← Machines Like Us
Machines Like Us
Social Media Bans Are Wildly Popular. They Might Also Be a Mistake.
00:58:16
Towards the end of last year, Australia did something no other country had ever tried: it banned social media for kids under 16. And a bunch of others are following with similar laws, first Denmark, then France, then Indonesia and Austria. All in, there are now more than 25 countries that have either implemented, or are actively considering, social media bans for kids. It seems like Canada is moving there as well. In April, the Liberal party adopted a non-binding motion to restrict young people’s access to both social media and AI chatbots. All over the world, you can hear parents breathing a sigh of relief. They’ve spent the last decade watching their kids become hooked on their devices, and now we’re doing something about it. It looks like we’re finally going to get our kids back. But researchers like Candice Odgers are skeptical. Odgers is a psychology professor at UC Irvine who’s been studying the digital lives of young people for almost 20 years now, long before anyone was wor...
Loading summary
Transcript114 lines
- [00:00]
A
Hi, I'm Hailee Steinberg, senior style editor at the Globe and Mail.
- [00:04]
B
I cover the fun parts of the
- [00:05]
A
news, from Taylor Swift's engagement ring to
- [00:08]
B
decoding the latest viral fashion trends.
- [00:10]
A
If that's the kind of journalism you care about, head to globeandmail.com subscribe hi, I'm Taylor Ohn from the Globe and Mail. This is Machines Like Us. Towards the end of last year, Australia did something that no other country has ever tried to do. They banned social media for kids under 16. And a bunch of others are implementing similar laws. First Denmark, then France, then Indonesia, and Austria. All in. There are now more than 25 countries who have either implemented or are actively considering social media bans for kids. And it seems like Canada is moving there as well. In April, the Liberal Party adopted a non binding motion to restrict young people's access to both social media and AI chatbots. All over the world, you can hear parents breathing a sigh of relief. They've spent the last decade watching their kids become hooked on their devices. And now we're doing something about looks like we're finally going to get our kids back. But researchers like Candace Odgers are skeptical. Odgers is a psychology professor at UC Irvine who's been studying the digital lives of young people for almost 20 years, long before anyone was worried about what social media was doing to their brains. Auger says that there isn't really any research to suggest that these social media bans will work, but her argument goes deeper than that. She says this idea that smartphones have caused a youth mental health crisis just isn't supported by the ev. So as governments all over the world start to kick kids off of social media and maybe even AI chatbots as well, Candice Odgers thinks we're making a serious mistake and I want to know if she's right. Candace Audgers, welcome to Machines Like Us.
- [02:24]
B
Thanks for having me.
- [02:25]
A
So I want to start with the story I just mentioned, which you know well, and it goes something like this. Sometime around 2012 or 13, young people got smartphones. They all signed up for social media, and following that, their mental health plummeted. What does that narrative get wrong?
- [02:45]
B
Yeah, so a couple of things. You know, we started in the field in 2008, actually following young people on their phones, and about 60% of them had phones there at that point. And it's been steadily rising during that time.
- [02:57]
A
Those would be dumped. Those would be like flip phones right at the time.
- [02:59]
B
No, those were smartphones. Yeah, smartphones. Social media took off around 2012. And I think that's where people like to draw the line. But it's interesting. Smartphones came sooner, smartphones came sooner. And, you know, it is the case that that use has increased during this time. And people try and tie that uptick to, you know, youth mental health, but, you know, the youth mental health. So I'll give you the example. In the United States, suicide risk started increasing in around 2008, right. In the aftermath of the Great Recession. Perhaps we shouldn't be surprised because adult suicide rates had been increasing since 1999. So between 2000 and 2020, 1.2 million children in the United States lost a caregiver to deaths by firearms or drug overdose alone. And that rate in the time that you mentioned just now, 2012 to 2021, it doubled among adults. And so we keep looking at young people and thinking, you know, what else could have been happening during this period of 2012 to 2020 other than Instagram? Well, in the US adults were dying, caregivers were dying at an increased rate. And we know that the number one predictor of youth mental health is the adults around them and how well they're doing. So that's one of the big things I think we're missing in the story.
- [04:21]
A
Are those trends sort of consistent across countries or is that an American phenomena?
- [04:26]
B
So many things are an American phenomenon, including exporting fear around this issue. So many of the bestselling books that are, are telling the story about this, are telling an American story, and then kind of cherry picking different types of stats around the world. It actually is the case in the youth mental health world and suicides in general, that America is an exception. And so, you know, suicide has been increasing in the United States. It's been decreasing in most other countries. So there's something especially wrong with American adults right now.
- [04:57]
A
Interesting. I mean, we do hear from a lot of parents that they're seeing their kids spend an inordinate amount of time on social media and they are seeing an effect. How do we. What is that? What's going on there?
- [05:12]
B
Yeah, so it's clear that young people are spending a lot of time in digital technologies. It's also clear that parents don't want or like this. And so I think a lot of the things that people, people that are leading these movements are getting, right, is that adults want something different for young people. Right. So they want them to be spending less time online. They want tech companies less involved in their lives. And that's a values and a normative based argument. And you know, I can certainly get behind that as a parent. And we. We see that, you know, really strong support. What, what people get wrong is when they try and, you know, weaponize the science and tell a story that science tells us that social media is a major contributor to adolescent mental health problems, or that is a suicide epidemic or is rewiring children's brains. If you want to make that argument, the data and the evidence aren't there. So you can make the value argument that this is not something adults want and they do not like it and they would not be allowed, they're not
- [06:06]
A
mad, and they'd rather government take it away. And all those things are sort of normative arguments that we can all.
- [06:11]
B
Yeah, and that's, you know, I'm not. I'm certainly not here to tell any parent how to raise their children or the number of hours they should spend online. But if they're making those decisions because they're scared, because they think that the science says that this is the causal pathway to suicide and major mental health problems, that's simply not supported in the evidence that we have today.
