AI for Kids

How a Teen is Making AI Education for Everyone (Middle+)

Amber Ivey (AI) Season 2 Episode 24

Send us a text

Kaashvi Mittal, a rising high school senior, is transforming how we think about AI education. As founder of Together We AI, she's breaking down complex technology into bite-sized, understandable concepts through free workshops that make artificial intelligence accessible to everyone.

The spark for Kaashvi's journey came during a Stanford AI for All summer program where she helped create an AI model analyzing satellite imagery to determine socioeconomic status of different regions. 

For parents and educators worried about children becoming overly dependent on AI, Kaashvi offers practical wisdom: establish clear boundaries between helpful uses (generating practice problems) versus harmful applications (writing essays). She emphasizes developing core skills first, using AI as a supportive tool rather than a replacement for critical thinking.

Resources Mentioned in This Episode

Together We AI
Sign up for Kaashvi Mittal’s free, beginner-friendly AI workshops: www.togetherweai.org
Instagram: @TogetherWeAI

Stanford AI4ALL Program
A summer program that introduces high school students to AI and its applications: ai-4-all.org

Learning Platforms for Kids:

  • Code.org – Great for learning the basics of computer science and AI
    code.org
  • Khan Academy – Offers beginner-friendly videos about AI and computer science
    khanacademy.org

Image Generators Mentioned:

  • DALL·E – An AI model that creates images from text prompts
    openai.com/dall-e
  • Midjourney – A generative AI platform for high-quality artistic images
    midjourney.com

Support the show

Help us become the #1 podcast for AI for Kids.

Buy our new book "Let Kids Be Kids, Not Robots!: Embracing Childhood in an Age of AI"

Social Media & Contact:

Listen, rate, and subscribe! Stay updated with our latest episodes by subscribing to AI for Kids on your favorite podcast platform.

Like our content, subscribe or feel free to donate to our Patreon here: patreon.com/AiDigiTales...

Amber Ivey:

Welcome to the AI for Kids podcast, the podcast for moms, aunties and teachers who want the kids they love to understand AI without more screen time. We keep it simple, safe and fun. No tech degree required. Each episode breaks down AI ideas and includes activities to help kids use AI in ways that keep them curious and creative. No pressure, no overwhelm, no extra screens, just clear, engaging learning you can feel good about. Let's get started. Hi everyone. Welcome back to AI for Kids.

Amber Ivey:

Today we have someone who's making sure that everyone, no matter their background, gets to learn about AI in a way that actually makes sense. Meet Kashmi Mittal, the founder of Together we AI. She's a high school junior yes, she's still in high school who runs free AI workshops breaking down AI into easy to understand lessons. She's even worked on AI research at Stanford and UC Santa Cruz, helping solve some of the big questions about AI and bias. Today we're going to talk about why AI education matters, how she got into AI and the real risks and cool possibilities of AI in the world today. But first let's get started with a fun game. But before we even do that, I want you to say hi to everyone. Kashmi, hi, kashmi.

Kaashvi Mittal:

Hi, it's so great to be on this podcast.

Amber Ivey:

I am so excited to have you here. I know I bothered you a few months ago, but I'm glad we were able to get you on and I know the audience is going to be just as excited. So, to kick off, we're going to do an icebreaker called AI Mythbusters. So I'm going to say a statement about AI and you have to tell me if it's true or false and then explain why. Are you ready? Yeah, I'm ready. Okay, great, ai can think and feel just like humans. What do you think?

Kaashvi Mittal:

That's a tricky question. I think AI is getting pretty close to humans, but I would have to say false. Humans have emotions and individual thoughts which AI can't replicate in the same way.

Amber Ivey:

Ding, ding, ding ding. You are right. We both agree today. That is false. We have no idea what's going to happen in the future with AGI and all the other stuff, but great job there. Next one If you use AI too much, it can stop you from being creative.

Kaashvi Mittal:

I'd say it depends on how you use it, but I had to go with an answer. I would say true, if you use AI to create artwork instead of making it yourself, or write stories instead of writing them yourself, you might stop using your own creativity.

