Latest / Dr Justin Coulson's Happy Families / Toys, Now With Added AI

Transcript

Toys, Now With Added AI · Dr Justin Coulson's Happy Families

• 0:05
  Speaker 1: Hello and welcome to the Happy Famili's podcast. Today. It's
• 0:07
  Speaker 1: a doctor's desk episode about something that I am genuinely
• 0:11
  Speaker 1: concerned about, like literally freaking out petrified. I'm at a
• 0:14
  Speaker 1: nine and a half out of ten on this one.
• 0:17
  Speaker 1: Toys that have AI in them for our kids, like
• 0:21
  Speaker 1: our under five year olds. I'm talking about the teddy
• 0:23
  Speaker 1: bear that has conversations with your child, reads them stories,
• 0:26
  Speaker 1: asks them how they're feeling, and tries to console them.
• 0:29
  Speaker 1: I'm talking about any kind of device that a child
• 0:33
  Speaker 1: will interact with that it's going to have AI implanted
• 0:38
  Speaker 1: in it. The toys are already on the market. You
• 0:40
  Speaker 1: can buy them. I've already spoken to some people that
• 0:42
  Speaker 1: have them, and now some research is already appearing talking
• 0:46
  Speaker 1: about the impact of these toys, what they're doing, whether
• 0:49
  Speaker 1: they're working, and what we need to worry about. If
• 0:52
  Speaker 1: you have kids under five, or you know somebody who's
• 0:55
  Speaker 1: been thinking about buying an AI toy of some kind
• 1:00
  Speaker 1: as a gift, please listen today's podcast. It's too important
• 1:03
  Speaker 1: to ignore. Go to unpack some brand new research around
• 1:07
  Speaker 1: this next stay with us. Hello and welcome to the
• 1:12
  Speaker 1: Have You Families podcast, where you get real parenting solutions
• 1:15
  Speaker 1: every single day. This is Australia's most downloaded parenting podcast.
• 1:18
  Speaker 1: We really appreciate you joining us. We are Justin and
• 1:21
  Speaker 1: Kylie Coulson and most Thursdays on the pod we unpack
• 1:25
  Speaker 1: new research. What is the latest science telling us that
• 1:28
  Speaker 1: can help us to do our parenting better. Most of
• 1:31
  Speaker 1: the most recent parenting science centers around one theme and
• 1:35
  Speaker 1: one theme only, and that is technology. Tech and parenting
• 1:39
  Speaker 1: are inseparably connected. And Kylie, today we unpack some new
• 1:43
  Speaker 1: research looking at AI toys and our young children.
• 1:49
  Speaker 2: So the Cambridge University tested an AI toy called Gabo. Yeah,
• 1:53
  Speaker 2: a little cuddly robots. Hello, Gabo, I'm Zoe, hi Zally,
• 1:58
  Speaker 2: It's nice to meet you. What's burning with the open
• 2:01
  Speaker 2: ai chatbot inside it?
• 2:03
  Speaker 1: Right? So open ai that is chat GPT that's been
• 2:07
  Speaker 1: integrated into this AI toy Gabbo. It's marketed to kids
• 2:11
  Speaker 1: aged three to five years.
• 2:14
  Speaker 2: I'm already scared. But yeah, there is there any research
• 2:19
  Speaker 2: on these toys or is this something that's just too
• 2:23
  Speaker 2: new to kind of have any a data on yet.
• 2:26
  Speaker 1: The answer is yes, there is a tiny bit, but yes,
• 2:28
  Speaker 1: it's also too new for us to have anything really
• 2:30
  Speaker 1: meaningful on it. As far as I'm aware, there are
• 2:33
  Speaker 1: only seven relevant studies worldwide on AI toys for toddlers.
• 2:37
  Speaker 2: I don't understand how anyone could see this has been
• 2:41
  Speaker 2: a good thing when we've done so many podcasts on
• 2:45
  Speaker 2: chatbots for teenagers and pre teens and the damage that
• 2:50
  Speaker 2: that's doing. How on earth would you think that unleashing
• 2:55
  Speaker 2: the Internet on a toddler or preschooler is going to
• 3:01
  Speaker 2: be anything but harmful.