- [06:31]
A
So those are two things, right? Like the suicide rates and serious mental health problems. But what do we know about what spending five hours a day or seven hours a day on social media actually does to us? Like, surely that's not good for us, but what is the effect it has on us? Do we know anything?
- [06:48]
B
Yeah, so this is actually interesting. So in our research, what we do is we actually follow young people on their devices and we get daily reports from them on their doing. We install, you know, tracking to understand kind of where they're going online, what they're doing. They also wear things like Oura rings or Fitbits, and we understand their exercise and their sleep. And the biggest thing that you find is there's a huge stratification like there is in everything else, based on socioeconomic status. So the amount of money your family has and the amount of resources you have for parents to pay for extracurricular and sporting and club everything that really just structures how much time is available in a child's life to spend time online. Right. And so we find big differences across socioeconomic status in terms of how much time young people are spending online. So when you throw out numbers like, you know, five hours a day online, you know, that's estimates based on reported data, but objective data is actually quite lower, much lower. And it's heavily dependent on what opportunities are available for young people. So we often view this as adults as, like, they're choosing to just do this, but we could choose to give them opportunities to do other things. And we have kicked them out of public spaces. We've taken away opportunities for community level sports and activities. And so when you offer young people those kinds of chances and opportunities, they'll take it. But we've taken many of many of those opportunities away.
- [08:12]
A
The debate often ends up using terms like online. Social media, phones, the Internet, all is sort of one thing. And the screen time debate definitely does that. Right. That just being on a screen is itself a gateway to a specific kind of activity or content.
- [08:30]
B
Yeah, we're on a screen together. This might be bad for us.
- [08:33]
A
Well, we are. Well, I mean we certainly are. I spend a lot of day on my screen, which makes convincing my 13 year old to get off them incredibly difficult. So what do we know about those differences here though? Is there certain kinds of social media that we think are worse than others? Is there when you look at people's, when you look actively and track actively the behavior of kids, do you differentiate between what they're doing versus just being on device?
- [08:59]
B
Yeah, and so that's been a big limitation of prior research, has just been kind of counting hours or minutes. And we know that's very different. Right. You could be on going through Khan Academy and refining your math skills or you know, communicating with a relative that's, that's far away or you could be passively consuming content that's really not appro. Age appropriate. And so there's a couple things that kind of stand out. One is that the most important predictors of viewing negative content or encountering negative things online are offline risks. So not being supported offline, having pre existing mental health problems. And we know that those young people are going to run into more problems online in part because of what they're seeking and then because the algorithms amplify that, that back. So that's a cycle that we worry about. Now it's really fascinating though, you know, we have a lot of assumptions about how this works, but there's been a few really, really large studies recently, one with 26,000 young people in the United Kingdom that found absolutely no connections between gaming, social media and internalizing. So anxiety, depression, the things we're talking about. And then another huge study of 10,000 adults that followed them every day, passively sensed and scraped what kind of apps they were on, social apps or gaming apps or information apps. And then they also had information about how they were feeling moment to moment. And they found absolutely no associations between different types of engagement, social media Apps. So even there, I think we need to check some of our assumptions about what the tech is doing to us. And the most compelling evidence we find is that what really drives these associations, when we find them, is how you're feeling influences whether you go online and how long you spend there and what you.
- [10:42]
A
And what you do ultimately and what you find there. Yeah, yeah. If the evidence is as murky as you say it is, or certainly uncertain, why are so many so confident of this connection? I mean, there's very few policy debates that I've engaged in where there's just such disconnect potentially there. Right. Between what people feel and believe and see in their own lives and in their kids lives. And what I think the research community. Yeah. Is broadly saying. Right. Which is. It's more complicated than you're making it out to be. Like, I think there's a lot of debate among the nuances of how complicated it is and what we know and don't know for sure. But like, there's a real disconnect there. Like what explains that?
- [11:23]
B
Yeah, it's interesting. I'm often labeled by some people as like the skeptic because I'm the person that kind of stands and puts my head up and gets people to yell at me for saying this. But this is actually pretty broad consensus that it's an unsettled or weak science. So the National Academy of Sciences convened a panel, they reviewed this evidence, they took testimony, and they essentially came to this conclusion that you see this mix of really weak associations, no way to distinguish cause from effect. We've seen some experimental research that's come out recently, virtually none with kids, but with adults, where you do social media reduction studies and you find the effects can't be distinguished from zero. Right. This is even when you tell people social media is bad for them and taking away is good, we find effects that are close to zero. And then, you know, even this week we see the smartphone ban data that dropped yesterday. That essentially.
- [12:13]
A
In Australia. From Australia.
- [12:14]
B
No, in the United States, a massive study dropped yesterday by hundreds of schools.
- [12:19]
A
In schools.
- [12:20]
B
Yes. Yeah. Zero. So close to zero effects on test scores, on bullying, on attention in class. You know, what you do see is an immediate rise in suspensions for kids violating the yonder. So I think that all that is to say that, you know, the evidence that we're seeing coming out of this, where you actually take the time and measure it carefully, is really mixed, small effects, nothing that would match this kind of level of concern and campaigning to isolate technology as the cause of those. And so then your question is like, you know, why is that? Why is there this massive gap? And I've really struggled to understand this. And this has been one of the most interesting parts about working in this field is like, why are we so far apart on this? And I, and I think it, you know, it comes back to this simple truth we all know, which is scary stories self, right? And the more often you hear something, the more often likely you are to believe it's true. I mean, 50 years of communications research has told us that. And so we are bombarded every day with headlines that tell us that our young kids are broken, that they've been destroyed by smartphones and social media, that their brains have been rewired. And we just constantly hear that message and parents are anxious to begin with. You know, we're an anxious lot. And so those messages resonate. And then you combine that with the fact that every generation looks back at the next one and judges them for how they spend their time and thinks they're doing worse off. So I think that's how we get here.
