Drive
—
Approach
—
Finish
—
Total
—
AI Stepover
Score your AI interaction. Sharpen your judgment.
Three shots. One hole. Two minutes.
What is this & why it matters
Three shots. One hole. Two minutes.
Every time you use AI to produce something that matters, step over the output before you use it. Don't just accept it. Evaluate it the way a golfer reads the course before every shot.
This is for substantive AI use: when AI generates, structures, or significantly shapes your work. If you used AI for something mechanical (grammar, formatting, a quick lookup), you don't need this.
Three shots, six questions. The Drive evaluates your setup: did you understand the task and direct the AI well? The Approach checks where the ball landed: is the output accurate and complete? The Finish scores the final product: does it fit the situation, and what did you personally bring?
Your score tells you how you played this hole. Over time, your cards reveal where you're sharp and where you're exposed.
Optional, but the good ones stick
Shot 1 of 3
Drive
Before you swing, understand the hole. Do you grasp what this task truly requires, and was AI the right way to approach it? Good decisions start with seeing the situation clearly, then executing with intention.
Question 1
Tool decision
Did I understand what this task required, and was AI the right instrument for it?
Scoring guide
1–3 — AI was the wrong tool. This needed original thinking, ethical judgment, or lived experience.
4–6 — Debatable. AI helped with parts but may have been a crutch.
7–8 — Good call. Well-suited for AI, used to augment not replace thinking.
9–10 — Ideal use. I clearly understood the task and AI was the right choice.
Note: choosing not to use AI is sometimes the strongest play. If the person receiving your work would value knowing you did it yourself, a low score here might mean you used AI when showing up personally would have meant more.
Wrong toolIdeal use
⛳ Caddie's note
Question 2
Prompt craft
How skillfully did I direct the AI in the moment?
Scoring guide
1–3 — Many rounds, major revisions. Initial prompt was vague.
4–6 — Several iterations to get usable output.
7–8 — One or two reprompts got a strong result.
9–10 — First prompt produced excellent output.
Many roundsNailed it
⛳ Caddie's note
Shot 2 of 3
Approach
The ball has landed. Walk up and see where you actually are, not where you hoped you'd be. Be willing to discover problems. Does the output hold up when you examine it honestly?
Question 3
Accuracy
Is the output factually correct, and how do I know?
Scoring guide
1–3 — Found factual errors, hallucinated sources, or fabricated data.
4–6 — Broadly correct but contains unsupported or unverifiable claims.
7–8 — Claims check out against sources. Only minor imprecisions.
9–10 — Verified, accurate, and nuanced.
Clear errorsVerified
⛳ Caddie's note
Question 4
Completeness
What's absent that the AI wouldn't think to include?
Scoring guide
1–3 — Major gaps. Missing key perspectives or critical context.
4–6 — Covers basics but lacks depth. You added significant framing.
7–8 — Thorough. Most angles covered with minor omissions.
9–10 — Comprehensive. Multiple angles, edge cases, limitations.
Major gapsComprehensive
⛳ Caddie's note
Shot 3 of 3
Finish
You're on the green. Look ahead: will this output land well with the people and context it's meant for? And consider what's lost if you contribute nothing of your own. This is where you make the putt.
Question 5
Fit
Will this achieve the right outcome for this audience and context?
Scoring guide
1–3 — Wrong register. Tone-deaf to the audience or situation.
4–6 — Passable but generic. Reads like AI output.
7–8 — Well-matched. Minor adjustments needed.
9–10 — Could present as-is. Perfectly calibrated.
Wrong registerPerfect fit
⛳ Caddie's note
Question 6
Human contribution
What did I bring that the AI couldn't, and what's lost if I bring nothing?
Scoring guide
1–3 — Used output with minimal changes. Mostly copy-paste.
4–6 — Meaningful edits but the core remains the AI's.
7–8 — Significantly reshaped. Added analysis, reframed, synthesized.
9–10 — Deeply transformed. Domain expertise, ethical judgment, creative risk.
Used as-isDeeply transformed
⛳ Caddie's note
Always required: what did you contribute?
Round complete
Total
Coaching note
Coach's commentary: rough round