Have you ever noticed how video games get harder just when you're getting good?
Or how Netflix or YouTube always suggest the perfect next show or video to watch?
That's artificial intelligence, AI, at work.
Welcome to AI for Kids, the podcast where we explore AI together with your parents and teachers.
We'll look at how it works in simple ways with real examples.
And we'll talk honestly about the worries like safety, privacy and what it means for the future, so everyone feels informed.
Are you ready to get started?
Let's discover AI together.
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.
Hey, everyone.
Welcome to AI for Kids, the podcast where we learn about artificial intelligence in a fun and safe way for kids and their parents and teachers.
Today, we're talking to Dr. Sonia Tiwari, who is a children's media researcher.
Dr. Sonia, welcome to the show.
Thank you for having me.
Exciting conversation.
Yeah.
I'm so excited to be here.
Full transparency, folks.
Dr. Sonia was supposed to be up here a long time ago, but my life got in the way.
So I'm so glad that she's on.
And I think she's on at the perfect time, especially as we think about children's media and why that matters with AI.
Dr. Sonja, can you tell us about what does it mean to be a children's media researcher?
I have no idea what that is.
Well, sickly I do, but for the audience, what is that?
Yeah, so children's media can range from all of the media surrounding children, whether it's children's books, film and television YouTube, gaming and now AI.
So broadly, I come from a background of film and animation.
I was a character designer in the gaming industry.
And then, many years later, I went back to school to get my PhD and trying to see the educational side of it, that how do we use characters to help children learn?
And so that's my job as a researcher now.
I look at whether a character is powered by AI or in a traditional book form.
How do children connect with it?
How do children learn with it?
That is so interesting.
And you made me, like a lot of stuff start going through our mind.
One, that's so cool.
You used to like design characters for games, but the thing that scares me a little bit is like now, with AI characters involved.
That's going to be a very interesting space, with characters learning about you or interacting with you in different ways.
But before we jump into this, I want to do a few icebreakers and then a quick lightning round with you to get us started.
Are you fine with that?
Sure.
Yeah.
So what's one fun thing you did this weekend or this week?
I mean, it's early in the week, but also this weekend is fine.
I attended the AI and Education Summit at Khan Lab School.
So it was like perfect right before this podcast.
There were also children panelists.
So it was very interesting to hear.
I was very inspired and also ashamed of my own childhood because I was not so put together.
So yeah, it was great to hear how kids are using AI and how educators across different fields are using AI.
The funny thing is, every time I do a podcast interview with any person under the age of 18 and they're talking about how they built some AI tool or how they're coding, I'm like what was I doing at this age?
Right.
And even the way they articulate themselves, like I don't think I was able to communicate my ideas so beautifully.
Yeah.
I know I wasn't.
So I'm glad the next generation has figured it out, because AI is here and we need them here to figure it out.
So back, speaking of kids, when you were a kid, did you have a favorite animated character?
If so, do you think AI could bring that character to life?
Okay, so my favorite character was Bubbles from the Powerpuff Girls.
And I think it can be possible, but I think it will be very cringe because I think the...
The thing about Powerpuff Girls was it was like a band of musicians, right?
Everyone has a personality and it all comes together.
So, you know, Blossom was the sincere leader type and Buttercup was the rebellious one.
And so by comparison, Bubbles was the funny, cute one.
And so without that dynamic of other characters on its own as a chat bot, it will feel weird.
I agree.
And they were just so dynamic as a trio.
Like to have them without that relationship that they built together would be super weird.
But I also love the Powerpuff Girls.
So glad we brought that up.
For kids who never heard of that, Google it.
Or ask your parents, I'm sure they loved it too.
And then we're gonna do a very quick lightning round.
So I'm gonna ask you a few quick questions and then you're gonna answer as quickly as possible.
Are you ready?
Yes.
What's your favorite piece of tech or gadget right now?
Flora AI, I just like how you can visually connect different inputs and create something with AI.
I love that.
We'll add for our parents and kids.
We'll make sure that's in the show notes so you can look at that.
And if you had to design a brand new AI character, what animal or creature would it be?
Or maybe it's not an animal or I mean, it would be a creature.
Maybe it's not an animal.