- [13:49]
A
I struggle with that too. But, but I also sometimes wonder if the research isn't always capturing some of the discrete nature of the harms that people hear about and experience, frankly. And you think of like all the whistleblower testimony that showed what the platforms themselves know is happening, right? Like the tens of thousands of people being direct messages by adults they don't know or the sex trafficking that's happening or the whatever. Right. The child sexual abuse material that's circulating. These things are real, but we're not studying them. When we look at suicide rates or long term mental health decline. Right. Like, are we studying the right thing? And is that part of the mismatch?
- [14:30]
B
Yeah. So I think, you know, both things are true. All of the, there are real harms online and those need to be cleaned up. And tech companies are awful and it feels good to hate them because they make awful decisions. Right. And social media, if I was to target and think about improving adolescent mental health, that is not where I would start. And so the harms that you mention, I think are interesting because one of the things that we've argued for is that, you know, this conversation we're having, which is really, what does tech do to kids? Like, and we basically point a causal arrow from technology to kids is the wrong place to start. Instead, let's start with like, what is the problem that we're trying to solve for young people? And we censor that. So let's take the harm example. Right. So I started my career working in child sex abuse in Vancouver with sexually exploited youth. And when we look at that question, we will ask, like, in what ways could technology facilitate and help young people who are struggling in those conditions to get support, to get out, and in what conditions does it amplify risk? And if we really cared, if we really cared about child sex abuse, then what we would do is instead of requiring companies to just report up all these instances so that they don't get sued and they're not liable, which is what's happening, which is what makes the numbers look awful, we would take money and resources and track down the perpetrators of those crimes. Right. We would have people on the back end. We would have FBI agents, we would have special units. We would go after the perpetrators of the crimes instead of punishing the victims by kicking them off out of these spaces. Right. So if we were trying to solve
- [16:02]
A
that problem, we could also make the companies change design features that are more susceptible to this or not. Like, there's other things we could make. Yeah, yeah.
- [16:11]
B
But kicking, you know, the fact that we have broken the Internet and we have broken various parts of society for children, and our solution is to kick kids off and pretend that it's going to be okay is not going to help kids in the end and will not help the sextortion or the child sex abuse that we see. And, you know, the vast majority of harm to children happens in their homes and their schools and their churches and their sports teams. We don't shut those down. We work to make them safer. Right. So anywhere that adults are young, people will unfortunately be harmed. And so the Internet is the same.
- [16:44]
A
I mean, I think it's fair to say that this debate has changed materially pre and post Jonathan Haidt's book. And I don't want to get into a debate about everything he said and hasn't said and so on and so forth, but his voice is very important, I would say, in the actions that governments are taking around the world. How do you explain that? What's going on there?
- [17:12]
B
I mean, he's a gifted storyteller. He's telling a story that parents are primed to believe, and it's a scary story. But at the same time, I think he's capturing this overwhelming desire among adults and among parents to hold tech accountable, to recapture, you know, their childhood in these various ways that we find troubling. And so it's really that normative value judgment movement. And, you know, I think, you know, he frames it as A collective action problem. And you know, I think that, that, you know, in part that is a value and a normative judgment that what I'm saying about the science, you know, may or may not matter. You know, where I take issue with it is kind of weaponizing the science and then telling parents, telling young people, shaming young people about use, that's very kind of normative and what that message does to them there and then really the opportunity costs. So from a mental health perspective, and so many studies have just found that social media is one of the least influential factors in predicting adolescent mental health. And in fact, that's like a direct quote from one of the most recent studies. And so if we really are addressing adolescent mental health as a problem, social media isn't where we start. Well, ask, in what ways does it amplify risk or particularly on the people
- [18:31]
A
most vulnerable, as you said before. Right. Which is like what he says too, right. Is that it might not have a dominant effect on the majority, but where it does, it's important. And I think that's probably what you're saying too. Right. But the way of getting at that might be different than taking it away for everybody.
- [18:46]
B
Right. And I think that the costs of doing that, you know, there's also, you know, it's what you often find is a bit of a double edged sword with young people who are most vulnerable are the most likely to report benefits online from the technology of finding community and support and safety in online communities and also experiencing the most harms and encountering the most harms online. So this is also like a delicate balance that we can't just assume that, you know, we need to restrict among people who are most, most vulnerable, which is really, you know, it's interesting. That's what a lot of the bands do, because highly educated and resourced parents. We know that one of the biggest things when new technologies come out is, is they amplify inequalities and youth with the most resources are going to get the most scaffold and personalized and safe experiences. And unless we build systems and tech and teach it in our schools, we will be leaving an entire group of young people behind. The ones that are kind of least positioned to benefit in lots of ways.
- [19:44]
A
Yeah, you mentioned earlier that you started studying this in 2008 and I, I wonder how you would characterize your sort of work in this space and how you've seen the lives of kids change over this time period. Because we're dealing with a range of Technologies since 2008. I mean, my God. It's been a constant pace of change in the world, frankly. And in Hoyuzi, how do you think through that arc between now and then almost 20 years?
- [20:13]
B
Yes. Let me start with youth mental health, because it has changed a lot. So when I first started working in this field, the big concerns were around high rates of youth violence and drug use, substance use, teen pregnancy. We were working to prevent those types of costly issues among young people, and those have plummeted. We are at historic lows. So violence among youth is down. Substance abuse, alcohol use is at historic lows. Right. So we have the most educated generation in terms of high school completion. If we look at most markers of how young people are doing. There was a great article in the Scientific American a couple weeks ago that shows that young people today are actually doing better than us on almost every metric you can measure. And that is not what adults think. Right.
- [20:58]
A
I mean, that's a good news story.
- [20:59]
B
That's a good news story. But they're sadder. Right? So they're sadder. And if you look around in the world that they've grown up in since, you know, the aftermath of the Great Recession and opioid epidemic and, you know, Covid and school shootings have increased during this time. You know, guns have become the number one killer of children in America. And so if we look at the. The things that have shaped their young lives, then, yeah, they're more anxious, they're reporting higher anxiety about the world that they live in. So I would actually look at these indicators and say that they're incredibly resilient. I would not say that this is an anxious generation or one that has been destroyed. I would say it's one that is really kind of thriving on many metrics that matter, despite the chaos and the suffering of the adults around them. So that's kind of the first part of that. Just independent of the text, just on that, what's improved?
