The 'explain the tradeoff' prompt

1. The habit

When you're in teacher mode — when you're trying to learn, not just finish — replace this prompt:

"Write me code that does X."

With this one:

"What are the two or three reasonable ways to do X? Explain the tradeoffs. Then I'll pick one."

Same problem, different relationship. The first version makes the AI the engineer. The second version makes you the engineer, and the AI the senior who taught you.

2. Why this works

Real engineering is mostly choosing. There are almost always two or three ways to do something, each one right for a different reason, and seniors know which one fits when. Juniors don't have that knowledge yet — and the reason they don't is that AI keeps making the choice for them, silently, in the background, and shipping one of the options without ever telling them the others existed.

When you ask for the tradeoffs, you force the choices into the open. You learn the shape of the decision space, not just the answer the AI picked. Next time, you'll recognize the shape and have your own opinion about which one to pick.

That recognition is taste. That recognition is also the thing that makes the difference between you and the model.

3. Variations of the same move

Same idea, applied to other situations:

• "I'm about to write a function that does X. What edge cases am I likely to miss?"
• "Here's my plan. What's a stronger alternative I'm not seeing?"
• "Walk me through the two or three patterns people use for this. Which fits when?"
• "Explain why your suggested approach is better than the one I started with — be specific."

All of them put the AI in the seat of senior who can see what you can't yet, and you in the seat of the engineer doing the work. Both seats need to be filled.

4. Try it

Take the next thing you're about to ask AI to do. Rewrite the prompt as "explain the tradeoffs, then I'll choose."