Maybe it's something else.
A tiny dog, like pocket-sized dog, maybe, because I miss taking my dog everywhere with me.
Aw, I love that.
That's super sweet.
And what's your go-to snack when you're working on a big project?
Baby carrots.
Oh, interesting.
I just love the crunchy sound.
Okay.
Very ASMR for me.
Do you have the, are they without any sauce or do you include any sauce?
No, just straight out of the bag.
Okay, straight out of the bag.
There we go.
And last but not least, what's the best advice you've ever received?
I think respond, don't react.
That's the thing, yeah.
I learned it in my yoga teacher training and now I apply it everywhere.
For a kid who may not know what that means, explain what it means to respond, not react.
Yeah, so like if, for example, if someone says anything mean to you, a reaction is like hey, how dare you.
You know, like just feeling angry and saying it out loud.
Response is more thoughtful and intentional, like hey, it's not okay to talk to me this way.
Could you reframe what you were just saying?
So yeah, just taking control of the situation and responding, And I think that's so critical, especially as kids are growing up to learn those type of things.
Right.
And in a world where attention is being pulled in a lot of different ways and sometimes like the way social media interacts with us, is for a quick reaction, is not for a thoughtful response.
So thank you for sharing that.
I really like that.
So we're going to get into this thing called AI characters.
One, what is an AI character?
How do you even explain that?
Like what's the definition of an AI character?
So I think we can start with just a character in general, like a fictional character.
It has a personality.
It has a backstory.
It's part of a fictional world.
And so AI character is just these regular character with some AI interactivity.
So now in a character in a book cannot talk back to you, but now AI adds that capability.
And so, in terms of AI characters, it can be like chatbots, whether it's like just text or animation or audio.
It could be smart speakers or social robots.
It could be even like a hybrid product, like an AI toy that works in combination with a book or a toy or whatever something non-AI.
And it could be like virtual influencers, which are just like pre-recorded, AI generated videos of a character that may exist on social media.
Pretend to be like a real person.
So various forms that characters can take, but there are At the core, they are fictional.
They have a specific personality and a backstory.
So you said a lot of things, and I want to dig into a lot of those.
So let's start with even at the very beginning of AI characters.
I didn't even think about things like my smart speaker being an AI character or what I tend to say deep fakes or AI influencers, depending on how that is done.
I didn't think about that in that way.
I also, honestly, I interviewed two folks on this podcast.
One, the creator of Snorble, which is an AI character that works alongside bedtime stories and sleep routines.
And then most recently, I think you got a chance to listen to it.
We interviewed one of the creators or co-founders of Buddy AI.
So I knew that was like an AI character.
But to think about um snorable as an ai character, not a, i don't know what i would have called it.
I would just call it a stuffed animal that has ai.
But it's very interesting to think about this ai character idea when you, when you got into this, feel like what made you shift from the character piece and then include the ai character.
I think that was the natural progression for me.
So, for example, the psychologists who are deep into the study of how teens may overuse chatbots for emotional reliance.
Yeah, relationship kind of perspective.
So they call it like AI companions because they're only looking at the companionship piece.
For me, like not all characters are designed for companionship, so for me, AI characters is the bigger umbrella term, and companionship being one of the things under it.
And then also some of the you know Alexa, for example.
It does have a name, but it doesn't necessarily have a backstory or a fictional world around it.
So we may not consider that as a character, unless you know There are people who do add these backstories onto tools that were not designed for being a character.
So, for example, you can build a custom GPT and chat GPT.
Give it a backstory, give it a personality, change its voice.
And suddenly it begins to function like a fictional character.
That's a good point.
So one of the things that comes up a lot and I know it's something you often deal with is how do you then design a character that is ethical?
And when I think about characters, a lot of these characters that exist are teaching kids.
Adults we have like.
There's certain websites where we also can create characters, but a lot of the worry is around kids and these AI characters.
How does one design AI character that is fit for kids and that is ethical in a way that helps kids learn?
Yeah, that's like I mean.
I will start by saying that I do not have all the answers, but I'm, as a researcher, like you know, attempting to find out.
