How I Use AI to Study
As the new school year approaches or has already started for many, I thought it might be useful to talk about a few ways that AI can be integrated into the study workflow. There is no doubt that AI has changed, and will continue to change, the way we learn. I know there are several skeptics who believe Large Language Models (LLMs) are curbing our ability to think and reason on our own.
However, I (and many others) have found that AI can also be an exceptionally powerful tool that can help increase my productivity and efficiency. Used with integrity, I think AI can bolster learning by taking on the “busy” work, and leaving me to focus on active learning and recall. As I have become more comfortable with AI models and prompting, I have also begun to use it for much more than explaining concepts I do not understand. Now, I use it as a combination of tutor and personal assistant to help plan my week, organize course material, build study resources and workflows, and identify and address my weaknesses.
In the interest of full disclosure, the model I primarily use (and pay for) is ChatGPT, particularly its new 5.6 Sol model. I have heard that Claude’s new Fable model is comparable in terms of the complexity of tasks it can handle, but I have not personally tested it for similar functions.
1. AI can help you plan your week
Let's be honest, planning around several competing priorities and deadlines requires quite a bit of cognitive effort. Trying to keep a mental map of everything is very difficult from day to day, and planning the week ahead can take hours as I try to figure out how everything fits together.
So, I have AI help me plan my time, particularly during busy weeks or weeks when I know there is a lot to get done but not a lot of structure (e.g., the week before final exams).
First, I dump everything into AI. I include a list of:
What I need to get done
Each item’s due date
How long I expect each task to take (usually this is a range of time)
Which items are most important
What hours I will have available during the week
Then, I ask it to turn all of that into a realistic schedule.
For example:
To do:
Convert Chapters 5–8 notes into flashcards (2-3 hours; due Monday night)
Do a pass through the flashcards (1-2 hours; due Wednesday night)
Complete amino acid titration practice problems (2-3 hours; due Wednesday night)
Study time available:
Monday: 8–10 a.m.; 12–4 p.m.; 9–10 p.m.
Tuesday: 2–6 p.m.
Wednesday: 8–9 a.m.; 6–10 p.m.
I find this particularly helpful because it forces me to confront the week ahead of time. If I have somehow assigned myself 15 hours of work on a day when I only have four hours available, I would much rather discover that on Sunday than on Wednesday night. Also, I can protect time to take care of myself, so that exercising, eating, and spending time with family and friends doesn’t get left by the wayside. Lastly, it helps manage stress by eliminating that nagging feeling of having 20 different things floating around in my head.
Once I have a plan, I can focus on whatever I am supposed to be doing that day without worrying so much about how I will manage competing priorities.
2. Before class, I feed AI my lecture notes and ask it to explain the fundamental principles
I find myself much less overwhelmed by the content of a lecture if I have a rough idea of what it will cover before going in. With AI, this is so easy to do. The night before or day of my class, I feed the PowerPoint or lecture topic into AI and ask it to give me a quick, very basic explanation of what the lecture is about, why it matters, and where we are going. That mental roadmap and vague familiarity with the major concepts and terms make the material much more manageable because when I see them in class, it is not my very first exposure.
3. I use AI to build a review sheet
After a lecture, I will ask AI to take my lecture notes and build a study summary sheet from them. Importantly, you can ask AI to tailor this to how you like to learn. For example, I like to know why what we are learning matters and how it connects to other material, so I ask AI to emphasize those connections.
You can even ask AI to help you write a reusable prompt for building your study sheets. For example, you could say:
“Each week, I want to turn my lecture notes into a summary sheet that covers the major themes, concepts, and examples. I learn best through lots of examples and metaphors drawn from everyday life. Please ask me a few additional questions about my learning style, then help me generate a prompt I can use to build an effective study sheet each week.”
There are also a few other things you could ask for explicitly:
A list of key terms and definitions
A list of learning objectives
For example, you may need to be able to solve a certain type of problem or identify how a particular condition affects an analytical technique.
Turning the material into a set of skills you need to master is a good bridge between reviewing the content and studying for a test. It can also help you identify weaknesses early.
Guidance for mastering each learning objective effectively and efficiently
This might include a step-by-step guide to solving each problem type.
AI can help you turn practice examples into workflows for different types of problems. It can also help you anticipate common “exam-style” tricks.
4. I turn that material into flashcards and active-recall questions
Once I have a study guide, I use AI to help me generate practice material from it. For major terms and concepts, I will usually include a definition flashcard. However, I try to avoid making the majority of my flashcards simple definition questions.
For example, I might have:
What is an SN2 reaction?
But I would also include questions like:
What clues indicate an SN2 reaction?
How do you differentiate an SN2 reaction from an E2 reaction?
Given this substrate, nucleophile, and solvent, which reaction pathway would you predict, and why?
Or, if you know you prefer a different type of flashcards based on your learning style, AI can help tailor those accordingly. Again, I recommend asking AI to work with you to identify your unique learning preferences and generate a reusable prompt that can turn your lecture notes into effective flashcards.
5. When something does not make sense, I go back to the foundations
Iterating with AI about a specific homework problem or concept I am stuck on is one of the most powerful ways I have found to use AI. I think of it as using AI as my personal tutor. Sometimes, I can tell that an explanation is not clicking because I am missing, or am not completely clear on, something more fundamental.
When that happens, I ask AI to go backward. If I am stuck on a problem, I ask it to “walk me through this problem step by step, building from first principles.” If I got the problem wrong, I ask the AI to walk me through it while also explaining why my answer was wrong and how to reach the correct answer.
In addition to sharing the problem itself, it is also very helpful to explain your thinking and reasoning – even if the thinking is to articulate your confusion – to the AI. It can use that information to identify what you understand, what you do not understand, and where specifically your logic has gone astray. The more specific you are, the better.
Even after I think I understand the correct answer, I will often briefly explain my understanding back to the AI. This gives it a chance to sharpen my reasoning, or simply confirm that I am on the right track.
One other trick I use when I am stuck but still want to figure out the problem myself is asking AI to walk me through it using hints without solving it for me. You can also ask AI to pose questions along the way to check your understanding. This helps you practice articulating why you are taking each step, which is useful for identifying points of confusion.
6. I note where I struggle and ask AI for problems targeting that exact weakness
Once I understand a concept, I want to know whether I can actually use it. Ideally, this consolidation occurs just after it clicks. Rather than repeating the same problem where I can rely on my working memory rather than new problem-solving, I’ll ask AI to generate another problem that tests the same underlying idea, or switch the numbers so I am forced to focus.
This question generation strategy is particularly powerful once you have a quiz or exam back from a specific professor. At that point, you have a sense of the professor’s style, and you can ask AI to help you identify the kinds of problems they are likely to give and generate new problems or quizzes in a similar style.
For example:
“Give me three more questions testing this concept.”
“Make a quiz in the style of my previous quiz using the content from my new lecture notes.”
“Make the next problem harder.”
“Give me a problem designed around the mistake I just made.”
In sum
Ultimately, I find AI most helpful when I use it as an active study partner rather than simply asking it for answers. It helps me turn an overwhelming week into a realistic plan, organize lecture material into useful study resources, trace confusion back to its foundations, and practice the exact skills I have not yet mastered. If you use AI as a substitute for doing the work, you’ll be in trouble, but I believe there’s a way to use AI to help make that work much more structured, targeted, and manageable.
You can also reach us directly at success@admitmd.com, or call or text 512-693-9228.
Comments