Amber Ivey:

That is such a good point. Someone shared with me a study from MIT where it's actually showing that folks using like generative AI for writing they're actually losing some of those skills. So that's a good thing to think about as we think about integrating AI. But I also agree with you. I think it depends right If you're using it as an assistant, but not like leaning fully on it. I think that's very different than like fully writing everything with AI or using it to make all of your art, like you said, yeah, definitely. Or using it to make all of your art, like you said, yeah, definitely. And then AI can help doctors detect diseases faster than humans in some cases.

Kaashvi Mittal:

That's true. Ai is such a powerful tool in the medical industry because of how fast it can process information and make predictions, and I think that in the years to come, we're going to see it being used in even more ways in the medical field, so I'm really excited about that.

Amber Ivey:

Same here. I'm really excited about that as well. All right, this one's an interesting one. Ai always tells the truth. What do you think that one?

Kaashvi Mittal:

is definitely false. Oh yeah, ai is not perfect and it can make a lot of mistakes, so it's super important to be aware of the fact that AI doesn't always give the right information and fact check.

Amber Ivey:

I love that. And last but not least, you need to be a coding expert to understand AI.

Kaashvi Mittal:

That one is false. Ai may seem super challenging at first, but when you break it down, anyone can understand it.

Amber Ivey:

I love that. So let's go into how you got involved with this whole AI and teaching thing. Like you're in high school right now, you're already leading AI workshops, which is mind blowing to me. How did you get into AI in the first place?

Kaashvi Mittal:

Yeah. So after my freshman year of high school, I participated in a summer program called Stanford AI for All. Stanford AI for All was an amazing opportunity for me to collaborate with others by coding an AI project. So our AI project was creating an AI model that could determine the socioeconomic status of a region by analyzing satellite imagery. And after that project I became really, really fascinated by a bunch of other ways of how we could apply AI, and that's really why I started to kind of delve deeper into how we are using AI in today's world and how we can use AI in the future.

Amber Ivey:

That is such an interesting project. So to ask you a little bit further, like what were you able to see? Like how could you tell one area over another had some of the issues you were trying to identify based on those scans?

Kaashvi Mittal:

models can see images or videos and extract information from that data. And what we were able to determine was we were able to look at different elements of the images. For example, if certain satellite images showed more buildings and less trees, that was usually an indicator of the fact that that region might have a higher socioeconomic status than another region. So we basically experimented with a bunch of different features, looking for a bunch of different characteristics, and doing that we were able to actually find some really great results.

Amber Ivey:

That is so cool and, honestly, something I've never even thought about. I do a lot of these interviews with different folks, but that's just like a really cool use of AI. You've also done some other cool things this whole idea of like AI research at Stanford and UC Santa Cruz. What's something surprising you learned from those experiences?

Kaashvi Mittal:

Yeah. So I think the most surprising thing I've learned was from researching AI bias at UC Santa Cruz, and it was really how much bias exists in today's world. I think that in the past, like a few years ago, we used to talk a lot about AI bias, especially when models like ChatGPT and DALI were released to the public, but since then, the conversation about the bias in these models has kind of died down. So a lot of people assume that these big companies have addressed the issue of bias. But when I started to research bias at UC Santa Cruz and saw the bias firsthand in those generated AI images, it was really shocking and it made me really aware of the fact that there's still a lot of bias in AI models today, and that's something that we all need to work towards.

Amber Ivey:

Can you give us an example of one of the things you saw in the images?

Kaashvi Mittal:

Yeah, so one of the types of bias we were looking at was occupation-based bias, which is basically looking at how the images generated of certain occupations were portraying more of one type of gender or more of one type of race, so we saw different types of bias. For example, most of the images generated of computer scientists were male instead of female, or most of the images generated of CEOs were not people of color. So those were some of the indicators of the fact that there was a lot of bias in these images.

Amber Ivey:

That's so interesting because when you think about, like, how computers are trained, if they're being fed certain images of only one group of people, then it comes out on the other end of that group of people all the time, comes out on the other end of that group of people all the time, even though there are other people with different genders or races who also are in those careers. That's really interesting. And what made you start Together we AI, Like what was your big mission with it?