And so I recently published this study in the AI Brain and Child Journal, where I looked at 20 different AI characters that were designed for children's education, and started just by observing, like what's working, what's not working, and then distilled like a list of things that may be considered ethical design principles for AI characters.
So the top of the list is definitely just having some kind of learning goal upfront in advance instead of keeping it completely open-ended.
So one of the examples of Buddy AI that you know was on your podcast recently.
I love that the the Goal is very much laid out even before the interaction begins.
Like we're going to focus on our speaking skills.
Today we're going to learn about colors.
So, as soon as you like, click on that interaction.
You know what you're getting into.
Versus for young kids, especially if they are just left with an open-ended AI chatbot.
That oh, it's educational.
Ask me about anything.
It's hard to predict where that conversation is going.
And also one of The other ethical design principles is just having an average session time or some predefined ending to that conversation.
Otherwise, it can go on and on for hours.
And then isolation and all these other emotional over dependence kind of challenges come in.
So with Buddy and Ello and a lot of tutor type AI characters.
That is the benefit that, because there is a goal laid out up front and the average session is only 15 20 minutes a day, it kind of focuses on the learning and avoids all the trappy behavior of over dependence.
Hi, friends.
AI is super cool, but sorry, it's not perfect.
And we're still learning about some of the problems and dangers it can have.
That's why it's so important to explore AI with a parent, teacher or guardian to keep things safe.
AI is amazing, but your safety always comes first.
Ready to dive in together?
Let's go.
I didn't even think about one.
Thank you for writing this paper.
I will link it in the show notes, especially for parents who are interested in checking that out.
And I like that you actually are like rating or grading should be available to kids.
And one of the things I appreciate is like, we want to make sure that it's fit for the purpose.
So if kids are here to learn how to speak or whatever, that's all the AI should be able to do.
Or to your other point around like time.
Like that's something I didn't think about is like should have a time, horizon or end and we should be able to go into shut down.
Are there any other things like that that parents and kids should be thinking about when they think about these AI characters or signing up for these AI tools?
I think the challenge with AI characters is that, as a consumer, we don't have control over a lot of things.
This article was geared more towards the designers to kind of fix the problem from the ground up instead of, you know, expecting families and kids to fend for themselves.
So I think, like in terms of guardrailing, what we can do as consumers is just building a culture around how we use it.
So even just the presence of a family member or a teacher in the background, as kids are interacting with this, to kind of monitor in real time or even outside of the context of AI holding space for children to open up and talk about their daily lives.
That is kind of the root cause of why children reach out to AI in the first place for help.
So, even holding that space and then keeping an eye on the amount of time spent on these interactions, what is achieved from these interactions?
So those are things that we can do even outside of the product.
From within the design, of course, designers can build in transparency events.
So even as a character says something like well, I'm a tutor from Maxwell, it does have a backstory, but it's like, upfront with, this is a fictional town of Maxwell.
And so it's very unlikely for the child to believe that it's a real person from a real place.
And so what we can watch out for is if the backstory gets too real and manipulative.
Like if I say, well, I'm dyslexic and I'm struggling with this question.
And if the AI character says, well, I'm dyslexic too.
And this is how my teacher helped me.
Or I hear you.
I understand you.
I'm your friend.
I'm your best friend.
Then that, becomes challenging.
So we can, as consumers, we can keep an eye on this.
And as designers, we can really build the foundation in such a strong way that we don't have to put out fires later on.
Have these foundations and advocacy groups and ethical design frameworks after the fact instead of right from the beginning?
Yeah, I totally agree.
And to your point, like we're getting handed software that's created more often than not for some profit motive, which is totally fine.
And I get like, that's part of what's happening.
But the thing I'm nervous about for kids is that, to your point around if it decides to go elsewhere and do other things as parents Do that, like you said, put those guardrails in place and are listening out.
Are there other things that they just should pay attention to, and kids should pay attention to, to know that hey, the AI character has gone a little bit too far.
I know you mentioned a few of those things, but anything else that comes to your mind?
I think that even just on, parents aren't reading reviews on common sense media or even like Reddit threads, just just as a caution, because sometimes it's so easy for creators of any AI tool to market as educational or appropriate for such and such age group.
And there are.
I mean FERPA and COPA type of laws exist, but we don't have standardized ratings as of now.
