Kim Mills: This past year, technology has been the glue that's kept many of our relationships alive. Birthday parties and business meetings moved to Zoom and we kept up with friends and acquaintances who we could no longer see in person with Facebook updates, FaceTime and group text messages. Now, even as life begins to return to normal, technology-mediated communication is here to stay. Although offices are starting to reopen, many workplaces plan to allow teleworking to continue, which will mean those dreaded Zoom meetings are not going away.
But even before the pandemic, we were already relying heavily on social media to keep us connected. Facebook alone has 2.8 billion monthly active users and 85% of Americans now own a smartphone. More and more, when we talk to each other, we do it with some kind of technology between us. So what does this mean for human communication? Is the version of ourselves that we present on social media authentic and truthful? Are people more likely to lie online or in a text message than they are in person? Do video calls exhaust us more than in-person conversations do? And perhaps more broadly, is all of this technology-driven communication good for our mental health and wellbeing or not?
Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association that examines the links between psychological science and everyday life. I'm Kim Mills.
Our guest today is Dr. Jeff Hancock, a psychologist and professor in the department of communication at Stanford University, where he founded and directs the Stanford Social Media Lab. Dr. Hancock is an expert in social media behavior and the psychology of online interaction. He is well-known for his research on how people lie and whether we can detect deception in texts, emails or in online reviews. His TED talk on deception has been viewed more than 1 million times. He's also studied how social media affects well-being and how artificial intelligence is changing the way we talk to one another.
Thank you for joining us, Dr. Hancock.
Jeff Hancock, PhD: Thanks Kim. I'm excited to be here.
Mills: So as I said, the name of your research group is the Stanford Social Media Lab. For a lot of people, when they think of social media, they think about Twitter, or LinkedIn, or Facebook or Instagram. You study those, but you also have a broader definition of what social media is. Can you explain that?
Hancock: Right, yeah. And in fact, I had a social media lab at Cornell before there was social media, so to speak, because for me, social media is any technology that we use to be social with one another. So I'm really interested in human-to-human communication, when there's some technology between us. But even more broadly, if you're talking to a robot, in a way I think of that as social media. When you're interacting with your car, that's a form of social media, and you and I here on a video conference to me is social media as well.
Mills: So let's talk about something that falls under that broader umbrella, which is video chat. Your research into Zoom fatigue has gotten quite a bit of recent media coverage. What exactly is Zoom fatigue? And why should we find it so exhausting to just sit in front of a computer screen and talk to other people?
Hancock: Well, that's exactly the question that my colleague, Jeremy Bailenson and I had, when the pandemic first started and us faculty members had all this extra time because we didn't have to commute. We weren't going into the lab, meetings were getting canceled left and right. But everything went to Zoom, and we found ourselves having these conversations about why would this be so tiring? We're here relaxed, we're in our homes. And so what we started to look at were the differences between Zoom conferences and face-to-face. And looking back, there was a lot of work in the '90s. My colleague here at Stanford, Pam Hinds, had done work on video conferencing and why they might be more fatiguing, but it's sort of, the research died out to some degree because it wasn't a huge deal. People weren't using video conference all the time. It was there if you wanted it, but most people didn't use it.
So what we started to talk about was this general feeling of kind of exhaustion or tiredness. Sometimes afterwards, Jeremy would say, “I just don't feel like doing anything afterwards.” Or I would say, “I don't want to talk to anybody after that.” So we found some sort of consensus on this idea that maybe Zoom meetings aren't necessarily more fatiguing than face-to-face, but have a kind of some unique fatigue. We think it's related to some of the non-verbal dynamics. So Jeremy laid this out in one of his articles really nicely. The fact that your face right now is really close to mine. So almost like at an intimate level and that's sort of physiologically arousing. Either we're going to be, come to blows or we're going to have a kiss, right? Like it's very unusual for us to be this close. When we're on with lots of people, even if I'm not speaking, it feels like a lot of people are staring at us. Something that we've called “hyper gaze.”
