Loading summary
Transcript32 lines
- [00:00]
A
You're listening to TechTank, a biweekly podcast from the Brookings Institution exploring the most consequential technology issues of our time. From racial bias and algorithms to the future of work, TechTank takes big ideas and makes them accessible. Welcome to the Tech Tank podcast. I am co host Nicole Turner Lee, senior fellow and director of the center for Technology Innovation at the Brookings Institution. From headlines about how students are using artificial intelligence to cheat on assignments, to lawsuits alleging that conversations with chatbots lead to fatal outcomes, the relationships that teens are forming with AI chatbots have been the subject of much scrutiny. But even so, more teens think AI will have a positive rather than negative effect on their lives. And this is not the only finding that we're hearing about. And I'm so excited about our guest today because we want to talk about a lot of this research that came out of the Pew Research Center. We're finding that class and race have huge factors and major determinants when it comes to who is getting online to use AI. Researchers like myself are seeking to better understand what teens think about these things as AI becomes an even bigger part of their daily lives. And a recent Pew Research study, which I'm going to talk about today, found that most American teens say they're using AI chatbots primarily to search for information and help to do schoolwork and to a lesser degree, get this, use it for emotional support and advice, which I think is at the heart of many of these concerns. So this is not a monolith, and there's some notable differences that I'm going to talk about today. And I am joined by Monica Anderson, the director of Internet and Technology Research at Pew Research center and author of a new report on how teens in the US Use AI. Monica has a deep bench of experience in public opinion research and is published widely on topics like the digital divide, privacy, social media, among other things. Monica, thank you so much for joining me today.
- [02:08]
B
Thanks so much for having me, Nicole.
- [02:10]
A
Listen, I want to jump right in because I was really fascinated by reading the recent results from the Pew study on how teens use and view AI. And it's no secret, Monica, that this is a very timely topic right now. So the first thing I want to jump into, why this study and why now?
- [02:31]
B
Great question. I think one of the things that is central to the work that we're doing at Pew Research center is to study emerging technologies. And there is nothing more emerging and top of mind than talking about AI. At the same time, we also know that teenagers are often the first ones that are adopting these new and emerging techs. And for us to be able to really understand how teens are navigating this full digital environment, we have to make sure that we're putting down those markers right now when it comes to AI adoption. And I think one of the things that we really wanted to do with this work and why I think it's so important is to make sure that we're grounding conversations in data. It's very important to learn about people's experiences, to bring in anecdotes. But we also see, see that there's a lot of value in doing these scientific representative surveys for us to get a better understanding of what this means for all teens in the US So
- [03:37]
A
I want to go a little bit deeper in this. Right. And just really start with the surprises. Right. So let's talk a little bit about what surprised you in the study, and then we'll go into some more of the study's findings.
- [03:51]
B
Yeah, for sure. I mean, I think one of the things that really stands out is just the diversity of reasons that teens are gravitating to chatbots. So you've got a majority are using them to search for information, but you've also got a little over half that are using it to get help with schoolwork. I really found it fascinating that you've got about 20% of teens saying they use it to get news. And while there's smaller shares that use it for more personal reasons, 16% are using it for casual conversations, and about 1 in 10 are using it to get emotional support or advice. Another thing that really stands out to me is that teens experiences are not a monolith. Right. Not all teens are using it or using it in the same way, but there's some really interesting differences and patterns when you start to drill down into race and ethnicity. And so across many of the reasons that we asked about, whether it was for getting help with schoolwork or summarizing content or getting news or emotional support, you see, in fact, that black and Hispanic teens, and at times black teens, standing out even further when it comes to using chatbots for this use. So I think it's really important, as you said at the top of this program, that it's for us to drill down into those experiences because they often match up with what we've seen in social media and Internet use more broadly.
- [05:25]
A
Well, that's so interesting to me, and I want to come back to the demographic differences that the study revealed, and I want to sort of stick on this, like, competence that young People are basically saying that they have. Right. We as adults or parents sometimes think that, you know, we have to train young people on how to use these tools. But I think in the study you saw about a quarter of teens describe themselves as extremely or very confident in their ability to use things like chatbots and had high awareness. I mean, how does this compare with some previous work that the Pew Research center has done on AI literacy? Are we seeing this as a growing trajectory where it's just going to get higher among young people and maybe the parents will be left behind? Right. Or is this something that is just sort of a trend or spike in trend that's happening now just given the prevalence of AI?
