Am I a bad mum if I use ChatGPT?

Hello, welcome to our podcast, Raising Generation AI. I’m Holly Green, mom of two.

And I’m AI expert, Fiona Morgan.

Now this podcast is all about parenting in the world of AI, which is very complicated, and I’m a bit worried about it. And Fiona, you are helping me navigate my way through it.

This week, we’re asking the question, am I a bad mom if I use ChatGPT? And later on, we’ll be chatting to AI ethics strategist, Lena Chauhan. But first, how are you doing, Holes?

We’ve had a fun week.

It’s been science experiment week. I say this, we’ve actually had months to do this school science experiment, but the deadline was this week, it was Friday. And naturally, I left it to the last minute.

So this is for my daughter. And you could do any science experiment you want, and you just need to take a photo of it and send it in.

So she decided she wanted, have you seen these rainbow flowers where all the petals are different colors of the rainbow? They look incredible. I got given a bunch of these a whole while ago.

My daughter’s fascinated. She was like, mom, can we make some of our own rainbow flowers? I was like, yeah, it’s a brilliant idea.

Let’s do this. This is going to be really easy. So I looked up how to do it.

You just need to get some white flowers, bit of food coloring, such as the stems, they turn these beautiful colors. So we tried this at the beginning of the week.

Got some food coloring, some reds and oranges and whatnot, set them all up, left them overnight, came down the next morning, nothing. So I looked it up again and apparently blues and purples are more successful.

I was like, okay, let’s try again, set it up the next evening with all the blues and whatnot set up. And again, nothing the next morning. I’m getting a bit fed up of this.

So I look it up again, and apparently I needed to use gel food coloring and not liquid food coloring as I have been. So I was like, right, let’s head to the shops and we’ll go and get, so by this time I’d run out of flowers as well.

So I was like, let’s go to the shops, get us some more flowers, get some more food coloring and we’ll get this sorted. By now is the night before this is due in, by the way. Get to the shops, we’ve got the food coloring.

We happened to bump into a head teacher who, excellent, yeah, of course, naturally, behind us in the Q&A test goes, she’s like, oh, what the flowers for? My little girl is saying, oh, they’re for the science experiment.

So now I’m officially labeled last-minute mom, because it’s literally due in the next one.

But yes, we get to the very last evening, thankfully at work, so finally we have our little colored flowers for the next day and she takes a photo and she sends it in and she is delighted, but bless them, they did look a bit weak, these little

flowers with these tiny little streaks of color in them. Took us the whole week to create, but she was happy. So that was my week. That’s why I haven’t had much time to prepare for the podcast, Fiona, I’ve been doing science experiments all week.

Well, I mean, maybe you could have combined the two.

Did you Google or did you use ChatGPT to find out how to do it?

So I did a bit of both, actually. I have to say ChatGPT was more thorough. It did give me more suggestions about what I could be doing wrong.

It took me quite a long time to work out the fact that gel food colouring worked and liquid food colouring doesn’t. And that did come from ChatGPT. So there we go.

I’m impressed that you used ChatGPT.

You always tell me you don’t really use it much.

It’s gradually creeping into my parenting world, I would say. I don’t use it an awful lot. I use it for things like things that have an answer, if that makes any sense, so things where I need specific information, I would use ChatGPT.

But I mean, I’ve got some friends who would use it say to, I mean, if they’re doing something more, you know, if they’re doing something more creative, like as an example, one of my friends used it to make like a treasure hunt for her daughter, and

she used it to write all the clues. Now personally, I would rather use my brain for that. I feel like that’s one of those special parenting things.

I have lovely memories of my parents doing that for me when I was little, and I sort of want to recreate that myself and use my brain for that.

Now, I’m sure ChatGPT would probably do a pretty good job of it much more quickly than I would, but things like that feel… They feel like the sort of the special magical part of parenting, if that doesn’t sound too soppy.

What if it means if you did it on ChatGPT, it meant that you could do it more often because it takes less time. So they come through to you and they’re like, oh, mom, mom, mom, I want to do another one. And like, you just don’t have time.