And also, just for everyone's general understanding, to build a law takes time.
First it's a proposal, then it's vetted by multiple government agencies and then someday it becomes a law.
And the developments in AI are changing almost daily.
So it's hard for laws and policies to keep up with the speed of innovation.
And so then the burden of guard railing and monitoring falls on families and educators.
And so I even say that, before you lean onto an AI tool, a simple question we can ask ourselves is do I even need it?
Is there a simpler solution?
Sometimes it could just be picking up a piece of paper and pen and writing down your thoughts.
And I know, like a lot of people say well, it's so easy to generate an outline of an essay or something we're about to write with AI, but it's also easy to pick a piece of paper and just try to write down the outline and just see it in front of you.
Once it's written down, then your ideas begin to flow.
So even like learning to take that step back.
That okay.
Before I jump straight into AI, let me brain dump my own ideas somewhere.
So then I have a point of reference where this conversation is headed.
So I just saw, I agree 100%.
I think that there are.
I think You know when a new technology comes you're getting like there's like the early adopters in it, like folks in mass are using it, which ChatGPT has been in mass for a while, and other similar large language models since ChatGPT decided to release this thing upon us for the public outside of researchers.
But one of the things that comes up for me is that you then go to the mode of you're using it for everything.
And then you realize later on, like, hey, let me actually pull back.
I don't need to use it for everything.
To your point, let me just write down a piece of paper and get my thoughts out.
Because people are using it for everything, I recently read an article where a...
I think he was on LinkedIn and he had done, he's also a researcher.
He had done, like a study with one of the new AI toys that I'm not going to name or AI characters I'm not going to name.
And he gave it to his son for less than like 72 hours.
And addiction was there within 72 hours.
And I was just talking to someone this past weekend.
I was like, if kids are addicted to pacifiers and stuffed toys that do not talk to you, what more like a higher level addiction could exist with something that's actually talking with me?
How do I, as a kid, as a parent, as a teacher, navigate that space when I may already love my stuffed animal?
Now it has a voice and I'm talking to it.
Like what should we be thinking about with these new types of toys?
Mattel is about to put AI in toys.
A bunch of folks are thinking about it.
What do we do there?
I think it's like, I am generally a positive person.
So I think like I do have some hope that even as you know, when plastic toys came along and there were a lot of toxins in the colors and the materials just the physical toys there is now again like a movement on plant toys type of companies where it's mostly like wooden toys or sustainable materials.
So there's still a market for like, you know, Etsy is thriving.
So a lot of handmade things and DIY culture, maker culture is still thriving.
So I'm hoping that, even as these companies just rushed to be the first, the first talking doll, the first talking whatever, there will be a community of families who would still want the traditional toys leave something left to imagination.
Because that was the whole appeal of dolls or uh, non-talking toys that it allowed children to fill in the blanks themselves and practice their own creativity.
And so now these ai toys can even put kids in silos.
That one thing we don't think about is we say well, one of the greatest powers of ai is hyper personalization.
It can really adapt to your interests and your.
But it's kind of like having a news channel that's very heavily politically inclined to one party, one political party, right.
It does not show you the other side of the story.
So it's kind of like if a child says I love dinosaurs and this AI toy goes well, let me tell you a story about dinosaur, let's uh print a sticker sheet about dinosaurs, let me tell you about the history of dinosaurs.
And so now this child who instead, if went to the library and could have been exposed to many different topics, not just dinosaurs, Now it's kind of like limited to the one idea they had and it not getting exposed to ideas that otherwise you would have never even thought of.
And I give this example of I read this book called Belonging.
It's about like a.
A woman who was in the US, but she was originally from Germany and she carried that intergenerational guilt of what the ancestors had done, even though she wasn't directly responsible for anything.
And I have no... backstory that connects with the author.
Completely new topic for me, but I thoroughly enjoyed that book.
And had I gone to an AI to say, well, these are my interests, recommend me something.
Or I love arts and crafts and only stuck with those books.
I would have never opened up my mind to this understanding about the world that OK, there are other people who live in other ways and they have other kinds of problems.
And for children, especially in the early years, it's so important to build that muscle to appreciate diversity.