- [06:10]
B
Yeah, I mean, I think what we're seeing overall for teens or adults is more people are familiar with AI in ways that they weren't four or five years ago. And teens are especially know about AI. 95% say they've heard of it. Nearly 60% say they've heard a lot. So that means that these conversations that are happening around AI are also happening amongst teens. And I think that's really important and interesting. The other thing that we wanted to do with this work is also understand whether or not teens feel comfortable with using these technologies. And as you point out, about 1 in 14 say that they are highly confident, and you've got about another third saying that they're somewhat confident. So teens, by and large, are feeling pretty confident when using these tools. That's not always the case when we've asked similar questions compared to older Americans. But I think that also just points out of how pervasive these tools are already in the lives of teenagers and even more so when compared to kind of older generations.
- [07:21]
A
Well, that's interesting. And I want to point out for the viewers, again, we will tell you where you can get access to this report. But one of the things I found to be interesting in terms of AI's impact, you know, as a scholar that is working on, you know, more of the humane impacts of AI, teens were more positive about AI and how it was going to impact them, about 36% to 15%. That was on the personal side. But then there was a little mixed reaction when it came to society, 31% of teens thought it was gonna have a positive impact, 26%. And then I think later, you basically ask a subset of those questions, and you see that teens are very, very much aware that AI is going to have an impact on society negatively. And there was a question I was kind of tickled that I do have Some friends out there that are young that understand this. You know, whereas 31% thought that AI was going to have a positive Future, there were 26% of teens that did think that there was going to be, you know, misinformation, loss of human factors, environmental impacts. Was that surprising to you as well, Monica, that teens are a lot more astute and not just the literacy on the tools themselves, but they kind of understand where this is fitting into society?
- [08:33]
B
Yeah, you know, absolutely. And I think, you know, we need to make sure that we're giving young people credit when it comes to their views and experiences with technology who can better tell us what they think AI is going to impact young people than young people themselves. So I think this is just something that we've seen throughout all of our work, that teens have really insightful and thoughtful things to say when it comes to the way that technology might be impacting them and society. And perhaps in some ways that us older folks don't quite get yet. And I think, you know, there's a pattern that we've also seen where people are more likely to think about the negative impact of tech technology on society than on themselves. Right. It could be, you know, you're thinking about your personal, positive ways that you can use this technology, but when you step back, you think about, oh, here are some of these societal things that might be happening. And so to your question, on understanding how teens are thinking about AI's impact on society, for those who said that they think it will be positive, you know, they really coalesced around three things they felt like. The reason why they felt this way is because it's going to make life better or easier, that it's simply good for learning and getting information, and it's going to lead us to be more productive. Now, on the other end of the spectrum, for those who said that actually this is going to be negative for society, there's two things that really pop. One is this concern about losing critical thinking and creative skills. Right. That was by far the most commonly cited reason among teens. And then the second most cited was around job loss. And that really is in line with other work that we've done, especially with adults, where there is a real fear about what AI might bring to jobs. Teens also mention things like misinformation, which I think is a really important part of this story and connects to what we talked about, which is AI literacy as well. And so I think it shows that, you know, teens point to a lot of different reasons, but these are the ones that were kind of most salient in their answers.
