Critique and Revise: Making the AI Improve Its Draft
Use criteria based critique loops to improve drafts, without falling into endless rewrites.
A taste of a lesson
I asked it to review my cover email and it said it was excellent. Then I said it felt weak and it agreed. Which is true?
Neither reply tells you much. Models lean towards agreeing with you, so both the praise and the agreement are partly echo. Get feedback with something to check against instead. Open a new chat, paste the email, and ask: 'List the three weakest points for a busy hiring manager, quoting the exact sentence for each. No praise.' Then judge each point yourself. Before you do that, write down one thing you want the hiring manager to remember after reading. That becomes your first criterion.
Written by the teacher as an example. In your lesson the tutor answers your own questions, and like any AI it can be wrong.
What you will be able to do
- Turn a vague wish to improve into three or four testable criteria
- Run a critique, choose, revise loop that keeps control in your hands
- Recognise sycophantic feedback and get more honest critique
- Decide when to stop revising before the draft goes bland
Lesson plan
- 1 From 'better' to criteria Define what improvement means for a specific piece before asking for help. Start
- 2 Asking for critique only Get a list of problems with evidence before any rewriting happens. Start
- 3 Targeted revision Revise only what you chose and keep the rest of the draft intact. Start
- 4 Sycophancy and pushback Understand why models flatter and how to get honest feedback. Start
- 5 When to stop Recognise diminishing returns and finish while the draft still sounds like you. Start
Try asking
About this tutor
Asking an assistant to 'make it better' usually produces a different draft, not a better one. This tutor teaches a disciplined critique and revise loop: you set clear criteria, ask for a critique against them, decide which points to accept, and then ask for a targeted revision. You learn why models tend to praise drafts too easily, why they sometimes change correct answers when challenged, and how to use a fresh chat or a sceptical reviewer role to get more honest feedback. You also learn when to stop, because after two or three rounds most drafts get blander rather than better. Bring a document you care about.
Reviews
4.3
4 ratingsSample
- Oluwaseun B.Sample
The accept, reject, unsure marking step sounds trivial but it stopped me from blindly taking every suggestion. My proposals keep their personality now.
- Hannah G.Sample
Learned why 'are you sure?' was flipping right answers. Practical lesson. Could use a short template sheet at the end.
- Elif D.Sample
Good ideas, but most examples were emails. I wanted more on longer documents like reports.
- Arjun V.Sample
I used to do eight rounds of edits. Now I do two with clear criteria and the results are better and less bland.
About the teacher
Teaches prompting one technique at a time, with plenty of before and after examples
9 tutors 339 lessons taught Sample
I teach people to write prompts the way I once taught people to write briefs: say what you want, who it is for, and what good looks like. I came to AI from editing and corporate training, so I care more about clear thinking than clever tricks. In my lessons we take one technique at a time, try it on...
See Chiara's profile and tutorsMore like this
Other tutors on the same or nearby topics.