Which tool to choose
- For explanations tailored to you and training questions: a general-purpose assistant, which you ask to adapt the language to your level.
- To study from your own materials with the sources under control: NotebookLM, which works only on the uploaded documents.
- To organize your time and keep the calendar: an assistant that builds the plan, plus a real calendar app for the reminders, because the AI doesn't alert you on its own.
How to do it
- Find out how you study right now. Ask the AI to take stock. The operational syntax:
Let me tell you how I study: [describe: you read, underline, summarize...]. Tell me which of these habits learning research considers ineffective and what you'd replace them with, one at a time.
- Move from rereading to retrieving. For each topic, instead of having a summary made for you, have it ask you questions to answer from memory. The effort of pulling out the information fixes it.
- Unlock difficult concepts with different explanations. The operational syntax:
I didn't understand [concept]. Explain it to me in three different ways: with an everyday analogy, with a numerical example, and as you'd explain it to a ten-year-old child. Then ask me a question to check whether I've understood.
- Build a realistic plan. Tell the AI how many hours you really have, not how many you'd like, and have it distribute the topics across the days, leaving room for spaced review (going back over things days apart, not all at once).
- Check and adjust every week. Ask: "let's recap what I've learned this week with five questions" and, based on how you do, have the plan retuned. A method is good if it corrects itself, not if it's perfect on paper.
A concrete example
Elena studies engineering and notices she forgets everything after the exam. She tells the AI her method: she reads and rereads. The AI points out that rereading deceives and proposes active recall. Elena starts having herself quizzed instead of rereading. On the concept of the derivative, which she'd never digested, she asks for three different explanations: the analogy with a car's speed unlocks it. She builds a plan around her real three evening hours. After a month, at the exams she remembers the concepts, she doesn't fish them out at the last minute. She changed her method, not the number of hours.
When it does NOT work (and how to fix it)
If you keep having the answers given to you instead of looking for them
It's the most common trap: the AI solves the exercise and you think you've understood. Remedy: ask it to give you only the first step and to let you continue, or to correct your attempt instead of replacing it. Seeing the solution doesn't teach you how to build it.
If the study plan is unrealistic and you abandon it
It happens when you declare the hours you'd like to have. Remedy: give the AI your real time, including the days when you don't study, and ask it for a "bad-week-proof" plan, with a margin. A plan that survives the unexpected is the only one you follow.
If you can't tell whether you're improving
Remedy: have it give you the same type of quiz on a topic at the beginning and after two weeks. If you answer better and with less effort, the method works. Improvement is measured by recall, not by the feeling of having studied.
A tip from someone who really uses it
Keep a study journal in which, at the end of the day, you write to the AI what you understood and what stayed foggy. It becomes your weak-points card: you reread it before the exam and you know exactly where to insist. This document is yours and follows you even if you change assistants, while the internal memory of a single AI does not.
Frequently asked questions
Can the AI figure out the right method for me?
It can propose techniques that work for most people and adapt them to your time, but the right method you find by trying. Use it to experiment quickly with different strategies and keep the one that makes you remember the most: the final judge is you, not it.
Does studying with AI make me dependent and less autonomous?
It depends on how you use it. If you have it do the work, yes, you atrophy. If you use it to quiz yourself, explain the obscure points and organize yourself, you develop a method that stays even without it. The tool that trains you to think on your own is the opposite of dependence.
Is it true that those who study with AI learn faster?
Faster, no; better, yes, if they use it for active recall and tailored explanations. Speed is a marketing promise: long-term memory has its own pace and the AI doesn't shorten it. What changes is the quality of studying, not the stopwatch.