- [21:51]
A
What's the cause of that improvement on those metrics?
- [21:55]
B
Yeah. So the, you know, in terms of the violence decline.
- [21:59]
A
Yeah, violence and broader sort of decline.
- [22:01]
B
Exactly. I mean, I think multi. I mean, I think with any of these, like, complex behaviors, there's not one single thing. And I think that's where the social media, smartphone thing, it's so compelling because people can point to this one thing and it's something you can turn off. But there's been lots of changes. Right. You know, some of it's been prevention, some of it has been, you know, population, demographic shifts. Right. That have happened in the United States, some of it has been like greater investment in kind of young people, you know, early on. And so I think it's a mixed causal story. But the main point is that the story that we're constantly told about young people doesn't match the data. Right. It just doesn't. You know, there are these upticks in anxiety and depression and we need to pay attention to that. But we also have to understand that against the context of what they're waking up to every day. And I think that the thing that we've learned is that as much as the technology changes, many of the things that young people need remain the same. And why they go to the tech, what's driving them there, remain the same. So when they hit adolescence, they're going there to connect, to be entertained, to be with friends, to consume youth culture. And their use looks really different from adults and it actually looks healthier and better calibrated than adults, which is one of the other things, the other takeaway messages. So when adults make these rules and pass these policies, they often have their own use in mind, which is not exactly healthy or helpful healthy either. Right. Like doom scrolling. The demise of democracy is not why young people go online. You know, they're online and they're chatting.
- [23:40]
A
No, it might be why we do.
- [23:42]
B
Right. That is exactly what we. You're right. Yeah, maybe.
- [23:44]
A
Wouldn't want to assume, but I can speak for myself. What do you think the role of social media is in this and how hard? What do you think it's harmful?
- [23:56]
B
I think it depends on how it's used. Right. And so, you know, we see firsthand what young people are consuming. And it's like any content, unless you know how young people are seeing that, remembering it, interpreting it and carrying it forward. It's hard to make a judgment on any type of content. And I think that the thing is that this is what adults don't get either. Social media isn't what it was five years ago. Social media is now like TV was. It is just this stream of content. And young people don't engage with social media. All of the action is happening in these offline discord or chat groups and they'll pull in content and share it, but they're not out there on the open intranet creating all this content and interacting with each other there. That's not where the interaction happens and that's not where the things that people are really concerned about like cyberbullying or, you know, a lot of the social comparison happens either. Right. So that's happening in these more private groups that are, you know, kind of self monitored or mediated. I'm Erika Alini, I'm a reporter at the Globe and Mail. The economy, new government rules or how companies behave all affect what you do with your money. Why is it so hard for families to find affordable apartments? Why doesn't your pension keep up with inflation? Why is vet care so expensive? I ask the big questions so you can make better decisions about your money and the Canada you want to live in. If that's journalism you value, head to globeinmail.com subscribe.
- [25:45]
A
I want to talk a bit about these bans themselves. I mean, we really are in the center of this debate here and I think almost certainly there'll be some form of age restriction applied potentially within the next couple of months here. So I think it's just really front of mind in the debate here. So I think maybe we should start with what's happening in Australia because it is kind of ground zero for a social experiment. And people promoting bans like to say that social media is a social experiment, but taking it away is also a social experiment and one that a lot of people are actively studying. And it's very early days in Australia, but I think it's fair to say there's a lot of people trying to understand what its effect has been. What's your sense of what we know and don't know about its effect and what it's accomplished and not.
- [26:37]
B
Yeah, so I think, I mean it's a fascinating kind of case study because they quickly move to ban like most countries have. And it's actually we've done surveys here. It's the only thing that Democrats, Republicans and Independents agree on is they want it banned. Right. Like this is fascinating.
- [26:51]
A
Oh, the numbers are off the charts. Off the charts. I mean it is truly remarkable. And that has to be a part of this debate, right? Yeah. If you're a politician and you're hearing 80% of people want it, I mean you're going to do it if you're
- [27:02]
B
up for reelection are doing this. Right. And kids don't vote and it doesn't care if, you know, if 14% of kids want this or not. So the first part of this is they rushed to do this and then discovered that they didn't actually know what social media was. So YouTube was not on the list and a number of other like commonly used social media platforms. Right. So then they had to figure out, you know, the people who are making this, is this disconnect between the people who are Making the policy and people who are actually users of tech, of the technology and the way that kids do it. So they put it back in. Fascinating. In December they shut off the accounts. And it's when policies go out and I was a professor at a policy school for years and usually you have to go like four or five steps out to see how iatrogenic effects might happen, like unintended negative consequences. But on day one, what they did is they shut down, I think it was 2 million YouTube accounts. Now, this doesn't mean that kids could not be on YouTube. They could still access all the content. They lost their accounts, which means they lost parental mental controls, age based content, moderation, safety filters. So on day one, they took away a lot of safety features. And you can argue whether or not YouTube has, whether they worked or not, or good or bad, but you just took it away. And the same month that Sesame street moved to YouTube. Right. So that was not great. And now I want to also caution this. I agree that it's early days and these things need to play out and people are going to jump on early data and say all kinds of things.
- [28:34]
A
But, but on all sides, Education on all sides.
- [28:36]
B
Right. So I don't have the great authority, but you know what you can go forward is say, okay, well do kids actually. Did this actually move kids off the platform? And it turns out for the vast majority it doesn't. And I'm sure you've seen these reports also, Stanford is following a cohort of young people and you see that the vast majority have remained online either because they have other accounts or because they're able to kind of trick the age verification system.