And to me then, one of the biggest ones psychologically is the mirror. So right now I'm speaking to you, I can see myself, in fact, my image is as big as yours. And while we've all seen ourselves in the mirror before, we've never really seen ourselves socially behaving. When we go into the bathroom and check ourselves out, it's usually just for appearance, but not for how we behave. This has a huge impact on us, I think and is very tiring. I've been trying to track myself just how much time I look at my image versus yours. And so it's hard to do this kind of introspection, but I'm guessing it's about half the time. My attention is drawn back to my image. So these are some of the examples of why Zoom could be so fatiguing.
Mills: Let's turn for a moment to your research on deception and lying. Your TED Talk that I mentioned a moment ago is called, The Future of Lying. What is the future of lying? How will it be different from its past?
Hancock: That is a really great question. For me, I was interested in how technology could change how often we lie, and then how we sort of end up trusting other people, kind of two sides of the same coin in a way. When we first started looking at it, we would ask people, where do they think people would lie the most? And what we found was this sort of folk theory that as soon as you couldn't see the other person, as soon as you couldn't hear the other person, lying would go up. And it’s sort of this non-verbal cue idea that if I can't see them, then they can lie more easily and therefore they will. And that’s wrong, it turns, out on a whole bunch of levels. And I think the most recent work that I think is the best on deception detection in general, is Tim Levine's, and it's summarized in his book, Duped.
And what we find when it comes to technology is his ideas from Duped work perfectly. So the first is that most of the time, most people don't want to lie, right? The best option usually is to tell the truth. And sometimes we're in positions where that's uncomfortable or awkward or could be harmful for us. And that doesn't really change when we go into technology. And I should also clarify that I'm not talking about the Russian spies or the scam artists, the prince from Nigeria or any of those folks. Instead, I'm talking about people that we know, people that we have a tie with.
And in that case, it looks like the psychology really drives lying, not the technology. So we found that people will be just as likely to lie in technologically mediated places as others, except there was one big difference. And that is, when the technology left a record. So in some of our earliest work, we found that people would lie the least in emails. And emails are one of these really recorded ones. So if I send you an email, not only do you have a copy of the lie I just sent you, but there's also ones on servers and everything.
So most of the time psychology is what's going to drive whether someone lies or not, not the technology, but there are some features. So if there's a record, people tend to think about that a little bit. Another one is whether we're synchronous or not. So when I'm face-to-face with you, there's only so much I can pause. Like if I pause more than a second, it gets kind of weird and you start to worry that maybe I'm having a stroke or something. And so we have to think of things really fast. And so lying occurs more often synchronously. Technology can give us a bit of a break on that. And sometimes we even see that people are more honest in those places.
So I think the future of lying is one that will be driven primarily by psychology. So do we have the need to lie? Are there other options? Also by some of these affordances of technology like, will there be a record of this or not? And so that will continue to evolve. Now in our latest work, what we're seeing is that people are getting really good at lying even in recorded media. And they're there, you just choose the kind of lies you can do. So for example, a student could say to me, “You know, Professor Hancock, my printer's not working.” And that's an okay lie in an email because the chance of me going to his place and checking out his printer is almost zero. But if he were to lie and say something about when he submitted something and then there's a record, I can check that. Then he's going to be in trouble.
Mills: Have you done comparisons among various media? Are we more likely to lie via text or in person or over the phone? I mean, how do we lie differently in different situations?
Hancock: Right. Great question. So we've done a number of these different kinds of studies. We use diary studies where we get people to record them. We've looked at log records, lots of different ways. And what we end up finding is that text and email typically have fewer lies than say phone calls. And it's because of that record. It's because we're not forced to say things really quickly. And it's sort of been hard to talk about it lately, because with disinformation, everybody just thinks that once you're on the internet, everything is a lie. And it is true that there is a lot of deception online, but I try to differentiate between what I call the known network, so people that we have a tie with and the unknown network, so that could be anybody from the Russian spy to just somebody trying to troll me, to someone trying to scam me.