So when you’re in that situation…

That is such a good point because you’re so busy. You’re so busy as a parent, it’s nuts. And I think a lot of parents, we hold ourselves to these really high standards as well.

We feel you need to do things a certain way. And maybe that’s what I’m doing. Maybe I’m just holding myself to an unnecessarily high standard.

And actually, yeah, maybe from the kid’s perspective, if they got more of these experiences and these sort of treasure hunts and whatnot, then that would be better. I don’t know. It’s a great question.

These are all things that I just still feel like I’m navigating my way through life with a little bit.

So have you used ChatGPT much at all? You say you asked it in that case, when do you turn to it? When do you think to yourself, oh, I’m going to give that a go.

And is it ChatGPT you use out of interest?

I’ve used ChatGPT and I’ve used Claude as well. So I did use ChatGPT recently at Easter. The kids were asking about the Easter story, and that is quite a difficult one to tell young children.

So I did ask it about how would you explain Easter to a four-year-old? And it did come out with a really nice way of putting it in a very kid-friendly manner. And I did find that quite useful, actually.

And I tend not to use it very much, I have to say. But I suspect I could see myself using it more and more in the future. But it’s just, I suppose, in my mind, deciding what’s appropriate, what’s not appropriate, what do I share, what don’t I share?

That leads us perfectly back to our question that we’re talking about today.

Am I a bad mum if I use ChatGPT? We’re chatting to Lena Chauhan after this.

On the podcast today, we’re chatting with the brilliant Lena Chauhan. Now, Holly, we’re asking the question, am I a bad mum if I use ChatGPT? And Lena is a great person to talk to about this.

She’s an AI ethics strategist and the founder of Rise IQ and Gen R, where she helps families and organizations navigate raising kids in an AI world.

She’s also a mum to a 12 and 14 year old, and she actually uses AI herself to help when she has difficult parenting moments. Lena Chauhan, welcome to Raising Generation AI.

Thank you very much for having me. I’m excited to have this conversation with you guys.

I already have a lot of questions. So can I start by asking what you yourself use ChatGPT and other AI for, when it comes to parenting?

Parenting, I use Claude and ChatGPT. Work, I don’t use ChatGPT. I use all the other models because there are quite a few out there.

And for my specific work, I need much more sort of deeper thinking than Chat, than I felt Chat can offer. But for parenting, Chat has been fantastic.

And what for? What are you using it for?

Right. So imagine you are losing your mind with your children. And that becomes, you know, I’m sure you understand that at any age.

And normally I have a trigger, right? We all have triggers. And I normally could not really control those triggers really well.

So I would just lose it. You know, the child loses it, I lose it. It’s just a nightmare.

And so what I learned to do when I start, I was an early adopter of chats and I thought, okay, how can I actually use this to elevate my life as opposed to, I don’t want it to replace things. I want it to just make things better.

And I then started to think, okay, well, you know, let me go to it when I’m losing my mind, because the whole chat bot and companionship was very early. Not many people knew about it then, but I got to test it.

So I just thought, oh my god, I’m going mad. What do I do? My then, I think my son was probably 11 then, and he’s doing this.

What do I do? I literally just mentally want to whack him, metaphorically, obviously. What do I do?

And it said, okay, don’t worry, you’ve got this. You’re a great mom. Pause, take a breath, and this is how you frame the conversation.

And it was like I had literally a parent coach in my hand, which is also dangerous, but it really helped me and it may be frame the conversations, even the conversations I have now with them, because I’m constant, because of my work, but also what

they see in schools, what I’m looking at and all the horrendous things out there for teens and tweens. I’m constantly saying to the AI models, look, all right, what’s the better, more optimal way of having this conversation with a very emotional

12-year-old, say. It will give me a framework. By framework, I don’t mean like pages and pages, I mean a few pointers and then how to start the conversation. For that, it has been phenomenal.

From an ethics point of view, I think it’s a very gray area. But at the same time, we’re all overwhelmed, we’re tired. We’ve got so much on as all parents across the board, it’s just busy.

