Methods · Creation, transfer & long horizon · Short guide
F17
Portfolio critique
Have a body of work assessed for development over time, seeing trajectory and recurring weaknesses, not only the latest piece.
When to use it
- You have several pieces across a term and only ever fix them one at a time.
- A portfolio review is coming and you need trajectory, not local edits.
- The same feedback keeps appearing and you want it named as a pattern.
What to ask AI
Type something like: “Portfolio (dated): [pieces + intent]. Critique development over time: what improved, what plateaus, what recurs. Cite specific pieces. Do not rewrite. End with ONE pattern I should deliberately break next.”
Ask for patterns across pieces, not line edits.
Keep the hard part on your side: you choose what pattern to break next.
How it helps
Portfolio critique assesses a body of work for development over time, trajectory and recurring weaknesses, not only local polish.
Local edits can leave the same global flaw untouched across months.
Naming one pattern to break is more useful than twenty micro-suggestions.
How to run it
- Assemble dated pieces with a one-line intent for each.
- Ask for development critique: improvements, plateaus, recurrences, with citations to specific pieces.
- Refuse rewrites. Demand one pattern to break next.
- Plan the next piece to attack that pattern on purpose.
- Revisit the portfolio after that piece lands.
Example exchange
Here is a dated portfolio: September lab report, October essay, November design rationale. Critique development over time, cite pieces, and don't rewrite. End with one pattern I should break.
One clear improvement is methods clarity from September to November; see section 2 of the November piece. Where do you think you still plateau?
My openings still delay the claim, and my conclusions overclaim relative to the evidence.
Pick one pattern to break on the next piece. If you choose claim in sentence one, say how you will force that before you draft.
Copyable prompt
Portfolio (dated): [pieces + intent].
Critique development over time: what improved, what plateaus, what recurs. Cite specific pieces.
Do not rewrite. End with ONE pattern I should deliberately break next.
The Tell
If every piece gets local edits and the same global flaw persists, you never zoomed out to trajectory.
If the AI rewrote highlights instead of naming a pattern, you received editing, not portfolio critique.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated weak-to-moderate. Portfolio critique over time is authentic assessment practice. Development talk is valued in studio and writing programmes; controlled evidence for AI portfolio coaches is thin.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI pattern-finding across long texts is speculative and may hallucinate continuity. You must verify citations against your actual pieces.