Kaashvi Mittal:

Yeah, so before I joined Stanford AI for All yeah, so before I joined Stanford AI for All, I was actually feeling really overwhelmed with the idea of learning AI. I had a lot of doubts like, will I actually be able to learn AI or will people know more than me? And after I learned about AI from the program, I realized that many people were also wary of learning about AI because they felt like it was this unapproachable and complex topic, and I wanted to make AI education accessible and easy to understand for everyone, which is why I started leading free AI workshops through Together we AI.

Amber Ivey:

What's interesting is that some of us, like I've known about AI for some time and I know it's like old enough to be a grandparent to the kids who are listening here today, or a great grandparent, depending on how young the kids are. But when OpenAI came out with their model, everyone's like what are we going to do? What's going to happen? And I love that you actually started offering those free courses. What are some of the I guess the most interesting things you've seen through teaching others how to get over that for themselves, as you've gotten over it through launching a whole program?

Kaashvi Mittal:

Yeah, so I think the biggest change that I've seen is how the students are at the beginning of my workshops versus how they kind of leave the workshops In the beginning of my workshops. Versus how they kind of leave the workshops In the beginning, most of the students are kind of closed off. They're not really sure what to say, they don't really raise their hands or want to participate. But that's why, through my workshops, I try to really break things down in a simple and easy way for them to understand and also bring up a lot of interactive activities where they can really just talk amongst themselves but also talk with each other. It's a really good way for them to kind of become more comfortable with talking about AI and also learning about AI. And so by the end of the workshop I noticed that they are a lot more involved, they are a lot more excited about the possibilities of AI and I know that they're going to look into AI even more in the future.

Amber Ivey:

I absolutely love that. I want to redirect us for a second before we keep digging more into what you're doing and how it can help kids. I want to talk about some of the things around AI, like the good, the bad, the weird, right. Ai is being used in everything from self-driving cars to TikTok algorithms. What's one AI application that you think is super cool?

Kaashvi Mittal:

One AI application that you think is super cool. One AI application that I think is super cool is AI image generators. It's still crazy to me how you can type in prompts with whatever you want and an AI model will create an image in seconds. I remember the first time I experimented with Dali, I was thinking of like the weirdest thing that I could put into the AI model to see if it would actually work. So I just for some reason thought, why don't I see if it can generate an image of an elephant on the beach wearing a tutu and a tiara?

Amber Ivey:

And I wrote down or I typed in that prompt and within seconds Dolly gave an image back and I just thought it was so cool that AI was able to create something that was in my imagination yeah, I'm still blown away, especially some of the more recent like updates to Dolly and Mid Journey where, like I can take, like I just did this this week I'm going to the Beyonce concert coming up and I'm still haven't found my outfit and I took a picture of a shirt I really liked and I was like, style this with. I went and do a red, white and blue theme because it's going to be on July 4th. Style this with this, and before it wouldn't take that image and be able to pull things in. Now it took the exact image of the shirt and then brought in the boots, it brought in the accessories, it brought it all in on that image and it's getting so much better.

Amber Ivey:

So, even like, the creative things are wild, but even just taking current images and like being able to manipulate them is scary, because no one would have known that didn't come from a magazine, right, it looks like the same exact picture. So very interesting stuff happening here and I'm glad that you brought that up. Ai can also have big problems, like bias, as you said earlier, and misinformation, or even lead to cheating. What's one AI issue that kids and teens should really be aware of?

Kaashvi Mittal:

Over-relying on AI. I think that it's so easily accessible to us nowadays, with it just kind of being at our fingertips, and it's so easy to think that, oh, I should just use AI to do this work for me or this work for me. But because of that, there's also this like worry, like what if you don't develop your own skills? And I feel like for kids and teens, it's important for them to also, you know, have those critical thinking skills, be able to develop their own skills before always just jumping to AI for everything. So I think it's always important to just think about how much you're using AI and make sure that you're also able to kind of just use your own skills as well.

Amber Ivey:

Yeah, I agree with you. That's the thing that, no matter who I talk to even kids that I speak to, especially teachers and parents they're concerned about kids losing their creativity or never gaining skills and it's like compare AI to a calculator. But I think it's way more advanced than that. But back in the day people used to like have to write everything out.

