I tried ChatGPT's new interactive quizzes on 5 subjects I thought I knew well — it quickly found the gaps in my knowledge
ChatGPT is surprisingly good at revealing your ignorance
ChatGPT has a new interactive quiz feature, designed to make testing yourself a more active part of using the chatbot.
Instead of asking ChatGPT to explain something and then nodding along because everything looks familiar, you can have it put you on the spot. It's part of OpenAI's steadily expanding learning toolkit, building on Study Mode and other features.
I picked five subjects where I felt reasonably confident, but it turned out to be a very efficient way to locate the holes in my knowledge.
How to use the quiz feature
The quiz feature is available to anyone. You just need to prompt ChatGPT by asking for it. I decided to start with a quiz about AI itself, specifically its history. But I didn't want to just have very basic questions, so I made sure to ask for the quiz to be at a certain level. And while you can ask for a short essay answer format, I opted for the classic multiple-choice style, and for help with where I went wrong.
My exact prompt asked ChatGPT to: “Quiz me on the history of AI at a college level with 10 multiple-choice questions. Don’t show me the answer until I respond. Explain anything I get wrong and give me my score at the end.”
The one-question-at-a-time format made a bigger difference than I expected. A conventional online quiz often gives you a page full of questions, followed by a score and perhaps a few explanations once you are finished. Here, ChatGPT reacted after every answer, briefly explained the result, and moved on.
I'll admit I was surprised at how difficult it became, but despite some errors, I did reasonably well, and actually learned some new facts.
Sign up for breaking news, reviews, opinion, top tech deals, and more.
I decided astronomy should be next. I have spent plenty of time reading about space, looking through telescopes, and collecting facts about stars, planets, and black holes, so I felt pretty safe. Unfortunately it felt like the AI took me telling it that ordinary solar system trivia was too easy after a few questions as some kind of personal insult.
“Make the remaining questions harder,” I told it, and ChatGPT took this to heart, with me facing questions on orbital mechanics and astronomical measurements that I barely got half right. Still, it's good to know that you can adjust the difficulty level on the fly.
That flexibility is something a fixed quiz cannot really offer. I could even tell ChatGPT exactly what was making things too easy, such as obvious wrong answers, and ask it to make all four choices plausible.
I decided I deserved something easier after that and chose 1990s pop culture. I grew up in the decade and figured it would be a breeze even if I asked for a high level of difficulty. I asked it to mix television, movies, music, video games, and technology questions, but to make it a little harder by "including questions that require more than just remembering titles. I managed a perfect score, despite not remembering exactly the composition of certain boy bands of the late 1990s and having to guess the right answer.
AI study buddy
I went for a more highbrow form of entertainment next with a quiz about Shakespeare, specifically asking ChatGPT to "avoid extremely famous quotes and obvious questions about plays like identifying the leads of Romeo and Juliet."
The questions about secondary characters and plot details made me glad it was multiple choice. If I were taking a real class on Shakespeare, I'd tell ChatGPT to note where I was failing and come back for further questions on it. I could even upload my notes to make sure I was being tested on what the class is covering.
My final category was world geography. At the suggestion of ChatGPT, I adjusted my prompt to make the quiz more dynamic in difficulty. I told the AI to "increase the difficulty of the questions every time I get three answers in a row correct."
If I was worried about getting cocky, that addition kept me very humble. The first six questions I breezed through, but enhancing the difficulty twice left me lost hunting for obscure mountains and the location of deep-sea trenches.
After I finished, I asked ChatGPT to analyze my answers, and the AI claimed it could distinguish between a silly mistake and a repeated gap in my knowledge. If I repeatedly struggled with a particular period, concept or kind of question, it offered to do a new round of questions focusing on that.
One crucial fact to remember is that ChatGPT can still get things wrong and hallucinate answers even in a quiz. If an answer seems off, you should check it independently. For more casual learning, the AI did quite well. You still have to do the studying yourself, but if you run out of practice questions, ChatGPT can help you keep studying ahead of any classroom quiz.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
➡️ Read our full guide to the best business laptops
1. Best overall:
Dell 14 Premium
2. Best on a budget:
Acer Aspire 5
3. Best MacBook:
Apple MacBook Pro 14-inch (M4)

Eric Hal Schwartz is a freelance writer for TechRadar with more than 15 years of experience covering the intersection of the world and technology. For the last five years, he served as head writer for Voicebot.ai and was on the leading edge of reporting on generative AI and large language models. He's since become an expert on the products of generative AI models, such as OpenAI’s ChatGPT, Anthropic’s Claude, Google Gemini, and every other synthetic media tool. His experience runs the gamut of media, including print, digital, broadcast, and live events. Now, he's continuing to tell the stories people want and need to hear about the rapidly evolving AI space and its impact on their lives. Eric is based in New York City.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.