And it's difficult sometimes to differentiate those. So is there a lot of disinformation online? Yes. It's a real major problem. But does that happen in between, say Jeff and Kim who have a professional relationship? No, very unlikely. In those kind of situations we see lying is more likely to be done on the phone, so vocally or face-to-face, than it will be in an email or text message.
Mills: What's been most surprising to you while studying the impact of technology online?
Hancock: Well, it was definitely the fact that people were lying the least in email, for example, one of our first studies. I remember, it was really strange. And then when we started thinking about it and thinking about this idea of the record, once you think of that, it's so obvious. But I think a lot of deception detection researchers up until the 2000s were really focused on non-verbal cues. This is the influence of Paul Ekman and his colleagues and students that we really focused on non-verbal things, and it made it into the pop culture. Now it's pretty clear that there aren't any really strong Pinocchio noses. There's no one cue that will always tell you if somebody's lying. So the fact that it happens online versus face-to-face, the non-verbals are just less important. And instead it’s some of these affordances of online, like the record that we leave behind.
Mills: There's this idea that we're all envious of each other these days. Not because people necessarily lie on social media, but because they present the best, most idealized vision of themselves, with touched-up selfies, or elegant dinners, or beautiful vacation photos rather than the day-to-day drudgery. Is the way that people present themselves on social media authentic or is that not authentic? Is that a big lie?
Hancock: This is one of the biggest complaints people have about social media, that it's all just like people showing their greatest stuff. And I guess I have two big responses to that. One is that, when we talk to people that we've just gotten to know, or we see on the street, or we see back in the hallway, we would not talk about all the crappy things or boring things that happened to us. We would say, “Hey, I just went skiing this weekend or I did that.” That is, we're always presenting a version of ourselves. And I don't think that's inauthentic. Instead, I think people are saying, “Here's what I want to project. Here's what I value. Here's me with my friends, or here's me traveling.” And so, okay. Is it a better part of themselves? Is it them looking the best out of the 50 photos they took? Yeah. But this is also like a sweater that I'm wearing that I think I look good in. I'm not here in my underwear.
Mills: I can attest to that.
Hancock: So is this authentic?
Mills: This is an audio podcast, but he's telling the truth.
Hancock: I am clothed. And sure, it's in the Zoom era, maybe we're coming more often in pajamas, but we still think about our presentation. So I think that we're able to do more kinds of optimized presentation online, but I don't think it's necessarily inauthentic. And my colleague, Sandra Matz, has done some really nice recent work showing that some people can present really authentically online. And it turns out online behavior and being engaged a lot, can be really healthy for them. High wellbeing. Other people report behaving online in a sort of inauthentic way, where they're saying, “Well, I did post that, but that's not really who I am or that's not, I only did that once.” And that can make them actually feel bad afterwards. So I also think we can move away from social media being all one thing or the other for everyone, and start to understand that if I behave in an inauthentic way that can have some negative ramifications for me, but if I'm behaving authentically online, it can be really powerful.
Mills: There's been a lot of worry in articles about whether smartphones and social media are addictive and whether they cause depression or anxiety, especially when we're talking about kids and teens. Are these worries justified?
Hancock: Let me start by saying, these are really prevalent worries. Any parent group I talk to, any professional group like educators, friends and family with kids, this is a serious concern. Because I think everyone sees kids, especially with the pandemic, on a lot of screen time. And even before the pandemic, my colleague Nicole Ellison and I were working on a book proposal about why we wanted to bring down the anxiety. One reason is there's not a lot of great evidence that using social media is automatically good or bad for you. Instead, our survey of the literature is that it depends much more on what you're doing.
So if you spend an hour connecting with an old friend or interacting with buddies on a video game or whatever you're doing on your phone, that can be really psychologically healthy. If you're doing something that's not, like stalking an ex or obsessing on something and using social media to do that, then that could be psychologically unhealthy.