Amber Ivey:

I remember my calculus teacher when I was in high school made us like not use a TI dash, fill in the blank I'm not going to say what number it was because I'll date myself. I'm not using TI calculator because she wanted us to learn how to actually do calculus. So we would have to do like one problem was like four or five pages of writing it out and then she let us use the calculator. So some of that like I think calculators are helpful because now I'm not in an industry where I need to do calculus and things like that, but at least I had the foundational knowledge. So I am concerned about kids losing a bit of that. Are there any recommendations further than what you just shared that kids or parents and even teachers need to think about to help kids not rely solely on AI?

Kaashvi Mittal:

Yeah, I think that just really emphasizing the fact that in the future, when you're working these skills, that like these soft skills, and also the skills that you develop, like these critical thinking skills, you won't always have access to AI like in every single situation and I think, knowing that in your mind, it's important to be able to build your own skills and just move past this reliance on AI and I think also just giving kind of instruction on maybe like when to use AI versus when not to use AI, especially when it comes to education, I think there's like a really clear boundary between when you should be using AI. If you're using it to help you study for an upcoming exam by like generating practice problems, I think that that is a really good way to use AI to support your learning. But then there are also ways that you can use AI that harm your learning, like using AI to write essays. That really takes away from your creativity and your own work.

Amber Ivey:

Yeah, you hit so many points on the head and I agree with you a hundred percent. The other piece that I'm concerned about I want to talk about a bit. It's like deep fakes are getting really realistic and a lot of people I know on my socials are saying like, oh, someone took my image and now they're trying to sell a product saying it's me and all these things are happening. If someone comes across a video that looks too good to be true, how should they figure out if it's real or fake when these things are getting so hard to tell?

Kaashvi Mittal:

Yeah, deep fakes are becoming really really realistic, and I think there are a few things that you should look out for when you're trying to determine if something is a deep fake. One would be unnatural body movements and especially having to do with the eyes. I've noticed that with a lot of deep fakes, there are kind of like these inconsistencies with the eyes, like, for example, people aren't blinking consistently or they're looking in a different direction than where they should be looking, and so if you notice that, like the body movements just seem a little unnatural, I think it's important to kind of just fact check and think like maybe this isn't real. The second thing would be the texture of skin. A lot of times, like in deep fakes, people's skin look very like overly airbrushed and smooth, but that's not how our skin is in real life.

Kaashvi Mittal:

And then the third one I think this one is like the biggest telltale sign of a deep fake is text. If you see text as very like wonky and weird and it's not really making sense, it's because of the fact that AI models view text as another part of the image rather than being like a logical text message. So because of that, it's really weird when you see text in AI image generated videos or images. They're usually not actually like proper words or coherent words, so I think those are a few of the different ways that you can tell if something is a deep fake or not.

Amber Ivey:

Good points. Kids, make sure you write that down, parents, you too, because those are really good points. What's one ethical AI issue? You think isn't talked about enough, but it should be.

Kaashvi Mittal:

Definitely AI's impact on the environment. When we use AI models, we often don't think about what's happening behind the scenes, which is that these AI models are using enormous amounts of energy and resources to function. Using enormous amounts of energy and resources to function. In fact, I recently read this article that stated that the average chat GPT conversation uses about one plastic bottle's worth of water. So I think it's super important to be mindful of how much we use AI, because it does have a really big environmental impact.

Amber Ivey:

Thank you for mentioning that. That's something I've been talking with folks about, and it's because of all the different data centers and what it takes to run these high power models that there's just a lot of energy going into them. I know there's articles out there that talk about if you use O3 on GBT versus 40, what that looks like and the difference, but a lot of the reasoning models or models that do more deep thinking do use even more water. So that's something for us to think about what should go in Google versus what goes into some of these more advanced models and how we use them responsibly. I love that and I agree we should continue that conversation. So I want to take us back to the school and back to some of the work you're doing. A lot of kids might think AI sounds cool and all, but I don't know where to start. What's the best way to begin learning about AI? Kids use of courses went up over the last few years substantially, but a lot of kids still are getting started.