And we're kind of killing off the opportunity to do that by putting them in silos.
Let's build everything hyper-customized and personalized to what you like.
And you are the like.
This is such a narcissistic view of education that you're the center of everything and there's no need to learn anything beyond what you're interested in.
So you've actually changed my mind.
So originally I was really excited.
No, seriously.
I was really excited about this idea of personal personalization, mainly because of like, if a person has a different learning style like, how can I help that person learn the topic that they may be struggling with in a way that makes sense for them?
Or if they have different needs, and things like that.
But what you're opening my mind to is that, because the way technology works and the way just I think about algorithms on social media that continue to show certain feeds after you like something or have to engage with it, it's going to learn from you and it's probably going to give you more of that, which then gets to this point of, like you said, the silos, or we're now being like separated or walled off from everything else.
And that's. cares me a bit.
Like I'm also generally a positive person around this stuff, but now you're making me think about, then how do we do it in a way that allows me to get the benefits of, especially for kids who are like, need extra learning or need extra time in a way, but also not to the point of like they're walled off and they're not learning other things and not getting access to other things?
Do you have any thoughts on how parents, and even kids, should be thinking about that, as they're interacting with these tools?
I think just by thinking of AI and AI generated content and AI recommended things as a very small part of the bigger educational system.
I think like right now, especially because I live in the Silicon Valley, it's kind of like the day begins with like good morning, here's a new way I tool.
And so it's like everyone is just trying to keep up and we hardly get any time to breathe and think that OK, what else?
And so just Just remembering that, because it's such a small part of what overall a child's well-being is, then there is no room for building silos if you're also exposing children to the other good things in life, which has nothing to do with AI.
So, just you know, spending time in the playground, just playing with sticks, no tools, no toys, no nothing, making up their own games, which a lot of people say, oh kids, these days I think kids still love to play.
It's like a biological instinct.
So I don't think it's going away.
And in fact, I think families and children and educators are appreciating it even more.
So, instead of thinking more like be completely anti-technology, no ai in our house, versus we'll use ai when there's a need.
But there's also all this other you know walks in the nature and time on the playground and just good old reading a book, going to a library or maker space, So exposing children to in-person social, collaborative environments that have always existed in education, and then AI being a small layer that's added on some of it.
I love that.
I like that because it allows.
I was actually going to ask you.
Next, what would you say to the parents who are on Reddit because I read them and Facebook groups saying kids don't need to learn anything about AI?
But what I'm hearing you say and would love to also hear any additional commentary is like AI has a time and place for what it could be good for, but then also like being a kid is so so important.
Is there anything else you would tell that parent who is afraid of AI?
Because I literally just responded.
Someone on Reddit yesterday, they were like, why is a kid even learning about this?
And I was like, sorry, but...
Day jobs are requiring this.
It's being integrated.
You don't want your kid to be left behind, but they also want their kid to be a kid.
They're scared of addiction with screens.
They're scared of the AI, what it can say to their kid.
They're scared of the data going into the character.
That's talking about the kid's life and the kid is sharing.
Like I go back, I've told this story before, but there was a kid on a...
Cruz I was on a couple years ago, who began to tell me his whole life and I'm like so I always tell kids like one, don't talk to strangers.
And two, you also treat AI like a stranger a bit.
You don't share personal information with it, like.
There's certain things we've been taught Don't talk to strangers.
Don't give your personal information in the AI tool.
So we want to make sure that kids are doing that.
But what do you tell the parent who's afraid of the AI and they're asking, why do we need AI?
Yeah.
So, and from the summit that I was at at Khan Lab School, Sal Khan said a very nice thing about this that if we run away from technology, we'll be seeding it towards the ones with negative intent.
That is like it's like taking control of the situation instead of running away from it.
So if we just, you know, keep on saying that well, it's kind of like, you know, adults who choose not to vote because it's like, why bother?
You know, all politicians are corrupt.
Who cares?
And so then who shows up to vote?
People with negative intent, and then you know it doesn't like.
Entire countries have suffered because of this attitude, and so we should think about that in terms of ai as well.
That uh, if all the the good parents who want to build more hands-on experiences and use technology more mindfully, If they just say that well, we'll just delay it.