So in our review of the literature—this is my colleagues at the Stanford Social Media Lab—we looked at over 200 studies, over 200,000 participants in all of these studies. And we can basically meta-analyze it so that you say how much social media use a person was having in the study and then whatever measure of wellbeing. And there's many types. We tracked six, things from like depression and loneliness, but also social connectedness and life satisfaction. When we did that huge, giant study, the effect size was R=0.01, which is essentially zero. It's a very precise estimate of zero.
So does that mean for all the parents out there like, “Hey, don't worry about it. You know, Professor Hancock says it's not a problem.” I don't think that's exactly what I would be saying, but I would say there isn't evidence for you to be really anxious and worried. Instead, what I would think about is, for your child and your family, how is this person's use of technology working, right? And so I have some friends, their kids are doing really well, they're thriving in the pandemic. They're learning all these computer skills and their friends are working with them to do better at homework and to stay connected. Others have really struggled. And I think this is another place we're getting where we need to move away from averages and start looking at individuals.
And there's really great work coming out of Europe, and this sort of, it's called the [inaudible] Wellbeing Project, where they're taking an n-of-1 approach. Where they're finding that some young people—indeed it is about 10%—find that using social media can make them more anxious for example. And there's another group, about the same size, that using social media can be really valuable for them, right? From a point like creativity, [inaudible] connectedness. And there was a whole bunch of people in the middle that it has no effect at all. There's zero correlation.
And so I think this is another thing where we need to start thinking about the individuals, what they're doing and how it's working psychologically in their life. And I think for parents, that's an important way to start thinking about it, instead of just how much time or how frequently they're using it, which is a kind of addiction model. Instead, think about, how functional is it? Is it working? What are they doing? What skills are they getting? Are they staying connected? And I think that's a more healthy approach to tech use.
Mills: So it sounds like you're looking at what people are doing when they're online. Are you looking at attention span? It seems that one of the things that's happening to us is that we're all over the place because we're distracted. You look at one thing, you've got an article, it's got five links in it. Pretty soon you've gone down some rabbit hole and you don't know how you got there.
Hancock: Right, Kim, I fully get that. I sense my attention changing as well, and that is a concern, but I kind of try to think of the longer arc of say psychological history. And a good story to anchor us is Socrates. And he was really adamant about a technology at that time, because he was worried about its effect on human memory, which to him was very much about the human soul. And that was the alphabet. He really believed that by writing things down, we would no longer remember them in the same way. And back there you'd have poets that would be trained to remember multi-hundred line poems, we don't get our kids to do that anymore. But I don't think anybody would say, “Well, let's get rid of the alphabet because we can't do 900-line poems.”
So yes, I would strongly believe that the way we are using technology, our immediate environment is changing, our brain is changing, our neuropsychology. And one negative outcome, I think, for me especially, is I feel like I'm pretty easily distracted, like doing 20 minutes of writing on one of my papers, I put a timer down now. So I just stay focused on that for 20 minutes and then I want to change it up. But perhaps that's allowing me to deal with a whole bunch of other kinds of information in our new kind of media environment. Perhaps there's some costs, but we're adapting and we have benefits there.
Mills: Are there other ways that we can incorporate social media into our lives that will maximize the benefits and minimize the harms?
Hancock: Well, I love the way you put that. I think that is the exact way to be thinking about it. And I don't have really high level advice other than, for each person, each family, to think about the degree to which this is beneficial for me and costly. There's been a number of studies that show that people that are really showing kind of problematic internet use, say overusing Facebook. They're often dealing with some other life stress, like they've lost a family member, they've lost a job, they're going through a divorce, whether it's some kind of financial issues. And so rather than thinking, what is social media doing to that person? It'd be more like, why is that person using it in that way? What are they trying to deal with? And for many people, it can be quite functional. It can be, I'm trying to deal with this, I'm trying to get through it. There are a small number of people, a small part of the population for which it's problematic. And it could be social media, it could be video games where they've just, doing it so much that it's interfering with other parts of their lives.