Kaashvi Mittal:

I think the best way to begin learning about AI is to really just start with the basics of what AI is and how it's being used in your day-to-day lives, and then I recommend for that watching videos from platforms like Codeorg or Khan Academy, because they do a really great job of breaking down complex topics. And then, once you've understood the basics of AI, I think it's really good idea to just start experimenting and trying out different AI tools yourself. So one of the things I frequently did when I began learning about AI was I would find an AI tool online, experiment with it a little bit and then research how the AI model actually worked, and so that way I was getting some of the hands-on experience using AI tools, but also understanding a little bit of how the AI tools actually work.

Amber Ivey:

Even how you talk about this stuff, you keep the word so simple and you're so clear on what you're trying to communicate, and that's how you teach as well. Can you give us an example or a game or analogy that you use to explain AI to younger students?

Kaashvi Mittal:

Yeah, sure, one of the applications of AI that I teach in my workshops is recommendation systems, which are basically any AI system that recommends you content like Instagram, netflix and Spotify, and I like to compare recommendation systems to your best friend. So imagine you're talking with your best friend and you ask them for something new, like a new movie to watch or a song to listen to. Your best friend knows your preferences, so they'll give you recommendations accordingly. And in the same way, ai recommendation systems are learning from the past content that you've consumed and your behaviors to suggest you new content. So, for instance, netflix will track the genres of movies you watch, your ratings of certain shows and movies and your searches to recommend you new content. So that's one of the kind of many analogies that I like to put throughout my workshops.

Amber Ivey:

Super easy to understand and I just love that example. What's something completely unexpected that AI has been used for that most people don't know about?

Kaashvi Mittal:

I recently learned that AI has been used to create personalized fragrances or perfumes for people by analyzing their responses to different questions, and I thought that was really fascinating because, like I would have never thought that AI could be used for something like that, but I think it's really cool if it's something that gets even more popularized.

Amber Ivey:

That's just making me think about all the like future personalization that's going to happen because of AI.

Amber Ivey:

I would never have thought to make a perfume based on, like, how I interact with this AI tool.

Amber Ivey:

So, for cool, and for those kids who are like, oh yeah, you know how to break things down and you know how to explain it, but it's too techie for me, there's a lot of kids I go into schools and I talk about AI all the time and I always ask, like, do you have a future career that will not be touched by AI or STEM? Kids always raise their hand because they're like sports aren't touched by AI, music isn't going to be touched by AI, cooking isn't going to be touched by AI. And we always get to the end point of like, literally everything's going to be touched by AI or STEM in some way and if you use it right, you actually are a better athlete. If you use these tools, you're a better cook. If you use these tools, you'll be a better fill in a blank if we have AI attached to it. So for the kids who might be nervous that AI is too techie or complicated and not for them, what would you say to someone who feels intimidated by AI?

Kaashvi Mittal:

I would say to just take it one step at a time. It can feel really overwhelming because there's so much information online about AI and people will commonly throw around these big words like neural networks and machine learning. But if you start with the basics, like understanding the definition of what AI actually is and how it's being used in your day-to-day life, you'll be able to slowly move on to learn even the most intricate topics.

Amber Ivey:

I like that. Start with the basics and you can build from there, and that's like everything in life. Right? We don't go from walking to immediately riding a bike, like we start with training wheels, then we take the training wheels off. I wish everyone would think about AI in that way that we're all learning. We're all learning how to use this newer or publicly available version of AI that hasn't been known to man just yet, but it takes time and there are resources and people like you are going to help them figure it out. Kashmi, I'm giving you an AI time machine and I want you to answer these questions rapid fire. Are you ready? Yeah, I'm ready. Okay, you go 10 years into the future. What's one way AI has changed the world 10 years in?

Kaashvi Mittal:

the future, self-driving cars will be everywhere.

Amber Ivey:

I agree. Okay, you go 50 years into the past. What's one way AI would have made life easier for people back then?

Kaashvi Mittal:

I think 50 years ago, access to mental health support was a major issue, so AI powered mental health chatbots could have made this support so much more accessible for thousands of people.