- [29:01]
A
Yeah, I've seen a survey anyway, I think by the regulator that 70% of kids are still using these platforms. There's a bit of ambiguity there, I think, on how you framed it, on having accounts versus using. And I think it's worth pointing out that it was designed to not do both. Right. That the policy was intended to get at the features that were seen to be tied to having an account, like targeted advertising, certain forms of data collection, recommendation system, so on and so forth that may or may not be more harmful because they know a specific amount about you due to an account. Right.
- [29:36]
B
Yeah. But the other side of this is that you'll push young people into less regulated and less safe spaces. I mean, they're going to be there. They're going to be there and they're going to find a way. You know, if you do research in schools and you shut off tech they'll open up a Google Doc and they'll all go into it and create their own community there. So, you know, I think can you
- [29:52]
A
say a bit more about that because you have a new study that that's I think about to come out on this. And so how do you define unsafe spaces and why do these kinds of restrictions push people into them either on platform or on other platforms?
- [30:05]
B
Yeah, so I think young people will quickly crowdsource and figure out where everybody's migrating or moving. So do we not, do we want people to move to signal, which is where some of the young people were moving? Right. Do we want them to move into spaces where we can't with zero we
- [30:20]
A
should say safety features or regulations or any kind. Right. Because it's an encrypted space in which we have no idea what's happening within it.
- [30:27]
B
Right. And so I think you can start from the assumption that young people are not going to give this up easily, given how like enthusiastic and early adopters that they are. The other thing that's very clear and like we just know this from developmental science. And if you're a parent or if you just spend any time with young people, the absolute worst thing you can do is come down with a top down rule and ban for an adolescent. And I'm not saying that they get to do whatever they want, but if you actually want to engage them in behavior change, you need to engage them in the process. And the solution, which is not what is happening with all of these bans. So they're coming down, top down, you know, and we, we see from like, you know, years and years of research where we basically ban something, you make it more appealing, right. You set up conditions in which deception and getting around the ban becomes the issue. And I think we're, you know, we're seeing this, there's great stories out of Australia, right. In terms of they the solution is that you use computer vision to guess the age. But like, you know, I coach middle school basketball, you go to a middle school and try and guess the age. Right. Like this is a problem that computer vision cannot solve. Right. You cannot accurately guess the age of a middle schooler. Right.
- [31:38]
A
And so now I think on photos they think it's about a year and
- [31:41]
B
a half and it's racially biased and apparently drawing a mustache works or using your mom's license or like, you know, in fact there were stories of like the dog being a substitute, which means like the companies really aren't trying because a computer vision problem that is easy as a dog versus a human to distinguish. But so it just sets up a game of like, you know, deception and getting around rules. We had this in my own house yesterday where I got a call from the middle school to my daughter's phone had been confiscated because of a violation of a yonder packet. And she was actually chased the day before by one of the yonder ladies that they're paying full time to surveil the school into a bathroom to. And so it's become like this ridiculous kind of cat and mouse game between young people and adults when this is something that we could be solving together.
- [32:34]
A
Yeah. Do you see potential benefits of doing
- [32:39]
B
this of the bands?
- [32:41]
A
Yeah.
- [32:45]
B
So to begin with, from where I sit with adolescent mental health, the amount of predictive power that social media has is zero or close to zero. And it depends. So in terms of mental health health, no. And if anything, I think we should be investing in solutions. Young people are here. We should be investing in digital mental health solutions in these spaces and working to make them safer. So experiments we should be doing are how to design these spaces in ways that support children's development, because they are going to be there. So I think that would be the best use of resources and the amount of money we are currently spending and social and political capital on a ban based approach. I mean, you just cannot ban your way out of a youth mental health crisis. Like we, we should know that. We should know that. But getting back to your point, adults love this. So this might be good for adults. Adults might feel a sense of accomplishment. Adults might feel like they've collectively solved something in fairness in schools. I'm sympathetic. I'm married to a teacher. I mean, I, you know, if this makes a teacher's life easier, which I'm not sure that it does, then sure. But don't at the same time say that we're solving the learning loss crisis or the mental health crisis, because those data are pretty clear that that's not what's happening yet. And maybe it will in the future. Maybe it will, but we're spending a whole lot of money, time and capital moving in this direction and the opportunity cost to me seems too high.
- [34:22]
A
I guess what I'm trying to get at here is I hear you on the disconnect and the leveraging of science here and evidence, and I think that's a real concern in this debate, honestly. But I also hear people's level of what you called general frustration. And you admit it exists. Right? And I feel it myself. And I think that's why you get 80% of parents supporting this? Because everybody's feeling the same thing, that they would rather their kids have some general experience of childhood that's different than the one they're having. And they see their kids on social media a lot of the time and they think that's probably just generally not great for them. Isn't that a good thing if we solve that problem? Are we just not being honest about the terms of the debate? But if we were, maybe there'll be all these benefits from this. Maybe kids would spend more time outside. Maybe it would force us to spend more time with our kids as well. Maybe it does solve a lot of problems we're all feeling, not just kids. If we take away this causal relationship between specific harms that maybe been over leveraged in this debate.
- [35:20]
B
Yeah, so I think, I mean, I think that's a fair thing to say is that this is like our idealized version of like what would happen is if we just shut off the phone, that suddenly we would all just go outside and enjoy nature with our, with our children. But these trends actually started before the phone, right? And so people have blamed things like childhood obesity on the phone and those rate, those rates were going up because of them. What we put the sugar we put in food and the, the taking away of unstructured time. And so I think that there will be a little bit of a disappointment in kind of this simple and singular solution. But this is interesting because people have actually approached me and said, okay Candace, back off. This is our opportunity to slay the social media companies. Like, so what if there's no cause? But my response has always been, if you're going to use children as the tip of the sphere to slay social media companies, what are they going to get back? Right? So it is highly effective to weaponize stories about child mental health to get social media companies and other companies to change their policies. And I think that that is effective and that is being effective right now. I think, I think change is happening because we have weaponized stories and there have been actual harms. So I don't want to discount the
- [36:35]
A
actual harms weaponized and, and told and
- [36:38]
B
amplified and amplified these stories. Absolutely, Absolutely. Absolutely, absolutely. But my issue with that has been, you know, I'm not willing to turn to an entire generation of young people that are better than us and say, shame on you, what you are doing is addictive. It's rotting your brains. You have no self control. Right. This is causing mental health redirect. All of our resources that should be going to support their mental health to this war on social media companies. So like hating social media companies puppies feels really good. But we are gonna fail if we hate them more than we care about getting like this actually, right. For our kids. And so that's the hill that I guess I'm dying on right now. So I get it. You know, this is highly effective. And then I'll also say, and you know, I'm not saying that AI is all a moral panic, but when I was young, you know, Congress convened to set an age limit on Pac Man. They thought PAC man was rewiring our brains, you know, that we would have no inattention. Right. So they blamed strings of burglary on Pac Man. Right. So there is this history of moral panics where we do have to stop and say, wait, are we all charging ahead after the wrong villain here?
