Prashant Gupta, a mathematics educator at The Cambria International School in Kalyan, Maharashtra, talks about his experience using AI in the classroom.
It has become almost an essential fact that educators should start teaching AI in their classrooms. The question is how to use AI in teaching. Well, now let us understand what AI is. We used to work a lot in earlier times when AI was not there; we had to complete the work in a specific time frame. But due to AI, it has been cut short. What has changed? For example, a work that takes at least 2 hours is done in 5 minutes or even some seconds. We just need to type a prompt that we have to get a certain work done using AI, to empower educators to improve the education system. So it has been made in one or two, empowering educators using AI to make learning easier and more engaging. Teachers in the USA and UK report that AI can significantly improve student engagement and participation. The retention rate of students has increased. Earlier, students used to bunk classes, but there is 40% higher knowledge retention when using AI-powered visual learning tools and 30% better understanding because we need to understand more via visuals. Improvement in student comprehension is reported by Indian universities using AI labs also.
How do prompts work?
When using ChatGPT, we just write a prompt. For example, I am a Maths educator. I need 20 questions on trigonometry. So I write, I need 20 MCQs on the topic trigonometric ratios and identities, such that 30% are easy, 40% are medium, and the remaining 30% are tough questions. So it makes 20 questions, so that 30% are easy, 40% are medium, and 30% are hard, and I do not have to work a lot. What else can ChatGPT do? ChatGPT writes your slides also. We are educators, so we have to teach using PPT. So we do not have to write every slide. What we have to do is just give a prompt, and it will give the whole content. For example, I am teaching indefinite integration. So indefinite integration needs some theory part also. So I will not write everything on the slide. What I do is I just write a prompt on ChatGPT, and after that, I get content. And then I copy it and paste it into Markdown, and then I get the Word file. After that, when I copy and paste it onto my slides, I have completed my class. I just have to move the slides and solve the questions. So ChatGPT and Massive Markdown have made my work and other educators’ work very easy.
How does AI help?
ChatGPT serves as an on-demand teaching partner, available 24/7. It is not like after 12 a.m., it will not give you content. It will give it. So, we need to create differentiated content for diverse learning abilities. In my classroom, there are, for example, 60 students. Out of that, 50 are below average, and 10 are above average and toppers. So I have to teach all 60, not only above-average students. So what I do is I explain the slides to all students, but for the rest of the 50, I give them extra questions. If extra questions are given and if they are not solved by any student, I just show the solution. If they see it and then solve it and get the correct answer.
Making history come alive with AI
This can be used by a history teacher also. For example, a history teacher wants to teach a lesson. Earlier, there were methods like we used to teach NCERT or other reference books, and it was quite boring. But history can be made very interesting. We have Leonardo.ai. What we have to do is, firstly, we will go to ChatGPT and paste the text from the snipping tool. Using the snipping tool, I will cut text from NCERT and then paste it into ChatGPT. I will write a prompt that I need 20 prompts to paste into Leonardo.ai so that I can generate 20 images to teach my students a visually perfect history chapter. For example, Indian independence or the French Revolution. Then it will generate 20 new prompts for Leonardo.ai. After that, I will open Leonardo.ai and paste each prompt one by one. Then I will write “hyper realistic”. Now my image library is created with 20 images. After the images are created, I will go to each image and set the image motion between 3 and 5 so that my image becomes a short video. Then I will download all 20 short videos. After that, I will go to ElevenLabs or Clip Channel. I need a voice-over also. So, the text that I have snipped from NCERT, I will paste it into ElevenLabs and select the trending voice-over. Then I will download the audio file. After all this is done, I will go to CapCut and paste all the audio files as well as my videos. Then I will go to a YouTube video to get background music for the History chapter. Once I get all these things and edit the videos, I will have created a very good movie-style history chapter within an hour. So it is that easy. Gone are the days when teachers used to teach: one day there was Subhas Chandra Bose, he was fighting, and he said this and that. Now trends are changing, styles are changing, and students are changing. So we educators also have to adapt AI in our teaching style.
Now what about Chemistry?
In Chemistry, we have to use some tools like MolView, ChemSketch, Avogadro, and MoleStar. How to use them to build and rotate molecules, visualise orbitals, bond length and hybridisation. Then come Chemis.org, Mechanism app, MoleView, and Reaction Mode. It will show electron flow, resonance, substitution or elimination step by step. pTable.com, Desmos Chemistry Graphs display real-time atomic radius, electronegativity and ionisation energy trends. PhET Simulation, ChemCollective Virtual Labs demonstrate UV, visible, IR and NMR principles interactively. If we use even 20% of these tools, our chemistry class will become more interactive and more interesting. Generally, we see that chemistry is quite a boring subject. But if we use AI and try to visualise trends, then students will engage more and be able to learn more chemistry.
To visualise molecular structure, 3D AI models show atomic bonds, electron configuration and molecular interaction in stunning detail. Simulate chemical reactions. We can watch reactions unfold safely in virtual labs. Observe colour changes, energy transfers and product formation. Interactive practice tools like Labster provide AI-generated experiments where students can adjust variables and see immediate results. Interactive practice is most important in chemistry. For example, a student doing salt analysis can check what salt will be formed after the reactions. It becomes more interactive.