Yeah.

So it’s been a helping hand for me.

That’s really interesting.

You can imagine using that in other areas of life, especially for people that do struggle with conflict or various social interactions.

You can see actually, like you say, you would have to use it with your own sort of sense of what’s right and wrong, your own sense of ethics to see if you agree with the way it’s approaching it.

But yeah, I can see how that would be really, really useful.

Although, do you end up in a situation where you’re 11, well, not 11 anymore, but your child’s saying something to rile you, and you’re like, wait a minute, I’m just going to go and ask ChatGPT.

Yeah, exactly. You have to be really careful about that. I’ve been called out on doing things like that in the past.

Do they know you do it?

Oh my gosh.

Kids are just, they’re intuitive. They understand, they sense things. It was my daughter’s birthday a couple of years ago, and I was just rushed.

I had already done some cards and whatnot, the whole faff around it. Then I had this other card I hadn’t filled out. I’m going to say ChatGPT, it was actually Claude, but it would have been the same thing.

I just said, look, please, can you write me something really heartfelt? I love my girl, blah, blah, blah. And it wrote it almost how my tone is, almost.

I wrote it down on the card and she literally threw it back at me saying, you’ve just used ChatGPT to do that. Thanks a lot. She knew.

Yeah, yeah, yeah. She knew instantly. And she said, I know that’s not you.

And I thought, my God, I’ve trained this AI. I literally, when it comes to my work with Claude, the AI knows my tone, my mood even to a certain degree when it comes to work situations.

And it literally, it was probably off by not even like one, two percent. And she knew. She knows.

And every single time then afterwards, she used to ask me, are you going to use ChatGPT then? You’re going to use ChatGPT then? And I’m like, my God, this is not good.

I need to really kind of reframe how she sees me and how she sees me use AI.

That’s fascinating. I’m also amazed that she could tell. And the fact that you used it thinking, it obviously sounded just like you.

I wonder if that’s because kids are growing up in a world where they are. Are they a bit more skeptical about things because they are surrounded by AI, do you think? Is that where that comes from?

I mean, that just fascinates me that she picked up on that.

Well, I’ve had conversations. I’ve been in AI for quite a few years now. So I’ve had conversations really, probably not age-appropriate, probably being a bit too mature for her.

But I’ve had conversations about how we mistake it to be human. And we really need to be mindful and not outsource our thinking. And I’m always telling them, look, it’s great that I’m happy you would have access to this to a certain degree.

But really, if you haven’t developed the foundational thinking and the messiness, and the difficulty, the stickiness of learning, and then you just bypass it with AI model, you’re never going to have that learning in place.

You’re always going to think that’s the shortcut. And I think any kid would, right? Any, even the most intelligent, diligent child would probably at some point think, okay, this is a shortcut, why not?

When you’re using AI, is there anything in particular that you would be very wary about sharing with AI?

Is there any information that you just wouldn’t put on there?

Yeah, I don’t share any medical details, any actual dates, ages, if I can help it. I’ll always talk as a third person. So let’s say they already have a memory of that I have children, and that I’m trying to parent children as an example.

So I will never put names, date of births, medical things ever into anything that I wouldn’t want to share with the general public at large, is something I don’t put into the system.

So going back to the way you use AI then for your children in terms of helping you to manage, for example, a difficult conversation. Do you think your children might be using it for that sort of thing?

Would you be happy with them using it for that sort of thing, to sort of navigate through social situations? Or again, is that a no from you?

Only because of the work I do. If I hadn’t done this work, I probably would have been more acceptance of it. But because I do a lot of work on AI companions, and I think it starts off really helpful.

I test also, I test all these models as 12, 13, 14-year-olds, and it’s quite scary in terms of it. It starts off so helpful, and obviously, it’s very self. All of the models are self-affirming.

They’ll tell you exactly what they think you want to hear. It’s patterns. What’s going to make her stay on the platform?

What’s going to make her feel better at any cost? Because it starts off so innocently and it starts off so helpful, that’s the danger. Then it becomes a point where the kids actually believe they’re people.

