Vithi vs ChatGPT: a study suggestion is not a study record

OpenAI’s guidance says ChatGPT can produce incorrect information and recommends checking important claims against reliable sources. A plausible study plan therefore needs your review, particularly when it includes dates, course coverage or external requirements.

Vithi’s mock form asks for an exam, score and date taken. Its portfolio activity form takes a title and description. Those fields organise information you provide; they do not independently verify a mark or a personal achievement. The same distinction applies to AI suggestions you encounter inside a student workspace.

OpenAI guidance on checking ChatGPT responses, reviewed 20 October 2026

Try this prompt with your own non-sensitive details: “I have 45 minutes tomorrow and these two question types to practise: ___. Suggest a short session with time to attempt questions, check answers and choose a follow-up. Ask if you need missing information. Do not invent my scores, exam date or course requirements.”

Inspect the proposed workload before using it. If the plan allocates all 45 minutes to answering questions, add time for checking. If it introduces a topic you did not supply, compare that suggestion with your actual course material instead of assuming it belongs.

This is a suggested prompt, not a report of a tested product outcome. You can use the same planning questions with a teacher, a friend or a paper notebook if a chat interface is unnecessary.

Suppose a proposed session includes a practice paper. “Take the paper” is still a future action until you complete it. Do not turn an estimated score in a conversation into an exam attempt. Use the marked result and actual attempt date when entering a record in Vithi.

For illustration, a completed paper might receive 63 out of 90. That example is hypothetical. Check that the selected exam matches the scoring scale, preserve the marked paper and record the actual score. A suggested target such as 70 is a different kind of number and should not be logged as an achieved result.

Keep a short correction note with the work: which error mattered, how you corrected it and what you will attempt next. The number alone cannot tell you whether the questions were easier or whether you repeated a misconception.

For a portfolio activity, draft the facts yourself before asking for wording help. State your role, the work you completed and the evidence available. If a rewritten description adds leadership, attendance figures or results you did not supply, remove those additions.

For example, “helped prepare the slides” should not silently become “led a team to deliver an award-winning presentation”. Save a description only when you can explain every claim in it. A polished paragraph is not stronger evidence than an accurate one.

Consider a conversation when you need to explore an approach, clarify a task or revise wording. Consider Vithi when you need a defined place for the exam attempt or activity you can substantiate. This is a manual workflow, not a claim of a ChatGPT integration, automatic import or equal capabilities across the products.

Use this handover record: suggestion received; facts checked; task chosen; work actually completed; source of the result; record entered; next action. Leave any unverified claim out of the saved result until you resolve it.

If your current notes already preserve these distinctions clearly, another app may be unnecessary. Try one real academic record before moving a larger archive, and judge the arrangement by whether it helps you recover accurate information later.

Try one verified record in Vithi