And so I like that approach that you just mentioned, which is how do you optimize? How do you maximize the benefits and reduce the costs? I kind of think of it like driving cars. Cars are far more deadly than social media, way more. And in the past were even more so, but we created new tech, there's airbags. We created regulations, you have to wear seat belts. We created new norms. You can't smoke in a car with your kids, but when I was growing up, that's what every parent did. It was not a problem. Now, if somebody tried to do that, that would be, not be okay.
So our tech changes, our policy and regulation changes and our social behaviors change. And it makes cars much safer. Are they perfect? No, but we're optimizing, we're creating the most benefit from the cars we can and minimizing the cost. And there's still work to be done. Most cars produce carbon, and that's not good for our well-being in the long term either, and we're working on that. So I think it's an ongoing thing, that people will need to just focus on the way they're using their tech to think about how is this beneficial for them.
Mills: So one of your more recent research interests is how artificial intelligence has begun to affect the conversations that people have with each other. This is really a really interesting idea that I suspect a lot of people haven't really thought about. How is artificial intelligence injecting itself into our conversations and what effect does it have?
Hancock: Right. I think it's fascinating too, Kim. So I'm glad you do as well. Most people think about, when we talk with an Amazon machine or Apple's Siri, that's not what I'm talking about. I'm talking about the way that we use AI to talk to each other, so AI sitting between humans. So when you and I were trying to figure out when we were going to have this conversation, my email would suggest, say three things. If you had said, “How about Monday noon?” It might say, “Sounds good, talk to you then,” or “No, I can't.” Very simple. When I look at them, they seem like, yeah, one of those would fit the bill.
It looks like there's about 16 billion of those messages sent every day, just through Google's platform. Now we see places like Microsoft that have introduced autocomplete sentences in Word. We're seeing it in text messaging, autocorrect, autopredict. And so what's happening is, when I go to talk to you in and of these environments, I'm given suggestions on what to say. And those suggestions often feel really right. But what we know from the way human language is produced is that, when I see language or hear language, it primes my semantic network space. So it actually activates some concepts. So that's how we can speak so quickly back and forth, is we kind of become aligned linguistically with what language we're reading or hearing.
So when I'm shown, “Sounds good” or “No problem,” it might feel right. It might feel like what I would've said, but I can't know because my brain has been primed to think those are reasonable things to say. And what's even crazier is, let's say you're using some system that never uses AI, never allows AI to be part of it. You don't have the options, but I use, “Sounds good. Talk to you then,” and send that over to you. Well, now you've been primed by AI, even though you have no knowledge of it, there's no representation of that, but your brain has been changed.
Okay. So does it matter? Well, it's at scale, so we're getting 16 billion and that's certainly a low estimate because that's just Google. What we've found is some work done at Cornell, by Malte Jung and Jess Hohenstein, and also we've replicated here at Stanford with Hannah Mieczkowski, finding that these AI systems tend to be overly positive. What I mean there is that, sounds good, sounds great, those are much more common in the AI-suggested messages than what humans would use. And we've done these in experiments, we've looked at the actual Google messages themselves.
And it makes sense. If you're a corporation you want to screw up that asymmetry effectively one way. You want to overdo the positive, right? And not more often say, “No, you're an idiot.” So that's smart on the creator's side. But when you have more positivity being injected into language at the scale of Google, it raises real questions about, what does that mean to our emotional tone? Is it appropriate? Is it going to change the way we're thinking? And then there are other big questions like, what kind of discourse are those systems being trained on? If it's around trust, for example, then we know that older, white, male discourse, right? The old Walter Cronkite type style of discourse is the most trusted in our society.
If the AI is being trained on that kind of language, then it's going to prioritize white, male discourse. And so there's real big ethical questions here as well when we're using AI to inject into human-to-human communication. And I think it's a fascinating area. It has a lot of potential for good, but I'm frankly a little alarmed at how invisible it is and how at scale it is.
