Exercise 2: AI vs. Human Systems Analysis
Layers Used: Layer 2 (Reasoning Receipt), Layer 5 (Divergence Test)
You will use the Error Taxonomy from Chapter 2, Exercise 1 to annotate errors in AI's systems analysis, not just factual claims.
What You Do
Now prompt both Claude and ChatGPT with the same scenario and ask each for a comprehensive analysis of all consequences. Compare both AI outputs against your own cascade map. Typically, AI produces a broader but shallower analysis — more categories, fewer connections between them. Create a merged map (Draft 2) that combines the best of human and AI analysis with clear attribution for each insight.
A comparison document with three columns: "Effects only I found," "Effects only AI found," and "Effects we both found." The merged cascade map (Draft 2) with every insight color-coded or labeled by source: Human (H), Claude (C), ChatGPT (G), or Synthesis (S) for new insights that emerged from combining perspectives. A brief note explaining which category had the most valuable additions.
I am comparing my systems analysis with AI-generated analyses of the same scenario. I have created a merged cascade map that combines insights from my own thinking, Claude's analysis, and ChatGPT's analysis, with each insight attributed to its source.
Please: (1) Evaluate my merged map -- is it genuinely better than any single source alone? (2) Are there insights I attributed to myself that are actually standard AI outputs? Be honest. (3) Are there synthesis insights (S) that are genuinely novel -- combinations that none of the three sources produced independently? (4) Rate the quality of my attribution -- am I being honest about where each idea came from? (5) What important systemic effects are STILL missing from the merged map?
Scenario:
My original map:
Claude's analysis:
ChatGPT's analysis:
Merged map with attribution:
Finally, complete the Thinking Score Card for this exercise: Independent Thinking (1-10), Critical Evaluation (1-10), Reasoning Depth (1-10), Originality (1-10), Self-Awareness (1-10). For each score, give a one-sentence justification.
Discuss with an AI. Question your scores.
Come back when you have your BEST evaluation.
What This Teaches You
You learn exactly where human systems thinking adds value that AI misses (usually in feedback loops and cultural/political dynamics) and where AI adds value humans miss (usually in breadth of categories). The attribution exercise forces intellectual honesty — you cannot claim AI's insights as your own when the source is documented.