops the risky part (a verdict about who you are). A model only sees the sample you give it.

The idea this guide practices: Use AI as a mirror, not a judge.

Do the exercise

  1. Choose a bounded sample

    Write what you started, what you finished and why you stopped for three projects. Use facts you are comfortable sharing. Leave unrelated private history out.

  2. Request competing explanations

    Ask for two plausible interpretations of the same pattern and the evidence for each. Make it distinguish missing records from a lack of action.

  3. Choose a behavior to test

    Convert one hypothesis into a small rule, such as completing a ten-minute demo before adding a new feature. Keep the rule reversible.

  4. Look for contrary evidence

    Apply the rule to the next project and record whether it helped. Ask what observation would disprove the original hypothesis.

A prompt to adapt

Replace the bracketed parts with your own details.

Review these three project summaries as a limited sample, not a personality assessment. Cite evidence for any pattern, give an alternative explanation, and name missing information. Suggest one reversible behavior experiment and what would disprove the hypothesis. Summaries: [summaries].

More prompts to adapt ↗

What this looks like

Stopping three projects after a prototype could mean avoidance, but it could also mean each prototype answered its question. The record needs the original goal before you can tell.

Check your result

Use evidence from your output. A confident explanation from the AI is not enough.

If it isn’t working

If the response uses sweeping labels about who you are, ask for specific actions and dates instead. Do not use a chatbot’s pattern summary as a diagnosis.

Where this came from

Adapted from the archived LinkedIn theme “I Asked My AI to Analyze Me. It Found a Prototype Graveyard.” See the post coverage and editorial method. One of the original LinkedIn posts ↗

Prepared September 2026. Tools and interfaces change; use current official setup instructions. Session lengths are estimates.

THE IDEA BEHIND THIS GUIDE

Read the story, then keep practicing

KEEP GOING

Your next useful step

Browse all 57 guides ↗