- [10:51]
A
And I think that's an interesting conversation that people need to have. Right. Because we as policymakers are, I guess I'm speaking for myself based on who's listening. Right. Oftentimes we think that including the person who is most affected by the policies that we're making, it's something that we don't often do. Right. And so in this case, I think your study really informs the views that young people have kind of made up their mind or have some perspective on this, and one in which we should actually pay attention to. And I'm saying this especially as more and more adults are taking the helm when it comes to privacy legislation or banning chatbots. I think your research for me was very insightful that, you know, we need to really think about how others on the other side really feel the users, as opposed to our own ideological or, you know, values, concerns. Which brings me to, like, this chatbot use. I mean, there is a lot. Let me say that again, Monica. Right. I mean, there's a lot of discussion when it comes to AI chatbots in particular. You know, we have read in the news about the impact of chatbots when it comes to really fatal consequences for young people, particularly chatbots that take on emotional dependency issues. We also know that there's concern about this general mantra that robots are going to take over our human experience. I'm curious from you in the report, there are various chatbot use cases that you talk about and the frequency at which teens actually use them. I want to kind of stay here for a minute because I think your findings also seem to speak to, I think, the concern that parents have with this dependency on chatbots. So where, where are the chatbot use cases? Like, let's clarify what you found in
- [12:31]
B
your reports, for sure. And you're absolutely right. There's a lot of conversation and concern about the way young people might be using chatbots for companionship or personal uses or emotional support. And what we see in our work is that while it's not the most commonly used reason why teens are turning to chatbots, it's completely absent. Right. So you've got nearly 20% say that they are using it to have casual conversations, and about 1 in 10 say that they use chatbots to get emotional support or advice. And one of the things that's unique with this work is that not only do we ask teens a set of questions, we also ask their parents. So we wanted to understand, okay, what are parents comfortable with their teens using chatbots for? And those two Things, casual chats, emotional support, advice. We're at the bottom of the list, right? And so in many ways, there's things that parents are comfortable using with teens using chatbots for, but that's where they draw the line. And I think one of the things that I would want to stress with this work is that, you know, this is the first time we've really done a deep dive on how teens are using chatbots. And this is a really important kind of first marker. As we continue to do this work and we do annual surveys of teens, it' to be really important to watch those numbers, right? Will they stay the same? Will they change? Will they go up over time? And that can also let us better understand what those experiences are. And another thing that I know we've chatted about, and I know you're kind of super keen on, is those demographic differences as well. And so even when we take getting emotional support or advice in this kind of series of questions, there's some really big differences by race and ethnicity. So about 20% of black teens say that they use chatbots to get emotional support. For white or Hispanic teens, that number drops to about 10%. And so I'm really interested to be able to track this over time to see if this pattern, you know, is this the same pattern we're going to continually see? But I think it also speaks to this broader conversation, is that, you know, teens are using these chatbots for a lot of things, but it's not a one size fits all. And in our work, we're seeing some really consistent patterns by race and ethnicity.
- [15:08]
A
And that is interesting to me. And I want to kind of jump into the race and ethnicity question, if you don't mind. I mean, I was quite surprised from the study about the differences in demographics when it comes to AI use by young people. And there seem to be some gender differences, like boys are more positive than girls about AI. But what was more striking for me, Monica, was like, black and Hispanic teens, I mean, they were more likely to say they use chatbots for schoolwork, which is so interesting to me, because on the other side of this mountain is this conversation around AI use of cheating, which, you know, for black and Hispanic teens sort of places another mark on their backs when it comes to their performance in school, which I talk a lot about. You know, we cannot assume young people. You know, I tell teachers all the time, the best indicator to know that a child used a chatbot is if you read the whole essay, and it was supposed to be about polar bears in Antarctica. And now it's about Colombia or someplace else because they didn't proofread it.
- [16:10]
B
Right.
- [16:10]
A
We can't always assume that AI is essentially the indicator of whether or not somebody's using it for a crutch. Tell me a little bit first on why black and Hispanic teens, you think are using chatbots more regularly. Like, what did you find? And then maybe we can extrapolate a little bit.
- [16:26]
B
Yeah, sure. And I think one of the things that really stood out to me when I was looking over this data was how much it reminds me of other findings that we have about tech use. So if we take our question about asking teens, how often do you use the Internet? And so that's something we've been asking for a very long time. And every single time we've asked that question, we have seen racial and ethnic differences. And so there's a lot of conversation about screen time, but there's not as much conversation about how screen time differs across those different dimensions. So about a quarter of white teens say that they use the Internet almost constantly. When we ask black and Hispanic teens that same question, that number doubles to more than 50%. We also see some consistent differences when it comes to using social media. So whether it's TikTok and even how often teens say that they use TikTok with kind of black teens standing out for using TikTok more than other groups. And so this has been a consistent pattern that we've seen, and this is just another kind of marker in that pattern as well. One of the things I'd also point out is that, you know, as you mentioned, black and Hispanic teens are more likely to say they use chatbots to get help with schoolwork. And one of the things I found really interesting is that as researchers, we want to do our due diligence. Right. And there's times where racial differences might be driven by things like income or other factors. And in fact, we're not actually seeing that here. Right. So there's substantively a difference in how teens are approaching this by race and ethnicity that's not being driven by household income.