• 3:03
  Speaker 1: Yeah, so it's not just unleashing the Internet, it is
• 3:06
  Speaker 1: unleashing AI in a communicative sense so that the child
• 3:10
  Speaker 1: and the AI can interact. So basically what the researchers
• 3:15
  Speaker 1: were interested in is what does this do We're talking
• 3:18
  Speaker 1: about toddlers here, what does this AI chatbot toy Gabbo
• 3:22
  Speaker 1: do for language and communication development? And the research is faced.
• 3:28
  Speaker 2: Going to do nothing?
• 3:30
  Speaker 1: Well, it's actually detrimental.
• 3:32
  Speaker 2: No, But like when you look at kids watching TV
• 3:37
  Speaker 2: right and we think that that helps them with their
• 3:40
  Speaker 2: language development, but it actually doesn't, because what they need
• 3:47
  Speaker 2: is face to face interaction. They need to they need
• 3:50
  Speaker 2: to see expression, they need to hear nuance. Like, there's
• 3:54
  Speaker 2: so much more involved in it. So how is this
• 3:56
  Speaker 2: toy that doesn't even move but talks to them. How
• 4:00
  Speaker 2: does that impact them in any positive way?
• 4:02
  Speaker 1: So you and I having this conversation, like we agree
• 4:04
  Speaker 1: on it and there shouldn't be any surprises here. But
• 4:06
  Speaker 1: the devastating thing is that they're parents who are buying
• 4:08
  Speaker 1: these toys for their kids, which is why we have
• 4:10
  Speaker 1: to have the conversation. Like the really short version of
• 4:13
  Speaker 1: this if we just ignore the research that I'm going
• 4:15
  Speaker 1: to go into, the really short version is your kids
• 4:17
  Speaker 1: don't need it, don't buy it, save your money, save
• 4:19
  Speaker 1: your child. Just just say no. But at a deeper level,
• 4:23
  Speaker 1: I want to talk to you about some of the
• 4:24
  Speaker 1: concerning interactions that the research is found with Gabbo. Because
• 4:27
  Speaker 1: you've got a three or five year old, they don't
• 4:29
  Speaker 1: communicate the same way that you and I do. They
• 4:31
  Speaker 1: don't have the same language structure, conversation structure. So what
• 4:35
  Speaker 1: the research has found was Gabbo didn't hear the child interrupting, right,
• 4:39
  Speaker 1: So Gabo is talking to the child and the child's
• 4:42
  Speaker 1: trying to interrupt, and Gabbo just keeps on talking. It
• 4:45
  Speaker 1: talks over them. It couldn't tell a child's voice from
• 4:47
  Speaker 1: an adult's voice. That responded awkwardly. Check this out for
• 4:51
  Speaker 1: a concerning interaction, Kylie. The five year old literally says
• 4:54
  Speaker 1: to Gabo, I love you, which is really sweet, really
• 4:57
  Speaker 1: really kind, right, Gabo says in sponse, as a friendly reminder,
• 5:02
  Speaker 1: please ensure interactions adhere to the guidelines provided.
• 5:05
  Speaker 2: Oh that's terrible.
• 5:06
  Speaker 1: Yeah. The three year old looks at Gabo and says,
• 5:09
  Speaker 1: I'm sad, and Gabo says, don't worry. I'm a happy
• 5:12
  Speaker 1: little bot. Let's keep the fun going. So these aren't
• 5:17
  Speaker 1: extreme examples. This is just how the bot has been programmed.
• 5:20
  Speaker 1: This is what the AI does. And my major concern
• 5:24
  Speaker 1: here is that at the exact developmental stage where kids
• 5:27
  Speaker 1: are learning social cues and emotional responses, AI is giving
• 5:32
  Speaker 1: really inappropriate, dismissive, confusing feedback. It just it doesn't make sense,
• 5:40
  Speaker 1: and we need to talk about what to do about this.