- [37:52]
A
Of course, the flip side of this is any benefits we might get from the ban also remove the positive benefits kids get from these. And expressing that there are benefits to social media can often them be tricky. What are they, what are, what are we taking away from kids if we do this?
- [38:11]
B
Okay, well let me, let me flip this. So I have an, I have an idea for how we could really improve the Internet. Right? So the group with all ears. Okay? So the. There's a group in our society that spreads the highest level of misinformation. And if we got rid of this group, misinformation on the Internet would go down I about 80 to 90%. Right. There's a group that is responsible for most of the sextortion online. And if we got rid of them, the Internet would automatically improve, especially for young people who need it the most. And there is a group with the worst mental health of any age group in our society. And that group is adult men between the ages of 45 and 65. And so I could propose that we could kick them off the Internet and improve the Internet for everyone. But everyone would rightly say that's grossly unfair. Right? What you're doing is you're cutting off opportunities for them to do all of these things, to work to get information, to connect with people, to, you know, the news, right. Like the entire world is now online. And so it would be grossly unfair to take those kind of rights to information, right. And access, you know, away. How would they tell these scary stories about our kids? Right? I mean, they wouldn't have the Internet. And so what we're doing is we're turning downward and we're saying we're going to take away this, what is now a public square, which is where young people actually get a lot of information. Every school has a social media account where they advertise their sports events, where they call out achievements of their students. Right. So at the same time, schools are banning social media. You know, my daughter's in the student government and they're walking around the school with a phone and posting on Instagram when the volleyball game is and how to fundraise for this specific event. And so it is where information is now shared.
- [40:00]
A
We don't need to be doing that kind of stuff on social media.
- [40:03]
B
Right.
- [40:04]
A
Like there's no reason my school needs to communicate. My kids school needs to communicate.
- [40:08]
B
Yeah. And so they'll do both, but they're gonna do, they'll do it on somewhere else, which is like social media. Social media is something where you have an account, your school would have to have an account and share content. So where would the school share that? An email know?
- [40:22]
A
Sure. I mean, it works pretty well for me,
- [40:26]
B
for our generation.
- [40:27]
A
Does it need to be broadcast publicly via an unsafe platform? Like do we need semi private messaging services like Snap being the default place where kids go to socialize?
- [40:35]
B
Right.
- [40:36]
A
Like, I think we've drifted into using certain tools that are not necessarily aligned, like you say, with the benefits we're trying to get out of them. Right.
- [40:46]
B
Oh yeah. And so I am with you that I am not endorsing the safety or the design features of all of these platforms, platforms. And I think that massive change has to happen for them to be better for all of us. What I'm saying is the distinction between making it better in general and making it better for kids is not as big as we think it is. Right. And so we should clean this up for everybody because kids are going to be there anyway. And the approach we're taking now, which is like kick the kids off, it actually holds tech companies less accountable. It gives them a path because the kids aren't there anymore. And I've had so many, I mean, people look at me like I'm, you know, all pro tech. I've been like yelling about child safety and data safety for a very long time and unequal access, like those are the issues that I care about. But the solution space right now where we think that we can solve it by kicking them off versus like making it better for everybody is, is just, I, I don't see how it works. But you know, I've been wrong many times, so. So this could happen again.
- [41:50]
A
Let's go to that thing, because that's what I want to talk about next, which is, look, we know there. And you've just named some of them, like design features that are harmful to adults and kids. In Canada, we're talking about a ban or a restriction potentially combined with a design based regulatory regime. Right. That goes at some of these. How do you see the solution playing out to those harms, to the harms that we know are connected to potentially the business incentives of the companies, the lack of responsible design of the companies, features that frankly shouldn't be allowed for anybody, but we know they allow them. How do we get at those problems?
- [42:31]
B
I mean, how do we stop companies from selling our data? Right.
- [42:34]
A
Of course, like David, a privacy law would be helpful, which we don't have a good one here either. So how do you see the solution set, I guess is what I'm getting at here.
- [42:41]
B
Yeah, so I think that the solution set is absolutely having more independent and kind of authority over the regulation of companies in terms of. And the design features, they're a little bit tricky because people want to separate them from content and you can't. Right. So the example is, you know, people talk about kind of autoplay or keeping people hooked, but is that design feature on its own problematic if it was just, you know, pictures of paint drying or, or us talking with people, you know, just constantly. So without the engaging content, these design features on their own, it's sometimes hard to see how they're harmful in themselves. So it's hard to separate it from content. And then you.
- [43:24]
A
But there is a relationship between content creation and design and the incentives of design. Right. Like we know certain kinds of content get created because it is incentivized. So there's a bit of a chicken and egg thing there, right?
- [43:35]
B
Yeah. And I mean m. People want to watch. The people are. It's engaging. Right. And so do you ban engaging content? So I, you know, I think we, we are in a tough, in a tough space where we have to have the companies come to the table and you either, you know, have a carrot or a stick. And we've, and I think we are in a space right now where the tech companies have so much power that the carrot or the stick is not going to work. And so people are frustrated. And that's why I think, you know, using children and in this fight has become so necessary. And what I would say is that there's, you know, there's costs to centering them in the battle. But, you know, the kind of safety by design is something that we should all strive for. But I mean this is a big, this is a big tent hard problem.
