Latest / Chatter that Matters / What happened to the Truth? - Gordon Pennycook
Transcript
- 0:04I've got an interesting guest today. His name's Gordon Pennycrock.
- 0:07Small town, Saskatchewan boy. Spent, I guess most of
- 0:11his time in school just trying to crack the class up.
- 0:15I was always fascinated with psychology. I guess he was the
- 0:19friend you went to when you had a problem. And he went on
- 0:23and started undergraduate in psychology,
- 0:27professor, third year, said you have an aptitude for that. And sort of the lights
- 0:30went on. Gordon started racing down this path,
- 0:34graduate degree, associate professor at Cornell. And he's really
- 0:38trying to understand us
- 0:42and why we're inherently trusting as a human race. If someone
- 0:46tells you something, you don't just immediately reject that
- 0:49as being untrue. Like it's. It would be kind of like normatively inappropriate to just
- 0:53completely question everything that someone says to you. But often lacking
- 0:57the skills to discern what is
- 1:00truthful and what is not, what is right and what is wrong, what is
- 1:04fiction and nonfiction news that I should share and news that I
- 1:08should ignore. Getting people a little bit more aligned with the truth
- 1:12will make them more discerning. We're not giving them new information. It's just getting them
- 1:15more sensitive to the fact that they might be making bad decisions because
- 1:19they're not focusing on whether things are true and they should be. And that's what
- 1:22he's spending a lot of his time on. I think you'll enjoy this
- 1:26conversation because it is about you and I and it is about the fact
- 1:30that we are under attack. That social media has given
- 1:33incredible powers to what he calls the elite and the people in
- 1:37charge of politics and how they encourage
- 1:41us to lock ourselves in these castles with like minded people and like minded
- 1:45content, validating each other and finding purpose in
- 1:48loving what we have to say and hating what the others do.
- 1:52It doesn't benefit a political oligarchic society
- 1:56with very wealthy political elites that are trying to control things, to have an
- 1:59educated masses, not just educated with knowledge, but educated
- 2:03for how to think and to question things.
- 2:10Hi, it's Tony Chapman. Thank you for listening to Chatter that Matters, presented
- 2:14by rbc. If you can please subscribe to the podcast
- 2:17and ratings reviews. Well, they're always welcome and they're always
- 2:21appreciated.
- 2:25Gordon Pennycrook, welcome to Chatter that Matters. Hello
- 2:28Gordon. I'd love to just claim some Canadian roots here. I know you're obviously
- 2:32an associate professor at an Ivy League school at Cornell, but just talk a little
- 2:36bit about growing up in Canada. Well, I'm from a small town called
- 2:40Carrot River, Saskatchewan. My dad wanted me to work in the pipeline in Alberta.
- 2:44In the oil fields. So I went into college without knowing anything about how anything
- 2:48works. And eventually I kind of get lucked into this third year psychology course
- 2:52where a professor just said, hey, you seem to be, have some like aptitude for
- 2:55this. Which I, no one's ever told me that at any point. And he suggested
- 2:58I apply for a, what's called an undergrad summer research award to the
- 3:02federal government in Canada. I got the award and I got into the lab and
- 3:05that's how I like actually started doing actual research. And they were doing research
- 3:09on human reasoning. And I actually was able to for the first time work in
- 3:13an actual like academic. Like I wasn't doing labor. How important was
- 3:17that third year professor that sort of knighted you and said you have an
- 3:20aptitude for this, that made you realize this could be the, the
- 3:24career path you were looking for? Jamie Campbell, the late Jamie Campbell.
- 3:28You know, in high school in a small town, you get kind of pigeonholed into
- 3:32a particular kind of lane. And even in first year, first
- 3:35like couple years of undergrad, like I had decent grades, but they weren't great grades.
