LLMs are, at their core, sophisticated wordsmiths. Anything that involves manipulating, synthesizing, or restructuring words is exactly where they excel. That is why a tool like NotebookLM has become so insanely popular; you give it a wall of text, and you ask it to make sense of that text for you. The LLM already has the material in its "brain," and all it needs to do is what it does best -- reshaping that material until it gives you exactly what you need to understand the topic.
I've spent a lot of time studying with ChatGPT, and it was a total game-changer for my workflow. But recently, I've started slowly switching over to Gemini. There is one specific aspect of the Gemini ecosystem that makes it objectively superior for heavy-duty studying and the kind of spaced repetition you need when you're staring down a massive exam.
Gemini quizzes are brilliant Interactive feedback is the game-changer
Making quizzes with LLMs is a trick that every student with a pulse knows by now. It's been an option since AI went mainstream. However, a while back, I was pleasantly surprised by ChatGPT's hidden quiz feature that generates quizzes that weren't just plain text blocks. You could actually click an answer, get immediate feedback, and see a score at the end. It felt like a real app.
But when I asked Google's Gemini to "quiz me," I was even more surprised by how integrated the experience was. Gemini's quiz module is much more comprehensive than the ChatGPT version I was used to. Here's the prompt I used in this case:
"This is last year's PharmD final year exam. It's 200 questions from different categories: pharmaceuticals, pharmacology & toxicology, pharmacognosy, medicinal chemistry, and pharmacotherapy. I want you to study it, understand the type of questions, and then make a quiz containing 15 questions from this PDF."
Sure enough, I get a quiz of 15 questions from the PDF I uploaded, but each question also has a tiny hint underneath. What's more, when I pick the answer, I get feedback from Gemini as to why that choice is correct or wrong.
The best part, though, happens when you finish. A completed quiz is your gateway to even more material. You get the option to ask Gemini to generate flashcards based specifically on the questions you got wrong. You can ask for "more questions" to get another 15-question set instantly. But my absolute favorite part is the "analyze performance" prompt. It gives you a high-level overview of your strengths and weaknesses across the different categories. If I'm acing toxicology but failing medicinal chemistry, Gemini tells me exactly where to pivot my focus for the next round.
The hints are usually too much of a hint and give away the answer. I try to avoid clicking them! The feedback from a wrong choice is usually more constructive.
The simple automation hack for studying Let the bot chase you instead of the other way around
This is all great in theory, but I can be super lazy and incredibly forgetful. Even though this routine has made studying almost frictionless, I still might not remember to open the Gemini app every night and manually ask for a quiz. Life gets in the way. Much like ChatGPT, Gemini has the ability to schedule and automate tasks. So, I took it a step further: I asked it to "Send me a similar quiz with new questions every night at 8 PM."
Gemini schedules a task, and that's it. Every night at 8 PM, I get a message in that conversation with a fresh quiz waiting for me. It turns the study session from a chore I have to remember into a notification I just have to click.
A few things to watch out for: the scheduling action isn't always flawless. It's not actually scheduled unless you see the specific "scheduled task" box in the response -- it has a little toggle and a clock icon. If Gemini just tells you "Okay, I'll do that," but doesn't show you that UI element, it didn't work. Also, don't be too polite. Don't ask it if it can schedule a task. Just tell it to do it, and it usually complies without any fuss.
I've used this single conversation thread for about a month now, and a problem eventually arises: the context window. Once the conversation gets too long, you won't get an explicit error, but the quality of the quizzes will steeply decline. You'll start getting repeat questions, or the logic in the feedback will start to get fuzzy. When that happens, I suggest you disable the task and just start a brand-new conversation. Fresh context equals better quizzes.
What we mean by studying with Gemini It's not just about getting quick answers
A quick clarification: when I talk about studying with Gemini (or any other LLM), I am not saying that the AI is doing the learning for me. I'm not asking it to write my assignments or summarize chapters so I can skip reading them. That's basically useless. If you don't do the mental heavy lifting, you won't remember the info when the screen is off. Instead, what Gemini does is take the friction out of the study process.
I've got the source material, and I've got my personal notes, but everyone knows it would be ten times more efficient to study those notes using flashcards or active recall. The problem is that making hundreds of high-quality flashcards is a soul-crushing chore. Most students spend more time making the cards than actually studying them. Gemini flips that script. It can churn out a deck of targeted flashcards in seconds.
Unlock LLM study upgrades -- subscribe to our newsletter Subscribing to the newsletter gives you hands-on guides, tested prompt blueprints, quiz and flashcard workflows, and analysis techniques for studying with LLMs -- practical steps to turn AI tools into effective study habits. Subscribe By subscribing, you agree to receive newsletter and marketing emails, and accept Valnet's Terms of Use and Privacy Policy. You can unsubscribe anytime.
So you see, Gemini builds the tools, but I'm still the one studying them. This logic applies to any supplemental material -- quizzes, summaries, or categorization lists. At the end of the day, they are just letters and words, and an LLM like Gemini is the ultimate machine for generating them.
Related I started using NotebookLM with ChatGPT, and the results were incredible
This NotebookLM and ChatGPT combo transformed my research, learning, and writing process
Posts 1 By Saikat Basu Scaling your knowledge with AI
My perspective on LLMs is that they amplify the capacities you already had. If you were already a stellar writer, you can use it to write better. But if you weren't, then it's not gonna produce what a stellar writer would. The same thing applies to studying. I was already a student before Gemini. I know what to study and how to study, but now with Gemini, I can make that more efficient.
LLMs are here to stay! And they are only going to get more integrated into how we learn. But as the many recent AI outages have taught us, it is always best to use these tools to build your own internal skills rather than outsourcing your brain to them entirely. Use Gemini to build the flashcards, use it to run the quizzes, and use it to analyze your gaps -- but make sure you're the one internalizing the answers. That way, if the servers ever go down, you aren't left stuck without the knowledge you actually need.