- [44:24]
A
But could I push that a bit more than maybe because I'm just so in the policy weeds of this. But like saying we should strive for safety by design is just unsatisfying to me. Right. Because of course we should strive for it and we should want it and prefer it, but it needs to be forced. We know that. There needs to be 20 years of watching social media. Right. So what do you think we should be building and doing and making platforms do? And how in this, do you look at digital regulators that are emerging in multiple jurisdictions and say, well, that is the broadly right path here we should be going.
- [44:56]
B
That is the broadly right path and benchmarking this and benchmarking it in terms of what's working and not what's working and also benchmarking. We also know that the thing we know is that young people are going to be the first people through the door on the tech. Right. They are early and enthusiastic adopters and we never designed for that with that in mind. Instead we come in later and say we'll fix it when it's broken and that's not good enough. And that's never been good enough. Right. And so embedding these, you know, toolkits and design frameworks that really consider development and this is why I think there's such a misfit with the bands, because you have to assume that they're there and they're using it to force the design that, you know, is accountable for that. Those kind of young users, if we put in a ban and they're bypassing it and they're all, you know, they're there anyway. We're not, we don't. We lose our leverage of forcing companies to design for the fact that they're there. Right. And we've been through this numerous times with the, the, you know, age 13 cut off and companies would say, oh, we actually don't save child data because you don't have children on our format. And we know from the logs of our kids in our studies and their reports that, well, 80% of them are there, so stop lying and let's fix it.
- [46:10]
A
Yeah, I mean, I do think it's important to acknowledge that a ban assumes they can never be made safe and design approaches assume the opposite, that they're not being made safe by choice and that we change those choices. But the other challenge, I think, I think opponents of the ban bump into maybe or the argument for design based regulations Bump into is that there are countries who are trying to do this. Europe broadly has one through the dsa. The UK does, France does domestically as well. In addition to the dsa, Australia has some version of this. And yet they are all also reaching for bans. So what was insufficient about those levers? Are they not not building them properly? Are they not the right policies? Or for someone who's like Canada who's trying to think of build this building regulator and then we say, well that's going to solve the problem. And then we look at other countries who have those and they're still going for a ban. What's the gap there that they're trying to address?
- [47:12]
B
I mean, it's a great question and I don't know if it comes back to this point that we started with, which is that a ban feels good and it feels like we're taking control even if we're not. Right? So even if it fails, it feels like we're doing something. It's so politically popular. Like you know, if you're running for any kind of office or school board, like this is, this is the first thing on your kind of mandate because nobody's gonna push back on it. And so I think separating that from whether or not it's effective or not, and I do think that the cost is not something that we can ignore. Right. So you know, the school based bands, it's about $30 a pouch la UN spending $7 million. There was $29 million earmarked in New York for the implementation of the ban, both between the pouches and the enforcement. And so, and these are sometimes taken from low income, low resource schools. And so I think we, it is a trade off. There is a cost to this, there is a cost to the ban, but it is very politically satisfying.
- [48:16]
A
So the change isn't happening fast enough. People are demanding faster change. Regulations take time, design shifts take time. Compliance frankly is not very high on the platform side. And compliance is therefore, let's leapfrog and
- [48:28]
B
just like, yeah, I mean I understand the impulse. I just practically, I think where I come at it a little bit differently is I have a front row seat both in my own home to my teenagers phones. But you know, following 3,000 young people around through their teenage years, you know, you see what the rhetoric is and then you see what they're doing day to day. And so where we've been really focused is that, that you know, we see this drift towards people embracing bans as, as a solution to these really hard problems. And there's going to be people on the front lines, and that includes, you know, teachers, educators, parents, who are going to be left with fewer resources when this isn't actually achieving the kinds of outcomes that you talked about. And so how do they.
- [49:14]
A
If they're mutually exclusive, right? If they're, if they're mutually exclusive, which they might be, but I guess not necessarily. So just before we close here, I just want to talk about something else that's emerged in this debate recently and certainly in Canada, in part because of some pretty tragic cases of harms we've seen inside AI chatbots and that have garnered a lot of kind of rightful attention. I think many, I would say, including many in the government, have started talking about, about including AI chatbots, both in the potential regulatory regime, but also in the ban. And I'm wondering how you view that, as someone who's watched different technologies emerge over the last 20 years, how do you view something like an AI chatbot? Do you see it as part of the continuum of social media and therefore we can talk about it in the same way?
- [50:04]
B
Yeah, so I think it is an evolution and I think it's going to be infused throughout all the main platforms. And so we've done some research funded by the National Science foundation that has looked at how parents, teachers and young people engage with conversational agents and LLMs. And we've just seen a rapid up take. And the thing that we know, the things that we know are that young people are going to be there first. And that is clearly the case. And they're actually outpacing adults in terms of their adaptation and experimentation with them. And the other thing that we know is that companies are really bad for designing with young users in mind. And so you have this massive conflict between they're the first ones through the door and companies are not thinking about it with any kind of developmental care. And so in this case, a safety first approach seems appropriate. And we have these horrible cases and these transcripts where you just look at it as, you don't have to be a psychologist as a human being and
- [50:59]
A
say, in what universe is this?
- [51:00]
B
Where are the safe safeguards here? What engineer did not put in? Where is the failure point? And there are many along this. And so I think people are right to say this is a chance to get out ahead of this. Now the question comes to what is the instrument we have to do that? Because I think there's always this balance between safety and innovation with the development of new technologies. And given where the world is, where the markets are, where the tech industry is that all the safety protocols are further and further behind and the push for innovation here is putting us into some pretty uncharted territory. So I think it will be tough to regulate around, but it is absolutely something that we should be, should be thinking about.