- 3:39But then when it actually came to the actual research, I thought, this is something
- 3:42that I can actually be good at. And then being pulled aside and have someone
- 3:46actually validate that, that was, yeah, that was, that was pretty remarkable. As I
- 3:49started progressing, I went to grad school at University of
- 3:53Waterloo. And one of my favorite stories is from my dad, who was
- 3:56very supportive. So I was defending my PhD, and
- 4:00it was very theoretical. It was about what are the cognitive
- 4:04mechanisms that lead people to what triggers people to reflect and
- 4:07deliberate. I had made no attempt at all to
- 4:11make any argument for why it's important in the world or consequential.
- 4:15I just said this is, it's just for the sake of knowing what happens when
- 4:18you defend your dissertation is like, you know, there's a, you do the talk and
- 4:22then there's a bunch of questions and then you leave the room and then they
- 4:25deliberate as to whether you get your PhD. And so my parents were there
- 4:28and my dad came up and he said, yeah, that was really interesting. What's it
- 4:32for? You know, what do you do with it? Because, you know, he,
- 4:36he was a farmer, he worked at a sawmill. Spending all this time,
- 4:39six years and you. And the combination is just me talking about something that
- 4:43has no apparent bearing on anything. Seemed kind of surprising. Eventually
- 4:47when I started doing work on misinformation, he started understanding why it's of use. But
- 4:51at the time it didn't make any sense. I hadn't heard about the class
- 4:54clown, but in some ways this irreverence seems to come through. I mean,
- 4:58one of your. I think the early name you made for yourself was
- 5:02tackling what you call bullshit. You know, phrases that sound profound but mean
- 5:05nothing. Was that the beginning of you defining your brand within
- 5:09the psychology space that you're. That rather than getting caught up in
- 5:13terminology that only a handful would understand that you would start
- 5:17tackling terminology that everybody could embrace?
- 5:21Absolutely. I mean, I think the kind of irony I guess is that even though
- 5:24after I had defended my dissertation that was very theoretical and really was of
- 5:28no consequence to anybody, I had for a long time
- 5:31really been focused on trying to do stuff that would actually matter or that could
- 5:35intersect with what people understood about how the world works. And that was not so
- 5:39densely theoretical or non consequential that no one would ever care about it outside
- 5:43of academia. And so then I started doing work on stuff like religious belief
- 5:47when I was in grad school. And the bullshit thing was that's. Yeah, absolutely. I
- 5:50mean you, you, you nailed it. Like it's. I don't spend a lot of time
- 5:53worrying if people think I'm using the wrong terminology. I try to make it so
- 5:56that it's accessible to the broader public and not just other academics.
- 6:00The irreverence part comes through. I might have a bias on this,
- 6:03but I see the academic circles really living for
- 6:07affirmation from other academics. I mean they, you know, they live for
- 6:11their citations and they live for their. The nods and that
- 6:15you don't seem to care about that your approach is much more, as you
- 6:18said, this intersection between real life and how people think,
- 6:22feel and act, has that caused you any issues
- 6:26within the academic world or people just appreciate your
- 6:29sense of being granular and more honest and transparent?
- 6:33I actively try to not be influenced by the
- 6:37prestige game and the like. The competition that people have, that's
- 6:41inane. And you can't ever win the competition because no one's smarter. Like no one's
- 6:44going to produce more than Isaac Newton did. I just try to do good work
- 6:47that I think is interesting and is important. I think there are cases in which,
- 6:50if I was in my early days more focused on
- 6:55the game of academia of like trying to publish in the best journals instead of
- 6:59just getting a paper that you think is interesting. I might have gotten to an
- 7:02Ivy League school earlier or something, I'm not sure. And by the way, it was
- 7:05never a goal of mine to reach an Ivy League school. Anyways, I just wanted
- 7:08to get gainful employment so that I can do the stuff that I love doing,
- 7:10which is the job. And then getting into place like this makes it easier for
- 7:14me to do that. I don't follow a lot of the, like, I
- 7:18don't wear a shirt with a collar when I teach, you know what I mean?