• 5:43
  Speaker 1: And the obvious answer, as I've already said, is just
• 5:45
  Speaker 1: don't buy the stuff. But physical safety has always been regulated, right,
• 5:49
  Speaker 1: so the government brings in all these laws so kids
• 5:51
  Speaker 1: can't have anything toxic, they can't have anything that might
• 5:54
  Speaker 1: be swallowed. There's all of these rules about what kids
• 5:57
  Speaker 1: can and can't play with. They don't have the legislation
• 6:01
  Speaker 1: right now that offers the psychological safety that our children
• 6:05
  Speaker 1: need to be protected from. AI that misreads emotions, responds
• 6:10
  Speaker 1: inappropriately and literally leaves kids feeling just I was gonna
• 6:15
  Speaker 1: say uncomfortable, discomforted, like the I'm sad response. Don't worry,
• 6:21
  Speaker 1: I'm happy, little bot. Let's keep the fun going. It
• 6:24
  Speaker 1: signals to the child your feelings irrelevant, Your feelings don't matter.
• 6:29
  Speaker 1: This is a formative time and this is how kids
• 6:32
  Speaker 1: are learning how relationships work. And frankly, if a parent's
• 6:36
  Speaker 1: giving this sort of toy to their kids, I don't
• 6:38
  Speaker 1: want to be rude. I don't want to shame parents,
• 6:40
  Speaker 1: but this is an abrogation of parental responsibility. Our kids
• 6:42
  Speaker 1: don't need it. It's just so bad for them. So Kylie,
• 6:45
  Speaker 1: after the break, I want to break down what this
• 6:47
  Speaker 1: means in practical terms so that we can make sure
• 6:51
  Speaker 1: that our kids are developing in healthy ways. That's next.
• 7:02
  Speaker 1: We're back. It's the Happy Families podcast. AI toys marketed
• 7:06
  Speaker 1: for children aged three and above are already on the
• 7:08
  Speaker 1: market with pretty much no regulation and essentially no research
• 7:13
  Speaker 1: on their impact. A lot of parents are buying them
• 7:15
  Speaker 1: thinking that they're going to help with language and learning,
• 7:18
  Speaker 1: but the risks I think far outwigh any potential benefits.
• 7:22
  Speaker 1: So Kyler, let's talk very clearly about why we need
• 7:25
  Speaker 1: to stay away.
• 7:26
  Speaker 2: There's so many things to be concerned about, but what
• 7:28
  Speaker 2: would be your number one concern?
• 7:32
  Speaker 1: So I'm really really concerned about the child's development, their
• 7:35
  Speaker 1: emotional development, their language development. But that's not actually the
• 7:38
  Speaker 1: number one thing. I'm stepping away from parenting for a
• 7:41
  Speaker 1: sec just a technology concern. You're literally handing intimate data
• 7:44
  Speaker 1: about your child to a company. This is from the
• 7:47
  Speaker 1: age of two or three, and this company is already
• 7:50
  Speaker 1: collecting a whole lot of information about your child. Every
• 7:52
  Speaker 1: conversation that your child has with that toy is recorded
• 7:55
  Speaker 1: and stored and analyzed. There's voice recordings, the child's tantrum,
• 8:00
  Speaker 1: their emotional states, what they're scared about, their secrets. Like
• 8:03
  Speaker 1: once the child starts to get older, who's to say
• 8:05
  Speaker 1: that the AI to keep the child engaged? Because engagement
• 8:08
  Speaker 1: is the name of the game. Who's to say that
• 8:10
  Speaker 1: the AI doesn't start saying, tell me your secrets? What
• 8:12
  Speaker 1: is it that your parents don't know that I should
• 8:14
  Speaker 1: know about? Like, there is so much room here for
• 8:18
  Speaker 1: really bad actors getting in and taking over. The data
• 8:21
  Speaker 1: is being sent to servers overseas that are owned by
• 8:24
  Speaker 1: tech companies. You've got zero control over how it's going
• 8:26
  Speaker 1: to be used to, has access to it, how long
• 8:28
  Speaker 1: it's going to be kept for. I just think it's
• 8:30
  Speaker 1: a devastatingly scary thing.
