ChatGPT AI Guide: Reconcile kurta Try-On Feedback Without Voting

By TryBuy Editorial

ChatGPT can organise a two-person kurta try-on discussion without turning it into a vote about the wearer’s body or style. Ask for an attributed log that keeps wearer experience, visible observations and preferences separate. The deliverable is a clear next-action queue, not a model’s fit diagnosis or a majority decision about which outfit someone should wear.

Collect the two voices accurately

Label the wearer’s comments and the observer’s comments before supplying them. Preserve the actual meaning rather than rewriting the notes to agree. Include the garment identity, selected size, supporting clothes and ordinary action being assessed. If one person did not identify the view or movement behind their comment, leave that context unknown.

State the question the session was meant to answer. You might be checking a sleeve position or comparing the effect of an outer layer. Keep unrelated opinions about colour or occasion in a separate preference field. This prevents the model from allowing a confident aesthetic comment to override the wearer’s reported restriction or uncertainty.

Copy the feedback-log prompt

Organise these labelled kurta try-on comments into a decision log. Keep wearer reports, observer-visible events and styling preferences distinct. Preserve garment identifiers, selected sizes and the action associated with each observation. Do not diagnose a construction cause, infer consensus or rank the wearer’s body. Where comments disagree, record the disagreement and the smallest clarification or repeat observation that could resolve it. Assign actions only to supplied roles and leave the final clothing decision with the wearer.

Ask for a source-comment pointer on every row. A clean summary can otherwise erase an important distinction, such as an observer describing how a sleeve looks while the wearer describes how it feels during a particular action.

Filled fictional example

Suppose the wearer reports that reaching feels restricted, while the observer says the standing photograph looks balanced. In a fictional log, both statements can be retained because they answer different questions. The model should not decide that the attractive photograph disproves the wearer’s report, nor should it infer a particular seam problem from the report alone.

The next action may be repeating the relevant ordinary movement with the same supporting outfit and recording the observation more precisely. If a professional assessment becomes appropriate, the log can prepare a neutral handoff. It should not tell the observer to pin or alter the garment to demonstrate a preferred shape.

Use revisions that preserve ownership

• “Show which statements came from the wearer and which came from the observer; restore any attribution lost in the summary.”
• “Move colour and silhouette preferences outside the fit-observation queue.”
• “Rewrite the unresolved row as a neutral question rather than a diagnosis or a decision for the wearer.”

If the wearer clarifies their experience, add that as a new labelled note. Do not erase the earlier comment simply to make the record appear consistent. A short explanation of the clarification is more useful than a tidy history that hides why the next action changed.

Review the log together

Read the summary back with the people involved where appropriate. Check that the language does not exaggerate, minimise or substitute their observations. Confirm that an observer’s opinion remains an opinion and that the wearer’s preferences have not become universal fit rules. The model does not experience the garment or inspect its hidden construction.

Use the original practical guide for running the practical two-person session and the men’s kurtas collection when another design is genuinely part of the wearer’s next decision. Your finished deliverable is an attributed log and a small clarification queue. It should help the wearer understand the evidence and choose their own next step, rather than adding an AI judgement to a conversation that already contains competing opinions.

For general instructions and follow-up techniques, see official OpenAI prompting guidance.