- [18:17]
A
Well, and that's interesting, right. In terms of other income differences, I mean, you also mentioned 20% of teens and families earning less than 30,000 a year say AI helps them if you just take out black and Hispanic or you intersect those qualities. But when you get to the families earning $75,000 and over, that number drops to 7%, which is so interesting to me because it also, you know, Monica, we've known each other for a long Time on our digital divide work.
- [18:44]
B
Right.
- [18:45]
A
I mean, it could indicate a couple of things, but why is it that higher income families drop to 7% and like where the other data points that you found to be interesting? Because I think I saw the same thing when it came to the use of chatbots. On the emotional side, the higher the income, the less likely there was sort of parent promotion of these tools.
- [19:04]
B
Yeah, I mean, I think one of the things we want to be able to is kind of continue to look at those questions and see if those patterns arise. But I often think about the way that we talk about the digital divide today might be very different than the way that we talked about it 10 or 15 years from now. You know, one of the things that we didn't do in this work, but I think will be interesting over time is right now, you know, a lot of these tools are free and available. Right. And so what does that mean when maybe having to pay for these tools come into play? Like, are we going to see those differences? But I'd say to your point on the income differences is, you know, when we look at other forms of tech use, income for teens is not that big of an indicator compared to maybe that it once was. Right. So smartphone ownership majorities of teens, regardless of income, have a smartphone. Right. Regardless of income majorities are using social media. What we're seeing here, especially on the school, schoolwork question, is one that we don't have like definitive answer as to why we're seeing it, but it's definitely something that we're going to keep a watch on because when it comes to income, there's not a lot of differences when using chatbots, but you do see it with how often they're using it for school work.
- [20:31]
A
Well, and it's also interesting, I mean, I think some of the concerns that parents are expressing when it comes to AI among higher income families, I would think. And I'm working on a piece on this. So I was so happy when your data came out because I was trying to find data and you gave me some good data points that as part of a blog, if you're listening and you follow the Tech Tank newsletter, I do have a piece coming out on this that sort of incorporates many of the elements that we're speaking about, but was also interesting to me is like maybe higher income parents have the resources for their kids to do other stuff, you know, and so maybe that could be a part of it too. Monica, you know that they are able to do the type of literacy training because maybe they're more exposed to AI in their workplaces. I don't know. You know, seems to me another study for Pew.
- [21:16]
B
Yeah, well, well, on that point, you know, we are able to look at what parents find acceptable by income. And there's some ways in which actually higher income parents are more okay with their teen using chatbots for. That's true when it comes for searching for information or entertainment. And there's not a lot of income differences. When asked about whether or not you would be okay with your teen using chatbots to get schoolwork, the one area that I thought was super interesting here is that that kind of flips when asked about getting some emotional support or advice there. It's actually lower income parents that say that this would be more acceptable to them when compared to higher income parents. And there's also some income differences in terms of. About talking to your teen about chatbots. So higher income parents are more likely to say they've been having these conversations with their children as well.
- [22:15]
A
Yeah, and I think that's, you know, part of the trend that we've commonly seen in digital divide research where higher income parents just have more resources available. Maybe it's, you know, their ability to sit down with their kids and talk to them about these tools because they see them more prevalently in the workplace or they follow the news. Or maybe it's because higher income kids come from families where they have choices. You know, I, I really am taken by what you just mentioned in terms of young people in lower income communities using a lot of the free stuff. Free stuff people. Right. Like Monica. I mean, I can only imagine, and maybe it didn't come up in the study, that maybe there are higher income students that use more of the paid stuff potentially, and so they don't show up in some of these other studies. I mean, I really have a question for you and I've got a couple more questions. I don't want to keep you, but it's like, how do we reconcile then if most of these kids are basically saying, hey, hey, teacher, we are using chatbots to do our homework, how do you reconcile that? Because you basically let the cat out of the bag among educators who are violently afraid that they're reading AI generated papers?