AI supercharges mathematics
Now, my subject is Mathematics. For example, I have to make slides. I snip content from teaching notes and paste it into ChatGPT or Claude to get content. Then I paste it into Markdown to get a Word file, and then paste it into PPT slides. I also get questions from ChatGPT. But it is not enough. Step-by-step guidance provides detailed explanations, breaking complex problems into manageable steps. I write on every tough question in Claude that I need to understand like a 10-year-old child, so that even JEE Advanced problems become easy for students. Feedback offers immediate correction and alternative approaches. Sometimes ChatGPT gives wrong answers, so I ask it to correct itself. It admits and corrects. Mastery achievement adapts difficulty based on progress until concepts are mastered. Khan Academy integrates AI to personalise learning paths, resulting in a 20% boost in test scores. We have to learn to solve JEE Mains and JEE Advanced Questions. So, I snip the entire question, paste it on Claude, and I give the prompt Explain the tough question using step-by-step as we would do with a 10-year-old kid. I get the explanation for JEE Advanced and JEE Main questions. If I don’t understand, I rewrite it, get it explained, and once I am clear about the whole thing, I copy it onto my slide. This way, a weak student can feel comfortable. Or I create slides using ChatGPT, paste the content on Massive Multum so that I can have all the slides covered. AI has a major role. If a student is weak in math, when I want to give a test on Set Relations and Functions and Trigonometry, AI understands the student, and it starts giving easy questions. The student can score more in this way and gain confidence. In the next set of questions, there will be 80% easy and 20% moderate questions. In the third test, it will be 60% easy and 40% moderate. In this way, the student can increase the percentage of marks he or she scores. Math will need human intervention first of all, and AI can help the educators. It cannot replace humans. It may commit errors that can be corrected only by humans. So we can use AI to improve our teaching style and make the class more engaging and interactive.
Smarter teaching, faster
AI dramatically reduces lesson planning time whilst improving content quality, allowing educators to focus on meaningful student interaction, generate complete lesson plans, and export in Markdown format. They can create quick snapshots and share AI-generated material instantly. Educators report that it is reducing preparation time by 50% while creating more engaging content. Till now, countries like Canada have achieved 15% reduction in dropout rates through AI-driven platforms. The UK has a 25% increase in History exam scores. India has 18% improvement in Chemistry pass rates. The US has a 20% boost in maths scores with Khan Academy AI integration.
Is teaching transforming with AI?
So what we have to do is to stop making slides ourselves. We need to increase visually generated AI content so students are more engaged, and use tools like CapCut, Leonardo.ai and 11 Labs. Practical tips are to begin with free tools like Claude, ChatGPT, CapCut and Leonardo.ai. We need to provide teacher training, start small and scale up. We can collaborate with tech partners. Language is no longer a barrier—content is available in Hindi, Kannada, English, Tamil, Malayalam and more. The future of teaching is AI-enhanced with personalised learning, enhanced engagement, and global accessibility. AI makes quality education accessible worldwide, breaking barriers of geography, language and resources. The fusion of human creativity with technological intelligence creates unprecedented opportunities to inspire, educate and empower the next generation.
Can AI redefine teaching?
Yes, AI reduces the workload of educators and makes the classroom more engaging and interesting. Students will not want to miss the classes. Teachers can use AI to improve their LinkedIn profiles. They can make their resume more impressive using Claude or ChatGPT. It can help us locate vacancies in various places, coaching classes or educational institutions we are looking for, etc. It can also help us understand the salary structure. To cite an example, I copied the text in History, French Revolution, from the NCERT textbook, snipped it and pasted it on ChatGPT with a prompt that I needed 20 images and that I had to paste it on Leonardo.ai to generate an AI video for this. I got 20 prompts to paste on Leonardo.ai, got images, and I gave the command “Hyper Realistic” so that it looked cinematic. In my Image library, I set the images to motion to have 4 in each image to become a short video. I downloaded the videos, opened ElevenLabs, and pasted the whole script for voiceover. I downloaded the voiceover and created background music from YouTube. I put everything on CapCut, edited, and my history movie was ready.
AI helps not only the weak students but also the top ones. When we keep giving prompts till we get a convincing explanation, it will explain it as it will for a small kid. It will give the answers in a simple way. AI is used to analyse the students’ performance. Then it is used for interactive sessions. And AI can be used to explain or simulate some of the topics clearly.
Can AI go wrong?
Of course, there are many drawbacks to using AI in education, the main ones being students getting details from AI or ChatGPT and submitting the answers, which amounts to cheating. But we should learn to guide the students to use it in a positive way. AI can also make mistakes, as it clearly claims. There are other benefits also. There are a few challenges, also. When I paste the snapshots into the mathematical tools like D/DX, it does not understand things like some concepts. So I have to write it again. There are some tools, like Sigma, that do not understand the prompts, so we need to give them again and again. There are a lot of limitations for AI, too.
Contact details
Prashant Gupta,
Mathematics educator, The Cambria International School, Kalyan, Maharashtra
M: 7523862113
E: prashantvns68@gmail.com