Anyone who takes the first AI draft as final
No metric reported. Ethan Mollick (Wharton) argues the best way to use AI is interactively rather than hunting for a perfect prompt. In the comments of his guide, Conor Grennan shares a technique he credits to Mollick: after an answer, ask the AI to critique its own response, poke holes in it, then improve it based on that critique.
Mollick's broader tip: give context and constraints, for example 'provide it a few paragraphs you have written and say using the style in the paragraphs below, can you write ___.'
Here is a draft for [audience] whose goal is [goal]: [paste draft] Step 1: Critique it as a tough editor. List the 5 biggest weaknesses: where it's vague, where it's boring, where a reader would stop, what's missing, and any claim that isn't supported. Quote the lines. Step 2: Poke holes in your own critique. Which of your points are wrong or would make the piece worse? Step 3: Rewrite the draft fixing only the points that survived Step 2. Keep my voice and length. Show the rewrite and a short list of what changed.
Legacy Builder Lab is Clifton Sellers' community for founders who write their own content with AI. Members get his prompts and bot setups, five new post templates every Monday and a monthly live teardown.
Join the Lab waitlist