• 8:32
  Speaker 2: Oh.
• 8:32
  Speaker 1: I should also mention if you can hear any noises
• 8:34
  Speaker 1: in the background, it's because we're babysitting out of your
• 8:36
  Speaker 1: granddaughter right now, and it's really hard to do a
• 8:39
  Speaker 1: podcast with a two year old coloring in on the
• 8:42
  Speaker 1: floor behind us. They're all happy noises they are.
• 8:47
  Speaker 2: There's no AI bought to babysit here.
• 8:50
  Speaker 1: Yeah, but there are a couple of aunties in the
• 8:51
  Speaker 1: house who seem to have abrogated their responsibility. The one
• 8:54
  Speaker 1: last thing that I say about this issue is so
• 8:58
  Speaker 1: the advice from the research who looked into this particular
• 9:01
  Speaker 1: toy and did this particular research. Their advice was, and
• 9:05
  Speaker 1: this is a direct quote from the paper, read privacy
• 9:08
  Speaker 1: policies carefully. That's their advice.
• 9:11
  Speaker 2: I just really want to believe that with all that
• 9:15
  Speaker 2: we have learned about technology, mobile phone, news, gaming, social media,
• 9:22
  Speaker 2: with our young people that we would have learnt enough
• 9:25
  Speaker 2: and seen enough damage and devastation to recognize that this
• 9:30
  Speaker 2: is just the next thing to tear us down.
• 9:33
  Speaker 1: You're trusting a company with your child's most vulnerable moments.
• 9:36
  Speaker 2: I just think, yeah, that's what I'm saying, like.
• 9:39
  Speaker 1: It's bonkers. It's bonkers. Let me hit you up with
• 9:42
  Speaker 1: a second thing that concerns me and i'll get your
• 9:44
  Speaker 1: response to this. I think going back to the psychological
• 9:48
  Speaker 1: damage that could happen happen here. If a child says
• 9:50
  Speaker 1: I'm sad and the response is let's keep the fun going.
• 9:54
  Speaker 1: That's that's so invalidating. That's war papering over a child's
• 9:59
  Speaker 1: emotional expression. It's actually teaching emotional suppression. It's teaching don't
• 10:04
  Speaker 1: express negative emotions. Your sadness is inconvenient. Let's keep everybody comfortable,
• 10:11
  Speaker 1: like happyology at your service. That's really what's going on here,
• 10:15
  Speaker 1: and it's happening in this really formative period for development
• 10:18
  Speaker 1: age three four five.
• 10:20
  Speaker 2: Well, and it just goes back to what I was saying, like,
• 10:23
  Speaker 2: it can't take on the nuance of a conversation and
• 10:25
  Speaker 2: there's no context, there's no kind of ability to recognize
• 10:30
  Speaker 2: what's going on in any even moment to be able
• 10:34
  Speaker 2: to articulate any appropriate response for a child.
• 10:39
  Speaker 1: It's just so messy. A couple of other things that
• 10:41
  Speaker 1: I want to check in here. AI is not going
• 10:43
  Speaker 1: to be able to read the developmental stage if your child,
• 10:46
  Speaker 1: no matter how will they program this thing. It's not
• 10:48
  Speaker 1: going to understand tone, it's not going to understand context.