Amber Ivey:

Oh man, that's a good one. And then you go a hundred years into the future. Skip over that. 10 years pass those self-driving cars. What's one AI invention you hope exists a hundred years from now?

Kaashvi Mittal:

One AI invention I hope exists is personalized medicine or using AI for personalized medicine.

Amber Ivey:

I agree, I want that like 10 years from now. And then last one if you could bring one historical figure to the present and teach them AI, who would it be?

Kaashvi Mittal:

Isaac Newton In school. I learned so much about his theories and discoveries, so it would be super cool to teach him something myself.

Amber Ivey:

I love that and it's interesting because, like all these great scientists who are no longer here now, like laid the groundwork for all the stuff that's happening. So being able to like pull them up to now so they can see like hey, that thing you were, that theory you're working on, or those concepts actually became something, that's such a cool idea. So I want to talk about the future of AI and who we should think about following as we get close to wrapping up our show. What's one upcoming AI advancement that you're most excited about, beyond personalized medicine?

Kaashvi Mittal:

And, of course, we already have self-driving cars, but self-driving cars I think that using AI for biology in general so not just personalized medicine, but using AI for biology in general is something that I'm really excited about. I think that there are a lot of different ways that we can see AI being used in that field. Ways that we can see AI being used in that field Like, for example, using AI for analyzing MRI scans something that I'm really excited about because I think that when we look at how doctors are doing it today, where they're having to manually go in, look at these images for long periods of time and determine if there's, you know, signs of tumors or any sort of illness and disease, I think that with AI and its ability to process information so quickly, we can really speed up that process and make it so much more accessible to people around the world. So I think that's one AI advancement that I'm super excited for.

Amber Ivey:

I love that In real life example. So I live with something called multiple sclerosis. It's a neurological disease, but I have to always get MRI, so beforehand I have to like and for kids who don't understand, it's just a disease that affects your brain and then it may cause issues, but they have medicine that I'm on for. That helps me through the process. But I have to get scans of my brain to see how far it's gone, and Kashmi just mentioned MRIs, which are those scans. So I got a new doctor recently and he's sending me to an MRI that has AI integrated. I haven't gone yet, but I'm really excited about that.

Amber Ivey:

I probably should do an episode on that when I get back, but before I had to take like CDs that I often lose, have to transfer stuff over, have a doctor look at those things and see if there's change, versus like just having an AI being able to have all my data and have all that and can see if things have changed. Like I am with you, I'm glad my doctor is more advanced and he's like I can't see everything. You're going to go to this place that uses AI so we can actually see and we can actually help you even more. So I love that idea and I'm glad some doctors are thinking about that. So who are some cool AI educators, YouTubers or programs that kids and teens should follow if they want to learn more about AI and dive into this space? Of course they need to follow you. Who else should they follow?

Kaashvi Mittal:

I think for me, learning from Codeorg and Khan Academy, those were great resources to learn about AI in a simple way, used to and comfortable with even the idea of learning about AI, so I would recommend those.

Amber Ivey:

And if every middle schooler or even high schooler in the world could hear one message from you about AI, what would it be?

Kaashvi Mittal:

AI is made for everyone. Don't be afraid to learn about AI. I was intimidated at first too, but if you take it step-by-step, learn the basics, experiment with AI tools, stay up to date on new advancements, you'll be able to learn artificial intelligence more and more every single day.

Amber Ivey:

Kashmi, this has been awesome. I love how you're making AI education fun and easy to understand and fill in a gap that's definitely needed. But before we go, where can people sign up to learn more about your organization or to get free AI workshops and follow Together? We AI?

Kaashvi Mittal:

People can visit wwwtogetherweaiorg to sign up for my free, beginner-friendly AI workshops, and I'm also on Instagram at Together we AI.

Amber Ivey:

Thanks so much for being here, kashvi. I can't wait to see how, together we AI keeps growing, and thank you so much for spending time and for helping us learn so much about the future of AI, thank you. Thank you for joining us as we explore the fascinating world of artificial intelligence. Don't keep this adventure to yourself. Download it, share it with your friends and let everyone else in on the fun. Subscribe wherever you get your podcasts or on YouTube. See you next time on AI for Kids.

People on this episode