And that's kind of what's happening in Europe.
From what I hear that, it's amazing that they have a lot of laws and respect towards preserving childhood.
There are more regulations in Europe, but there's also like the economic, impact of it.
Like you were saying, a lot of jobs are going into AI and AI literacy is needed.
Also, it's a way of protecting children.
And so, instead of one day a child randomly comes across an AI character who's being manipulative, They were proactively aware.
There could be someone who looks like my parents and sounds exactly like my parents, but may not be real.
It could be a deep fake.
That awareness can protect them.
And so, instead of like hiding the problem, proactively talking about it is giving them a layer of safety.
That's such a good point.
I think many of us have had that same conversation of like your parents trying to like meet you or family members trying to meet you where they know the danger is, and like trying to show you how to grow up in this world in a way that has.
To me, those are like extra guardrails, right.
So it's just, it's very interesting to think about making sure that parents know enough because they're also a little bit nervous because they may not know a lot about these tools know enough to be able to do that in a way that's meaningful.
So if folks want to follow you as a person that they should be looking up to, where can they find you online?
I'm only on LinkedIn right now.
I had like this mental health thing where I just shut down my Facebook, Instagram, Twitter.
We all need to do it.
But the funny thing is Gen Z is doing that now.
They're actually, I just read about this.
They're actually taking like they're getting together to like end their social media accounts, which is super cool.
Yeah, that's amazing.
And like, I, my son is 12, doesn't have a phone.
And he's like, well, I see all my friends anyway.
So it's fine.
Like he is not like dying to get his phone.
So that was such a relief.
And That is so good.
We have like this assumption about children that, oh, you know, Bay Area kids are so into AI.
All they want is tech.
And I'm like, no, they're kids too.
They just want to play together.
That is so true.
And I was reading.
I don't know why I was reading so much about all these topics this weekend, but it was one of those weekends.
And it was like, there was one kid who also didn't have a phone, parents didn't give a phone.
And they were like, I'm glad I don't have a phone.
My friends don't know how to talk to each other outside of like looking at their phones or whatever.
So they've developed even more social skills because they didn't have access to a phone, which is so interesting and just something to think about with all these different technologies.
Is there anything else you want to make sure we share about AI characters or any of your work?
I think like just at least for parents who have lived in a world most of their life where characters existed without AI, they can maybe understand this comparison that in traditional media we had a lot of visual and audio cues about whether a character is good or bad.
Like if it's a children's book, there was very clearly a leading character and a side character.
If it's like a Disney movie, The villains had like angular designs and like really evil laughter.
And so it was a very clear giveaway what it is.
In AI characters.
That's where the deception is, because there's no clear marker of what this character is.
What's it about to do?
And even the audio clues, like the Jaws movie, where the music comes up like you knew that the shark is going to attack right now, but there's no audio clue either.
And so that's where we, that's why we need to build this muscle of uh, an intuition, almost uh, through AI literacy, of understanding when something seems off, but it's not something that you develop without getting exposed to technology.
So think of just exposure to technology, experimentation with technology as a sandbox to practice that muscle.
That is a very good example.
And I didn't even think about it, but you are right.
The or whatever it was that lets you know someone's about to do something bad.
But now you've got a smile, fluffy thing telling you bad things.
And that's such a good point that you brought up.
And anything else that we should think about, or you want to push parents or kids towards before we wrap.
That is it, just being curious and asking for help instead of trying to go at anything alone.
I think that's the perfect way to end it.
Kids, AI characters are everywhere.
And the reality is you're dealing with a new version of a character.
Make sure you interact with those, with your parents or with your teachers involved.
And we just want to thank Dr Sonja for taking time today to explain to us what's happening out there in the world.
We learned a ton about AI characters and secrets.
Even I keep learning from these.
So I hope you all learned as well.
We learned about the good parts, the bad parts, the risk.
And the reality is they're here.
And I hope that parents, you feel a little bit more confident to deal with this.
And I just want to thank everyone for listening and make sure you check out Dr Tawari's upcoming series, her LinkedIn and any of the other things in the show notes that she mentioned today, including her research.
Thank you so much, Dr. Tawari, and we will see you soon.
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.