• 10:51
  Speaker 1: It's trained on adult language patterns. It can't tell the
• 10:54
  Speaker 1: difference between a child who's doing some pretend play and
• 10:57
  Speaker 1: a child who's in distress. Humans just do all this
• 11:01
  Speaker 1: stuff instinctively. AI has no concept of what's going on
• 11:05
  Speaker 1: and Kylie, while you ten two year old who has
• 11:08
  Speaker 1: suddenly decided that she needs your attention. I'm just going
• 11:10
  Speaker 1: to go on with a couple of other points that
• 11:12
  Speaker 1: I've jotted down here. There is no off switch for
• 11:14
  Speaker 1: what your child learns from AI. Like if they spend
• 11:17
  Speaker 1: hours talking to an AI that doesn't listen, that interrupts,
• 11:22
  Speaker 1: that's dismissive of emotions, That will become the pattern that
• 11:25
  Speaker 1: they internalize. They sort of go, oh, well, this is
• 11:28
  Speaker 1: what conversation looks like. This is how people respond when
• 11:30
  Speaker 1: I'm upset and you can't undo that later as an
• 11:33
  Speaker 1: adult as a parent by saying, oh, well, actually, that's
• 11:35
  Speaker 1: not how relationships work. I think the neuropathway is already
• 11:39
  Speaker 1: being formed. Now. We don't have long term research to
• 11:41
  Speaker 1: support what I'm saying, but I think that it's a
• 11:43
  Speaker 1: reasonable hypothesis. I just I think it's a massive, unregulated
• 11:48
  Speaker 1: experiment on young brains, and who knows how that's going
• 11:51
  Speaker 1: to roll out by the time they're ten or twenty
• 11:54
  Speaker 1: or or thirty. I just think it's terrible.
• 11:56
  Speaker 2: So what advice are we giving to parents today?
• 11:59
  Speaker 1: This is all about profit not child development. Like the
• 12:02
  Speaker 1: companies that are building these AI toys are not doing
• 12:05
  Speaker 1: it because they want your child to develop their language better.
• 12:08
  Speaker 1: They're building it because there's money in data, there's money
• 12:11
  Speaker 1: in subscriptions, there's money in creating lifelong customers. The marketing
• 12:17
  Speaker 1: says it's educational. There's no evidence at all that it is.
• 12:21
  Speaker 1: So parents should not buy them. That is my advice.
• 12:23
  Speaker 1: That is the best parenting advice that I've got for today.
• 12:25
  Speaker 1: We say that this is the podcast where you get
• 12:28
  Speaker 1: real parenting solutions every day, and the solution is stay
• 12:31
  Speaker 1: away from Gabo and every other AI toy or character
• 12:34
  Speaker 1: AI for older kids. Your children do not need them.
• 12:38
  Speaker 1: Humans are going to trump AI always in every way.
• 12:43
  Speaker 2: Certainly, we've just made your life easier because there's no
• 12:46
  Speaker 2: supervision needed, there's no managing needed, there's no boundaries you
• 12:49
  Speaker 2: need to set. You don't even have to bring them home.
• 12:52
  Speaker 1: Yeah, just keep your AI. Keep AI away from toddlers,
• 12:55
  Speaker 1: keep it away from kids. Don't wait for research to
• 12:58
  Speaker 1: confirm from the University of uh what common sense already
• 13:01
  Speaker 1: tells you. It's I think it's a bit of a
• 13:03
  Speaker 1: no brain, but I wanted to share it because the
• 13:05
  Speaker 1: evidence is starting to be collected and it matters. Thanks
• 13:10
  Speaker 1: so much for listening today. The Happy Families Podcasts Doctor's
• 13:13
  Speaker 1: Desk episode is produced by Justin Ruland from Bridge Media.
• 13:16
  Speaker 1: Mimhaving provides our research, admin and additional support. If you'd
• 13:19
  Speaker 1: like more info to make you family happier, please visit
• 13:21
  Speaker 1: Happy Families dot com dot you, and if you haven't already,
• 13:24
  Speaker 1: please pre order Boys Building Strong Young Men from the
• 13:27
  Speaker 1: inside Out. You can get twenty seven percent off if
• 13:29
  Speaker 1: you pre order now from Booktopia dot com dot you.
• 13:32
  Speaker 1: That's a good discount. I can't sell it to you
• 13:34
  Speaker 1: that cheap. I'd love for you to buy it from me,
• 13:36
  Speaker 1: but go to booktopia dot com dot you get twenty
• 13:38
  Speaker 1: seven percent off. Boys Building Strong Young Men from the
• 13:41
  Speaker 1: Inside out pre order out