- [23:28]
B
Well, I mean, I think the first thing is just how quickly this is evolving. Right. Some of my colleagues, maybe two, three years ago, did a survey of teachers and had a question about AI, and even then teachers were concerned about kind of how AI might be used. But these tools are evolving and changing so quickly. And right now as schools are trying to create policies, you've already got more than half of teens saying that they are using it and they are using it for schoolwork. And so in a sense, you've had, you know, not only are parents maybe having to play catch up, but so are educators and administrators. The one thing that I do want to point out is that we also asked teens about whether or not these tools had been helpful to them for their schoolwork. And you see that, on balance, far more teens say that these tools have been helpful for them than unhelpful. And even getting back into some of those demographic differences that we talked about, black and Hispanic teens are more likely than white teens to say that these tools have been incredibly helpful to them when helping complete their schoolwork. So not only do you have these demographic differences in using these technologies, many of these teens are saying that these tools have been very helpful for them and their education.
- [24:54]
A
Yeah, I love the way you said, I mean, low. If you're an educator, the cat is out the back. And you're so right. We are creating policies that may be about, like, banning or not including them in the classroom, but really, teens are picking up the tricks of the trade when it comes to how to use these tools. And it makes sense. Right, because many jobs in the future are going to be requiring the use of AI tools or, you know, may be replacing them. You know, unfortunately, when it comes to AI, so it's important to know where they fit on the food chain. You know, just one last question. I mean, this is not your first rodeo when it comes to AI and public perceptions. And again, I'm so proud of you. Many of you don't know, but a quick shout out, no. Monica, for almost 20 years now, 15, 20 years, you previously led reports around America's perception of AI. Are there differences in the results that you're getting from teens versus some of the previous work you've done in terms of public perception on AI, or are you seeing these opportunities and potential harms sort of consistent among both groups?
- [25:54]
B
Yeah, there are some clear differences that we see compared to teens and older groups. One, teens are just using these tools more than older Americans do. And that's, you know, no surprise when you look at social media or other kind of emerging tech. This is something that is uniquely positioned. And in a way that kind of young people are often the first ones adopting these new technologies. So we see that kind of adoption and use is definitely already widespread among teenagers in a way that we don't quite see amongst older groups. The Other thing is that teens tend to be a little bit more positive about AI when it comes to older Americans as well. But at the same time, I will point out that one of the things that is really important for this work is we often, when it comes to attitudes, we give our folks that are taking our surveys a chance to say they're not sure. Right. So, like, what impact do you think AI will have on you over the next 20 years? About 20% of teens said that they weren't sure. And that's something that we see across our work. Right. In that sense, a level of uncertainty is a substantive answer. Right. So even when we ask parents about would it be okay if their teen uses a chatbot to have a casual conversation with, you've got about a quarter of parents saying they're not quite sure what the right way is to go here. So with any kind of fast moving technology, I think it's also really important to make sure. And what we've seen is a pattern of people might be a little bit more positive here, a little bit more negative there, but in a lot of ways, people are still trying to make up their mind about what they think these impacts are.
- [27:42]
A
Well, I think this is all great stuff. I am looking forward to the next report because I know you all are planning for something else. That's how Pew does it. And I think these results are really informative as to what the future of this might look like. And I think it's a really eye opener for those of us that have not heard or factored in the critical voices of young people. Monica, thank you so much for joining me on this discussion today.
- [28:04]
B
Thank you so much, Nicole. It's always a pleasure.
- [28:07]
A
No, I mean your work is really helping us better understand the diversity of perspectives on AI use, chatbot use, and we're just really looking forward to more of it. For listeners at home, you can find the report where Monica, let them know where they can get this wonderful report.
- [28:20]
B
You can find all of our reports, including this one, and that's p research dot org and keep watching that space. We're going to have more on AI. Bye soon.
- [28:29]
A
Pew research.org please explore more in depth content on tech policy issues at Tech Tank on the Brookings website, accessible@brookings.edu. your feedback matters to us about the substance of this episode. So leave a comment, let us know your thoughts, suggest other topics or just like it or share it, because that's how we build community. This concludes another insightful episode of the Tech Tank podcast where we make bits into palatable bites. Until next time, thank you for listening to Tech Tank, a series of roundtable discussions and interviews with technology experts and policymakers. For more conversations like this, subscribe to the podcast and sign up to receive the Tech Tank newsletter for more research and analysis from the center for Technology Innovation at Brookings.
- [29:22]
B
IT.