- 7:21And I, and it's like there's a bunch of little things that, that kind of
- 7:24add up where I, I kind of eschew the, the stuffy academic
- 7:28norms. I'm not comfortable or familiar with elite spaces. I write a lot
- 7:32about how social media has put us in our own castles with like
- 7:36minded people loving like minded content. And you validate who you
- 7:39are by hating. And one of the things that you
- 7:43challenge and part of that is a sense of partisan and
- 7:47fake news, that it's not that we necessarily match
- 7:51what our politics are. And I was curious how you could build on
- 7:55that because I was always of the opinion that once they've locked you in this
- 7:58castle and the algorithms shape how you think, feel and behave,
- 8:02very much it's validated by who's around me versus what I
- 8:06believe. Internally, most of the errors that people make
- 8:09are due to lazy thinking. They're not spending enough time
- 8:13engaging in deliberation and reflecting on their. Whether their
- 8:16intuitions might be wrong or just in general whether the stuff that they're engaging with
- 8:20online could be erroneous or misguided or disinformative or
- 8:23whatever. There's a whole other kind of class of thinking about
- 8:27human error and that like political bias that is way more
- 8:31focused on kind of like pointing fingers in many ways where it's like
- 8:34partisan bias occurs partly because people kind of
- 8:38want to be partisan. They are actively kind of protecting
- 8:41their cognition from threats and they convince
- 8:45themselves that the things that they want to be true are true. I mean, a
- 8:48lot of the studies don't find evidence for that and I just don't think it's
- 8:50true. I think that most people kind of are trying to figure
- 8:54out what's true and it's hard to do that, especially in a kind of polluted
- 8:57and confusing the information ecosystem that we're currently inhabiting.
- 9:01How much has to do with social media and their
- 9:04understanding of algorithms and where they can just sort of fly fish in
- 9:08the right headline and the right bait at the perfect time to get people to
- 9:12just jump in and either continue to validate this
- 9:16journey they're on, or if anything have a
- 9:19rock that they can put in their slingshot and fire to someone else. Yeah. I
- 9:23mean, we live in an attention economy in a context where
- 9:27the most precious resource is just someone's paying attention
- 9:31to what kind of stuff that you're presenting. That means you have to present something
- 9:34that will grab their attention. And how do you do that? Well, it's hard to
- 9:37do that if you're constrained by the truth. So there is research
- 9:40showing that false things have an advantage on social
- 9:44media. Once you take the truth constraint out, then you can say more things that
- 9:47will grab attention. And so there's a market for that sort of
- 9:51thing. And then there's also, of course, a market for things that are highly emotive,
- 9:55that make people angry or create outrage. It's not that people want that. If you
- 9:59ask people explicitly, they're not searching for, you know,
- 10:02false rage, bait content. It's just that's what grabs their attention. Naturally,
- 10:06it's the same. And a good example of this is clickbait. You know, the
- 10:09headline is something that makes it seem like it's really interesting. You click on it.
- 10:12There's nothing of consequence in the, in the article. People hate it.
- 10:16It tricks them. They do click on it, but they don't want to see it.
- 10:20And so it's a decision for the company to say,
- 10:23this is something that we're going to allow on the platform. And what we've done
- 10:26is we've just kind of abrogated that. We said we don't really blame them. And
- 10:30that whole narrative has even shifted where the social media companies aren't really being put
- 10:33to blame for the fact that they are taking advantage of the kind of weaknesses
- 10:37of our psychology. You've countered with some habits
- 10:41that people can do to just spend a little bit more time,
- 10:45make us pause before we believe something or share something.
- 10:49What's your research showing that we should be thinking the average listener
- 10:52should be doing as they're scrolling through their social media feeds?
- 10:56When people are sharing misinformation online, they often aren't even thinking about whether it's
- 11:00true or false. When you share something with other people, you share it because it's
- 11:03important or that it might be, you know, or maybe like something that people would
- 11:07like or whatever. And then like down the list is, wait, is it true?