Mills: Ha. It's like putting Prozac in the water, right?
Hancock: Right. Oh, that's good. It'll be the title of our next paper.
Mills: So obviously, as we've been chatting here, it's clear that technology changes really quickly. Is there anything that people aren't thinking about much now that you think will be the next big question?
Hancock: I think the two things I've been thinking a lot about for the future is exactly what we were just talking about. This idea of AI-mediated communication. AI being used to help humans talk to other humans. There's new tech out that technologists know a lot about, but I think most folks don't, which is called GPT-3. This is a natural language generation system. And it is truly impressive on the kinds of language it can write. So you train it up on whatever content area you're interested in and then it produces new content. I think it raises questions for disinformation. Now you have to pay people a little bit to get a bunch of disinformation out there. With something like GPT-3, you're looking at essentially infinite amount of disinformation for really low cost. So there's these kinds of things.
Also, if you're a young student and you're interested in marketing, for example, GPT-3 will be used to create massive amounts of marketing for a much lower price. And so if you're into brand management graduating right now, and you don't know about how AI can create language, you're going to be in trouble, I think. So those are things that are right at the cusp. And there's lots of things to be worried about, lots of things to be excited about.
The other brings us back to the Zoom fatigue that we were talking about earlier. It's pretty clear that most large companies that we've been talking to are interested in a hybrid model going forward. And there's lots of reasons for it, from cost to reducing carbon, all of these things. But companies are worried about culture, their kind of corporate culture. They're worried about bringing young people in and having them get known within an organization. So I think video is here to stay.
And one question is, how do we move from just this video where you're seeing into my living room, I'm seeing into your office, there's no real organization. If you were to come into my office at Stanford, you would see bookshelf and there'd be a desk. And we would kind of know how things are going there. We've just sort of opened our homes into business. And so I think there's a lot of things to be done here around VR. So how will virtual reality help us incorporate culture at a distance? We'll see changes in the way that Zoom conferences or video conferences are done. So I think we'll see big improvements in technology. And also kind of institutional changes. Let's not have Zoom or video conferences as a default all the time, let's use it when it's important or necessary.
And I think those are going to be more of a sense of, what's the right word? More of a sense of investment in people and their places, if we're going to ask people to stay home. And so we'll start seeing corporations, I think, invest in people's home offices and do it in a ways that will help support that culture, that they're very interested in keeping. So yeah, I'm excited for the future. I think we've opened up a new kind of way of working together and connecting that could be healthier for the planet, healthier for us individually, but it just doesn't have to be video conference all time.
Mills: I'm trying to figure out how we're going to come up with some sort of a hybrid model that works, so that when some people are in the office and some people are not, like what's happening with kids in school right now, right? Where the teacher is so busy dealing with the Zoom kids that she's not dealing with the kids who were in the room.
Hancock: Right.
Mills: And are we going to experience something like that when we go back to work?
Hancock: Those hybrid situations, whether it's some people in person and some people not are the most difficult for sure, because the ones that aren't there are a little less visible. And I think that's where we might see some good advances in technology. There's likely to be changes where I can put some glasses on and see another person who's not here right now, but the glasses make sort of visible or at least selling it, not perfectly, but at least will remind me that they're here, they're part of this conversation.
Mills: Well, this has all been really interesting. I appreciate you taking the time to talk to us, Dr. Hancock. It's been a pleasure to talk to you.
Hancock: Thanks, Kim. Really enjoyed the conversation.
Mills: You can find previous episodes of Speaking of Psychology at www.speakingofpsychology.org or on Apple, Stitcher, or wherever you get your podcasts. And please leave us a review. If you have comments or ideas for future podcasts, you can email us at speakingofpsychology@apa.org. That's speakingofpsychology, all one word, @apa.org. Speaking of Psychology is produced by Lea Winerman. Our sound editor is Chris Condayan.
Thank you for listening. For the American Psychological Association, I'm Kim Mills.
Jeff Hancock, PhD, is founding director of the 
