Retri
Give each conversation a finish line
Before asking another follow-up, decide what you want to leave the conversation able to explain or do. “Understand the tradeoff between caching and freshness” gives a session a finish line. “Tell me everything about caching” can keep expanding.
When you reach that finish line, close the explanation and try to describe the idea yourself. A blank spot gives your next question a direction. You don’t need a complete mental copy of the conversation.
Try this prompt:
I have enough background for now. Ask me one question that checks whether I can explain the caching tradeoff, then help me correct any gap in my answer.Keep a reference list and a recall list
Some information belongs in a place you can search later: a link, a long procedure, a detailed specification. Other information is worth having ready in memory: the meaning of a term, a distinction you need in conversation, or a rule you repeatedly apply.
Before making a flashcard, ask: Where will I use this without the answer in front of me? If you can name a situation, the card has a purpose. If the answer is vague, bookmark the material or leave it in the chat.
For a language learner, a phrase needed in conversation might belong in the recall list. For someone researching a new hobby, a supplier’s address probably belongs in reference notes. The right choice depends on what you’re trying to do.
Check recall rather than familiarity
An explanation can feel familiar when you reread it. That feeling alone doesn’t tell you whether you can produce the answer later. In Karpicke and Roediger’s 2008 vocabulary study, learners’ predictions of their performance did not track their actual performance. Read the study.
Use a small question to make the gap visible. Can you explain the difference without reopening the answer? Can you give an example? If not, return to that specific point.
You can use spaced repetition to organize later attempts. The practical suggestion here is to build a routine around the ideas you actually need, rather than save every interesting sentence.
Give your review queue a time budget
Decide how much time you’re willing to spend reviewing on an ordinary day. You might begin with five minutes and adjust after a week. Treat that as an experiment and adjust it to your material and schedule.
Review the cards already due before adding a large batch. If the queue consistently takes longer than your budget, pause new cards. Suspend low-priority material and rewrite questions that repeatedly leave you guessing what they mean.
Each saved card is a future review commitment. Being selective now helps keep the routine useful later. A large deck you avoid opening has little practical value, however thorough it looks.
Turn one useful chat into a small practice loop
After a conversation, write down the main idea in your own words. Pick a few questions that would help you use it. Check the answers, save the cards, and attempt them again when they’re due.
From this conversation, suggest three questions worth remembering for my current goal. Explain why each would be useful. Show me the drafts before saving anything.With the Retri ChatGPT plugin, you can save approved cards to a deck and ask to review the due cards later. Follow the step-by-step workflow for setup and review prompts.
On days when you have more questions than time, review what you already chose to learn. You can return to the reading list when there’s room for another topic.
Make room for work beyond the cards
If you’re learning a language, use the phrases in a conversation. If you’re learning to code, build something small. If you’re preparing for an interview, practice answering a question aloud and responding to a follow-up.
A card can help you recall an ingredient of a skill. Using that ingredient in a real task gives you a different way to check what you understand. Keep notes on the problems you encounter; those gaps can guide your next learning session.
For your next ChatGPT conversation, choose a finish line before the first prompt. End with one idea you can explain and a small amount of practice you’re willing to repeat.