- 11:11And so we came up with this, like very simple intervention was just subtly reminding
- 11:14people about the truth, you know, ask them question about accuracy. We have
- 11:18little advertisements that were like, think before you share that kind of stuff.
- 11:22Those subtle reminders improve the extent to which people are sharing true versus
- 11:26false stuff subsequently. So the message there is simple. It's like, just be
- 11:29more mindful when you're engaging online. It doesn't fix all the problems, but getting
- 11:33people a little bit more aligned with the truth will make them more discerning. We're
- 11:37not giving them new information. It's just getting them more sensitive to the fact that
- 11:40they might be making bad decisions decisions because they're not focusing on whether things
- 11:44are true and they should be. More often, you know, when I look at. We
- 11:47always use Walter Cronkite, you know, as the journalists that we all trusted and
- 11:51believe. But even within that newscast, there was a certain element of bias
- 11:55and US Propaganda. And we know it was only when they started showing the body
- 11:58bags of Vietnam War that people realized, you know, that it wasn't all
- 12:02going so well for the Americans. So how do you counter? Because today
- 12:05maybe the Walter concrete is a Joe Rogan. How do people come to terms
- 12:09with it? Because often they expect other people to do the curation
- 12:13and validation of what they're saying. There is no
- 12:17context in the modern world where you can determine the truth
- 12:20for everything yourself. Very few of the things that you would
- 12:24believe are things that you can verify with your own experience. Everything is
- 12:28based on other reports. So we need trust. The problem is
- 12:32that there's bias, which is, you know,
- 12:35presenting a kind of like, particular angle on things that's not inherently bad. I
- 12:39mean, there's like cases in which, like, if you watch a sports,
- 12:43you know, the hometown announcers will have a bit of bias in it. You can
- 12:47understand that it might make a bit more fun. And you can. You can, you
- 12:49know, appreciate that and take it into account. The problem is that if you have
- 12:53bias that is unappreciated or unspecified,
- 12:57you don't know that you're in that. And you think that it's just an honest
- 12:59accounting. That was the issue. Like take Walter Crockett. There was like, you
- 13:03know, imperative at the time for news, news people to
- 13:07be unbiased and they were presenting a biased view. And that was the problem
- 13:10here. What we have is political news reporting that
- 13:14the whole idea of being neutral is just completely evaporated. The
- 13:18other element, which is like, it has to be truthful. There are
- 13:21components of the like, quote, unquote news atmosphere that are
- 13:25not doing that either. Legitimate journalism, mainstream, they do a
- 13:29decent job of like, not. There's no inaccuracies. They don't, like, make up sources. They're
- 13:33not stretching the truth to such an extent that it's like, that's now Untrue.
- 13:37But there are lots of like seemingly legitimate news sources that
- 13:41are essentially just parodying falsehoods. And that is even bigger
- 13:44problem than just the bias problem. And I would argue too that today
- 13:48the people people are listening to aren't just
- 13:52reading the news, they're putting their opinion behind it. How does
- 13:55somebody come to terms with the fact that we'll go back to Joe Rogan.
- 13:59I don't think he's worried about sources as much as he's worried about
- 14:03his audience and providing the kind of opinions that people
- 14:07will go away with and say did you hear Rogan and his interview
- 14:11with and oh my God. I think this is probably the worst
- 14:14development in the news ecosystem is the injection
- 14:18of opinion into news. So if you have a 24 hour news station, there
- 14:22isn't 24 hours of news right. Of events to cover.
- 14:26Unless I mean you might cover all the things in like on in the world,
- 14:29but it's like there's not 24 hours of things that are worth covering. In
- 14:33Canada we have a 24 hour news station, but it's just the same hour of
- 14:35news repeated over and over again. It's not news for entertainment. If you
- 14:39go to the Fox News or MSNBC for 24 hour news,
- 14:43a very, very small amount of that is actual news. The rest of it is
- 14:46prognostication. People talking about events, that's just not journalism,
- 14:50it's just entertainment. It's not different than listening to Joe Rogan. He
- 14:54ostensibly is not doing it for news ad obviously he's not a journalist. It's just
- 14:57people talking and now people cannot distinguish between one and the other. How
- 15:01much of it has to do with. I always call it the campfire. In the
- 15:04past the vast majority of people sat around the campfire and listened to someone
- 15:08share a story. Today with social media, I'm expected to add a
- 15:12log to the campfire. I'm expected to have my own point of view.