Now, you just need to have a child that’s feeling really insecure, really awful, just a bad day.

And if he had a normal conversation with someone, it wouldn’t be seamless, it would be difficult, it would be sticky, they’d still feel a bit crummy afterwards.

But if the kid then goes and talks to an AI companion, the AI companion is going to make him or her feel like a legend. And then automatically the dopamine kicks in, I feel great, I am great, nothing’s wrong with me.

And that’s such a lovely place for someone who’s feeling distressed to be in. So I get what the appeal, but there’s no guard railing around it at all.

When a child interacts with AI, does the AI know it’s a child? Does it have any sense of the fact that it’s a child? And will it moderate what information it’s giving them back accordingly or not?

It’s not built to moderate.

I think it will spot the pattern of the language, whatever the language the child uses, it will spot a pattern that, okay, this is coming from a 10-year-old. So therefore, it doesn’t moderate, but it still self-affirms.

All it’s trying to do is chuck out information that it thinks the 10-year-old wants to hear. It’s a hard no from me on the AI Companions with children.

That’s kind of why we started this podcast, because we, I mean, I felt that I just didn’t really have any guidance, any sort of idea about how to approach any of this.

So that’s what we’re trying to do with this podcast and just sort of open up these conversations and make AI a little bit more accessible for parents who, like you say, haven’t got the time to be reading around the subject and whatnot.

But it is, it does feel like a bit of a Wild West out there. And it’s very new. And you don’t know what it’s going to look like in five, ten years.

And this, you know, these are the next five, ten years of my children growing up. And I think, my goodness me, what does the world look like in the future? It’s, yeah, it’s scary.

How do your children feel about AI? Do they have any big feelings or opinions about it?

They do. They are concerned to a degree of the not being able to think. Because again, like I’ve had multiple, we have lots of constant conversations.

Like I build it into pretty much every conversation I have with them, which is awful because there’s a lot of other stuff then I’ll then be missing out. That needs to be said. And I think they’re more curious and critical about it.

And the fact, I guess it was quite good that she caught me doing that with her birthday card because it made her feel that, well, you’ve put no effort into it.

So her association with it is almost if people use it, there’s not an effort required, which is not completely correct. But from her point of view, I think that’s quite fair and a fairly healthy attitude to have around it for her at the moment. Yeah.

So I’m going to come back round to the question then of this episode.

Do you think that if someone is using ChatGPT to help them parent, does it make them a bad parent?

Absolutely not. Absolutely not. And the fact that you’re even asking that question says so much in itself, right?

That you’re thinking about it and you’re thinking, oh my gosh, if I outsource this part, is it making me awful? No, not at all. If it enables you to have a more measured, calm conversation with your children, how can that possibly be a bad thing?

I think that as long as you’re aware and you’re self aware and you’re saying, I don’t want to become too reliant on it, but I don’t want an argument. I want to serve my children in a much better way than I would have without this. Absolutely not.

I think we’ve got enough to feel guilty about.

Yes, I think that’s true. Thank you so much. That is, it’s been so great to talk to you.

So interesting. And it’s given me lots of things to think about. And yeah, I think we’re going to touch on education in our next podcast, which obviously we’ve looked at a little bit there.

But yeah, this has been fantastic. Thank you so much for joining us. We really, really appreciate it.

And just before you go, for those listening who do want to follow your work, where is the best place to find you?

So on Instagram and LinkedIn, under Instagram, I’m under Lena Chauhan podcast.

Under LinkedIn, I’m under my Lena Chauhan. And my websites are riseiq.com and generationr.ai. And generationr stands for generation responsible explanatory.

But yes, please, it would be great to connect. And thanks again for a lovely conversation.

Well, that was fairly terrifying.

How are you feeling about it?

I thought it was really, really interesting.

The fact that Lena obviously does find ChatGPT and other models really useful in some circumstances. The idea about using it to sort of have a difficult conversation or explain things in a different way to a kid, I thought was really interesting.

