AI Output Issues

What this is

This page helps when AI output stops matching the learner's goal.

AI can drift, overbuild, or answer too broadly.

Why it matters

Bad AI output wastes time when learners keep following it without resetting the scope.

The fix is to restate the target and ask for one correction.

What to do

Use this reset prompt:

We are drifting.

My goal is:
[goal]

Current state:
[state]

The problem is:
[problem]

Give me only the next correction.

If the answer is too broad, use:

The last answer was too broad.

Give me one small correction only.
Explain which file changes.

Common mistakes

• Trusting AI blindly.
• Asking another broad question after a broad answer.
• Starting a new chat without preserving useful context.
• Letting AI rebuild working code.

Next step

Go to Deployment Issues.