- 15:16The logs I put on the fire is very much how I earn my likes.
- 15:20So is the problem it's the people putting opinion to the news
- 15:24or. The fact is the reason why there's such an appetite for those opinions
- 15:28is other people feel that that's their new role in society, that they have to
- 15:31comment. You know what, actually I think that people don't comment enough.
- 15:35Most people don't post hardly anything online. What
- 15:39you have are people who produce a huge amount of the content.
- 15:43Often the people who have who are the least like not the people that you'd
- 15:47want to be listening to. Because they're the most interested in garnering influence
- 15:50in a social media context. There is feedback, you get
- 15:54likes and comments and you get like. But most people don't see the comments that
- 15:57other people write. To some the primary post and mostly it's just likes and
- 16:01re shares and stuff like that. And so the incentives for the people who are
- 16:05generating content are basically perverse. It's the sort of thing that will
- 16:08generate eyeballs attention like I mentioned before. And in order to do
- 16:12that you often need to be kind of like salacious and make things interesting.
- 16:16A lot of people who read it, if they were to say something, they'd say
- 16:18that's dumb or I don't like it, but if you get enough people in front
- 16:22of it, then you'll get lots of likes. And so you have this trade off
- 16:25where people are generating a lot of stuff. If people were giving their honest opinion,
- 16:28they'd realize that most people don't like what they're saying. But because most people, if
- 16:32they don't like something, they just don't say anything at all. They don't get negative
- 16:35feedback, they only get positive feedback and then they keep on going and it snowballs.
- 16:38So if people were more likely to like fact check each other online, for example,
- 16:42then that would produce more accurate information that people are sharing
- 16:46because they wouldn't want to be like shown to be false. You know, you talk
- 16:49about the attention economy. In the past it was monetized
- 16:53with TV advertising and television and radio and it
- 16:56was today obviously it's that money's
- 17:00disappeared from mainstream media and it's moved to what I call the
- 17:03hairpin turn journalism. Like I'm waiting for something to happen to crash the
- 17:07sensational, the headlines, the doomsday scrolling. What
- 17:11happens to democracy, civility, society if
- 17:14we lose the kind of journalism that you're talking about
- 17:19where it's not so much opinion, it's fact and
- 17:23fact that makes people think. What happens if that disappears? Well, we're seeing it right
- 17:26now in the US we're seeing backsliding of
- 17:30democracy, essentially extreme levels of partisanship. If you
- 17:34add in the fact that there's two sides, explicitly
- 17:37sports, essentially there's the red team and the blue team and then you
- 17:41have all of this other people make a living
- 17:45off of the sport, off of increasing partisan
- 17:48animosity and there's incentives for gaining
- 17:52attention. We lack disincentives for
- 17:56accuracy. I don't know a way in which you don't go from that to some
- 17:59sort of collapse or some sort of like There needs to be some sort of
- 18:02change, some sort of substantial change, positive. I mean, I do have. The
- 18:06research shows that people do care more about the truth than they seem to. If
- 18:09you give people evidence, they will change their minds. People are more reasonable than we
- 18:12give them credit for. But it's. They're fighting an uphill battle. And you talk a
- 18:16lot about conspiracy theories. Is our appetite for conspiracy
- 18:19theories, is it increasing as we move into this world of social
- 18:23media where anybody can put anything or we always, I mean to me,
- 18:27I would have said the National Enquirer in the past and some of those tabloids
- 18:31used to see the grocery store with the two headed Martian. Have we gone
- 18:34from being the sideshow in the circus to mainstream is really the norm for
- 18:38conspiracy. So I don't think people's underlying interests have shifted. That's driven
- 18:41partly by what people are exposed to. The market is stronger