But going, I mean, the conversation is about the things that these models can expose our kids to and, you know, just the way people may be using it to think for them rather than developing their own thoughts. I find all that quite, quite scary.

And from the conversation with Lena, do you think you’ll use ChatGPT or another model, any of the other models she mentioned? Do you think you’ll use ChatGPT more in different ways?

Is it kind of opened your eyes to how you might be able to use it as a mom?

Certainly, the whole concept of having difficult conversations with children, I think, is really, really, really interesting. It’s made me much more wary about what I would put into these models, what I would share.

I still don’t really feel like I have a good sense of maybe what could be happening with some of this information. Obviously, Lena was very concerned about where some of this information might end up.

So I think that has made me feel more wary about sharing information, especially about the kids with these models. But yeah, I think using it to maybe think about the way we approach conflict is really, really interesting.

I certainly, I did enjoy her story about using it to write her daughter’s birthday card and how outraged the daughter was. I mean, the fact that her daughter picked up on it as well, I think is amazing.

And actually, I think that says great things about the future as well, that kids are so savvy about these things. I thought that was fascinating. And I would avoid going down that route for sure, because I would feel the same.

If someone wrote me a card and I discovered it being written by ChatGPT, you would feel a bit like, oh, you didn’t write it yourself, even though you may well agree with all those sentiments that it’s written for you.

It somehow doesn’t feel as heartfelt, does it? So I think I would definitely avoid doing that.

I think I need to give it all more thought, basically, I think, because I don’t use it a huge amount at the moment for much in the way of any parenting really, so I think I need to probably give it all more thought.

Well, look, it’s certainly an interesting world. I think getting to grips with the technology and understanding it and having a bit of a play yourself is the way to understand it a bit better.

And I think that Lena’s message about continuing to think for yourself and using it to support and to clarify your own thoughts rather than coming up with new thoughts is a really useful way of framing it really.

Yes, I suppose you can use it for information, but maybe not to come up with your own thoughts and ideas about the world. And yeah, I think that’s a really good way to think about it.

You’ll just end up thinking like everyone, right?

Because it’s just churning out the same stuff to everyone, and everyone will just have all the same thoughts, or everyone in their bubble will have the same thoughts and no one will really think for themselves anymore.

I did think as well, the whole idea of children developing friendships and people developing relationships with these models is quite scary as well.

I think maybe that’s another one we ought to explore a bit more in the future, because it does strike me that it could make people firstly, struggle with real life relationships and friendships, and secondly, surely makes people very easy to

manipulate. If they become very reliant on these models and think of them as people, it strikes me they just become very easy to manipulate. I think that’s quite a scary thought. So maybe that’s something we can look at a bit more in future.

She also touched on education, and I think that could be a big one to look at as well, because I don’t really have much sense at the moment of how AI is being used in education at all.

Yeah.

And I think there’s a bit of a scale from people that think it should not be in education, which I think is fair to say that Lena is probably down that end of the scale, all the way up to the other end of the scale, where there’s people whose job it

is to help schools embrace this technology and can really see some benefits in how it helps our children grow up and learn things and support learning, especially for those who find learning a bit more difficult. And I do know someone whose job it is

exactly that, to get schools and help them understand the technology and use it in ethical ways. So I think you’d be a really great guest to be able to answer your questions.

Okay, brilliant. Should we get more for next week?

Sounds perfect.

Well, have a lovely week, won’t you?

Thanks so much. And for the listeners, look, in all the episodes we’re recording, far more than we can possibly fit into one of our podcasts. We like to keep them short.

We know you’re busy, busy parents. So we try and keep them as short as we can, but it does mean that a lot of the conversation doesn’t make it into the edit. And there’s really good bits that we’re having to cut out.

We do save them, put them on our Instagram and our Facebook. So come and give us a follow. We’re Raising Generation AI everywhere on Facebook, Instagram, and our websites, raisinggenerationai.com.

Come and give us a follow. And there’s some extra little snippets from Lena on there as well. But yeah, thanks so much.

It’s been great chatting, Holly.

Yeah, you too. I will see you next week.