- [51:47]
A
I mean, we certainly don't. I mean, looking back at the pace of, we thought the pace of change around social media adoption was fast, but this is happening at a different level. And I think the idea we would wait 15 years or 20 years to get our handle on AI feels just totally out of step with the impact it's going to have on society. So I wonder how we speed up this process quite radically on the design side and the safety side side, or we're going to end up in this ban for sure.
- [52:14]
B
Right. And so we see young people with higher rates of mental health problems are more likely to engage with conversational agents and to seek advice. Right. And they're actually more likely to go online anyway to seek advice and help. And so this gets back to this broader gap that we know. Young people, when they're struggling, go online for help and they've been going online for other types of platforms and now they'll go to conversational agents also. So we can try and ban that or we can try and get in quickly and try to put in safeguards. You know, some people are very optimistic in thinking about kind of digital mental health chatbots and extending services and doing those types of innovative approaches. But you know, regardless of our approach, I think the thing we're going to also learn is from prior attempts is that the ban alone is not going to stop it. Right. And so if we try and ban young people from engaging with conversational agents, it's not going to work. So we're going to have to make them safer, but we're going to have to mobilize quickly to do that and
- [53:09]
A
we have to do it fast. Yeah, so it's going to. Yeah, I couldn't agree more.
- [53:12]
B
Yeah. And we have to have young people in these conversations because the other thing that we know is adults are really bad at understanding how young people use technology, why they use it, why they want to use it and like, and there's a lot that they can teach us about use. And that's actually a pretty good intergenerational bonding exercise.
- [53:31]
A
You know, it's interesting, we just finished an eight month project with. It was a citizens assembly of 118 to 23 year olds on how they want AI to be regulated or not. One of the more Striking things which took me by surprise, I have to say, is that they had a far greater support for a chatbot ban than for a social media ban, which is interesting.
- [53:57]
B
Well, they also, 1823, they grew up with social media. Right. And so anything we grew up with, anything we grew up with, we don't think is so bad. But what the next generation is going to grow up with is going to be terrible for them. And so this is, this is interesting.
- [54:09]
A
That's 100%. Yeah. So look, just to close here, I mean, I think, I think so many people gravitate to bans one because they seem like a simple tool. But also I think they imagine a future without social media and they remember their past and they think, well, you know, I probably wouldn't mind that for myself and my kids. Kids, right. That maybe that's a better future where kids are playing with each other more outside, more, or whatever it might be. And I think there's something very compelling about that narrative, clearly. What's your counter to that? What's your narrative of the future that is equally as compelling with social media?
- [54:48]
B
I think I'll come back to. First of all, the retrospective view is that if we actually went back to our childhood, it wasn't idyllic, right? And so we might be about the same age, but rates of substance use and violence and all accidents, I mean, they were much, much higher. Kind of actual risk for physical harm were much higher. And so a lot of what we do when we go back to, you know, I think about my free range childhood, and I'm not sure I'd want that for my children, the risks are, risks are pretty high. So there's that part of romanticizing how we grew up, which you look at all the data, young people, again, are actually doing much better, except for the anxiety, depression. So I think that we might have that a little bit kind of confused. I think, you know, moving forward, you know, we have choices, right? And we have choices for how we want to spend our time with our children. And the most disturbing thing that I see in our date and others is that the number one cause of conflict in the family is fights over screen time. Time, right. And so we know family conflict is not great for parent child relationships. And we're introducing a lot of that in this fight. And. And because we're scared, right, because we're hearing these stories that these things that we're letting our kids do are going to destroy their brains or lead them to extreme depression. And that is anxiety provoking for a parent. It causes stress and leads to a lot of conflict. So, yeah, like, we can, we can make choices like as. As for how we want to spend our time with our kids. And one of the things that I have to remind myself about is to not look at the phone at the tech first. Right. It's the thing that we didn't have growing up. But, you know, if I want to think about how my kids are doing, I think, you know, are they doing the things developmentally that they should be doing to be well, like, are they getting enough sleep? Are they enjoying school? Do they have friends? There's something extracurricular going on. And then the screen time part heart. Like, are there ways that I can engage with that with them? You know, that that brings us closer. Right. Because they're clearly drawn to it. Right. They're finding humor in it. They're finding connection. They're finding out information that maybe they're afraid to ask us. And so I try and reframe that a little bit as an opportunity for them to teach me about this new world that they're early adopters and they're, they're moving, moving through. And so I think that, you know, their world is going to look different than ours. And we don't like.
- [57:20]
A
Like that or we're nervous about it, right?
- [57:22]
B
We're nervous about it.
- [57:23]
A
And I think that's a human reaction, right, to be concerned about things you don't know and something fundamentally different.
- [57:30]
B
Yeah, I mean, I think it is. I think that we, you know, adults need to take a close look at their, their own relationship with technology, their own mental health and their own way of kind of being in the world. And, you know, when you put up that comparison with how young people are today, my money's actually on them. I spent a lot of time with them and kind of in research and life and coaching, and my money is on them getting this right and doing this in a better way than we have.
- [58:07]
A
Machines Like Us is produced by Paradigms in collaboration with the Globe and Mayo. The show is produced by. By Mitchell Stewart. Our theme song is by Chris Kelly. Our executive producer is James Milward. Special thanks to Angela Pacenza and the team at the Globe and Mail. If you like the interview you just heard, please subscribe and leave a rating or a comment. It really does help us get the show to as many people as possible. Machines like us to support by the max bell school of public policy at mcgill university. Learn more at mcgill ca maxbellschool machines like us is also supported by cifar, a global research organization based in canada. Explore their work at cifar ca.
- [59:03]
B
Hi, I'm Samantha Edwards, online culture reporter
- [59:06]
A
at the Globe and Mail.
- [59:08]
B
I write about the Internet and how it shapes our offline world, from TikTok trends and niche online communities to the rise of artificial intelligence and the influence of big tech. My job is to help readers make sense of this complex and ever changing landscape. If that's journalism you value, head to globemail.com subscribe.