- 18:45conspiracies. I mean, if you remember back in the X Files days, going back to
- 18:48when I was a kid, I was always fascinated by these like paranormal beliefs and
- 18:52things like that. If you just watch X Files, the conspiracy theorists in there, they're
- 18:55almost like quaint. It's not like that now. That whole space is very, very
- 19:00kind of negative and like in some cases violent. A lot of that is top
- 19:03down. People are feeding into it because it's benefiting them politically. Think
- 19:07about like the election fraud. Probably the most widely believed
- 19:10conspiracy is that the 2020 election
- 19:14was fraudulent. Majority of Republicans still believe that to be true
- 19:18and there's no evidence for it at all. Most conspiracies were actually fringe. That is
- 19:22not the case anymore. But that's not because people are more interested in conspiracies. It's
- 19:25that people are being thrown constantly all these different conspiratorial
- 19:29ideas. And it's not once you're exposed to all that information, it's hard to kind
- 19:32of overcome it. Some of your latest research with AI
- 19:35shows that you can sometimes chisel away
- 19:39someone's position in terms of their conspiracy beliefs. The
- 19:43common trope that people have about conspiracy theorists is that they're down the
- 19:47rabbit hole because they kind of want to be down there. Even in the, in
- 19:50the field of psychology, there's all this work on the importance of people's needs
- 19:53and motivations for why they believe conspiracies and that people
- 19:57believe them kind of because they convince themselves that they must be true. And that
- 20:01explains why like they believe a conspiracy that's so flat, laid against evidence,
- 20:05like the earth being flat and Scientists covering it up. However, an alternative is that
- 20:08they believe them because they just like saw a bunch of stuff that was seemingly
- 20:12compelling and they haven't seen good counter arguments because they're not in the right
- 20:16part of the information ecosystem where they get the other kind of side of the
- 20:19information. And so we do in these experiments is we used artificial
- 20:22intelligence to provide individualized, highly
- 20:26informative counter arguments that were all evidence based. And so people
- 20:30would write about a conspiracy that they believe in in their own words, provide the
- 20:33reasons why they believe in it, and then they have a conversation with, with an
- 20:36AI, there's GPT for it. With just some extra prompting, the AI just
- 20:40gives them like direct counterarguments for everything that they say. And
- 20:44then they say, okay, what about this? And the AI says well, actually that's wrong
- 20:46because of this and that and the other thing. And they keep on going back
- 20:49and forth. And after about like eight minutes of that conversation,
- 20:5225% of the people who believe the conspiracy don't believe it anymore.
- 20:55So that's a big shift. People, he would have thought as like the key,
- 20:59like primary prototypical example of people who won't change their minds according to evidence.
- 21:03If you give them good evidence, they do change their mind. Do they have an
- 21:06appetite? I can see it in the research circle and you're getting to do it.
- 21:09But is there an appetite for the average human to say, I
- 21:13want to be challenged? We know that the people in these studies
- 21:17really enjoy it. I mean, it is the case that people who believe conspiracies,
- 21:20they like talking about them and they also want to know if they're right. I
- 21:24mean, they think they're right. There's a subset of course, that are like just
- 21:28unwilling to accept the possibility they could be wrong.
- 21:31But if you give people, most people, enough evidence
- 21:36in the right sort of way, I mean within the context of conversation, people like
- 21:39it. Getting people in the door, that's the harder question. Getting
- 21:42them interested in even having this kind of detailed conversation. It's going to
- 21:46work for some people. Some people really do want to figure out the truth and
- 21:49they're misguided. Some people. It might be a harder to ask, but
- 21:53we have to still to tackle that. Essentially.
- 21:57We come back, Gordon Pennycook and I talk more about
- 22:02AI, the good, the bad and even the
- 22:06ugly foreign.
