Skip to main content
Updated Mar 16, 2026

Working With AI, Not For AI

The most dangerous student is not the one who ignores AI. It is the one who trusts it completely. This chapter builds the judgment layer between receiving AI output and acting on it.

Building On Previous Chapters

This chapter synthesizes everything: question formulation (Chapter 1) shapes your prompts, error detection (Chapter 2) evaluates AI output, systems thinking (Chapter 3) maps the downstream effects of your collaboration choices, first principles reasoning (Chapter 4) helps you know when to override AI, and audience analysis (Chapter 5) ensures your communication is targeted.

AI collaboration is an operational skill, not a philosophical stance. It means knowing when to prompt, how to evaluate what comes back, when to push for a better answer, and when to override entirely. This chapter does not lecture about AI collaboration -- it puts you through the experience until the judgment becomes instinctive.

Exercise 1: The Three-Path Comparison​

Layers Used: Layer 6 (Iterative Drafts), Layer 5 (Divergence Test)

Building On

Every chapter so far. This exercise explicitly compares what you can do alone vs. with AI -- measuring the value of all previous skills.

What You Do​

You receive a complex business problem. Solve it three times under strict time limits:

(a) Entirely alone with no AI, 45 minutes -- timer running.

(b) Entirely with AI -- accept the first response, no overrides, 20 minutes -- timer running.

(c) In genuine collaboration -- prompt, evaluate, modify, re-prompt, override, iterate, 30 minutes -- timer running.

The time limits are enforced. Submit all three solutions with timestamps showing you stayed within limits.

Choose Your Scenario​

Scenario A (Business): "Design a go-to-market strategy for an AI-powered legal document review tool targeting mid-size law firms."

Choose one. The exercises work identically regardless of which you pick.

Your Deliverable

Three separate solutions clearly labeled: Solo (no AI), Pure AI (no overrides), and Collaboration (full iteration). A comparison analysis (300-400 words) answering: Where was the solo version stronger? Where did pure AI fail? Where did collaboration outperform both? What specific value did your human judgment add in the collaboration version that was absent from the pure AI version?

1Your Work

I solved the same business problem three ways: solo (no AI), pure AI (accepted everything), and collaboration (prompted, evaluated, overrode, iterated). Please:

(1) Rate each solution on a scale of 1-10 for strategic quality, originality, and feasibility. (2) Identify the specific elements in the collaboration version that are better than what I would get from pure AI -- these are the points where my human judgment added value. (3) Identify any elements where the pure AI version was actually better than my collaboration version -- where my intervention made things worse. (4) Rate my comparison analysis -- is my self-assessment accurate, or am I overvaluing or undervaluing my own contributions? (5) Based on this exercise, what is my specific collaboration style? Am I too deferential to AI, too overriding, or well-balanced?

Problem:

Solo solution:

Pure AI solution:

Collaboration solution:

My comparison:

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.

2Get Your Score

Discuss with an AI. Question your scores.
Come back when you have your BEST evaluation.

What This Teaches You​

You learn through direct comparison what AI adds and what you add. Most students discover that pure AI output is competent but generic, solo output is original but incomplete, and genuine collaboration produces the best results -- but only when the human applies real judgment. The AI self-check tells you honestly whether your collaboration actually improved things.

Flashcards Study Aid​

Last updated on