- 22:11Chapman host of chatter that matters. You lock your doors, you
- 22:15protect your valuables, but how about your online activities?
- 22:18Cybercrime has become a household word. One of the biggest concerns
- 22:22facing businesses today, a big shout out To RBC for creating the
- 22:25Vault, an online cyber safety playbook packed with tips and steps
- 22:29you can take to counter cyber threats. Visit rbc rbc.com
- 22:33cyber and download your playbook today helping you protect your
- 22:37online activities. Well, that matters to rbc.
- 22:44If you're going to interpret whether something's true or false, you have to
- 22:48by necessity bring to bear whatever knowledge you have.
- 22:52And so imagine someone who only watched Fox News, maybe by accident of
- 22:56history, like that's what their parents watched. They don't really care about it, but that's
- 22:59what they're exposed to versus someone who wants to listen to msnbc. The range of
- 23:03things that are going to be plausible for them in the political realm is going
- 23:06to be way different. Right? And it doesn't mean that they're being biased, it just
- 23:09means that they have a different basis of knowledge. Those are the
- 23:12words of Gordon Pennycook. He's an associate professor at Cornell University.
- 23:16Really making a name based on the premise of trust
- 23:20and maybe how we're too trusting. And you'll learn about
- 23:23some techniques and tools where we might spend more time thinking
- 23:27about what we're being offered versus immediately running to liking it.
- 23:31And Chariot.
- 23:35You know, I've made an argument over the years in my social media that we've
- 23:39lost the middle ground where a lot of democracy and humanity
- 23:43and advancement happened. You know, it wasn't always perfect for
- 23:47one person or the other, but through consensus and collaboration.
- 23:51But I feel that middle ground's become this toxic moat as we continue to
- 23:54be imprisoned within our own circle of people that have like minded
- 23:58beliefs. Do you think what you're talking about in terms of
- 24:02AI countering challenging will
- 24:06create some drawbridges to have people leave their castles and come
- 24:10back to the middle where they're much more open minded to somebody else's
- 24:13point of view? We're actually testing that now. If you can use
- 24:17the AI to not just change what they think about a particular topic,
- 24:21but to make them more open minded in general, that's harder. Obviously. That's essentially a
- 24:25trait that people have, that people vary in how open minded they are, how willing
- 24:28they are to question their beliefs, how intellectually humble they are. I think it's
- 24:32plausible at the very least. One thing that we find is that people do trust
- 24:35the AI more and they're more interested in the conversation. So in the more
- 24:39limited case, having one of these conversations would make people more
- 24:43interested in having more of them. And so then in a certain sense that makes
- 24:45them more open minded. But if more In a more general sense, we don't know
- 24:48yet. Do you think the reason why the AI is proving to be so effective,
- 24:51it takes away the emotional side. So if I'm arguing with you over my
- 24:55position versus yours, it gets very emotional and I stop listening to
- 24:59what you have to say and the volume of conversation
- 25:02and the intensity goes up. Where AI is much more pragmatic and logical.
- 25:06When we've done experiments where people think they're talking to a human and it's the
- 25:10AI or. And it works the same, it doesn't change anything at all. But
- 25:13that's still in a like anonymous. They don't know who the person is. It's not
- 25:16face to face. All of the. And they're certainly not like you talking
- 25:20to your uncle, where there's a history and like all these things that makes
- 25:24it more complicated. So that's partly emotions, but it's also partly like
- 25:28legitimate reasons that people would have to question the motives of the other person and
- 25:31all these sorts of things. And the AI doesn't have motives, it's just giving you
- 25:34information. So that's, that's different. At the same time, we have done, and
- 25:38this was just brand new data, we did a version where it was explicitly a
- 25:41political topic, where there was like people were being persuaded that they were wrong about
- 25:45something that was political. And we made the bot to be
- 25:48essentially aggressive and rude. It was engaging in a, in
- 25:52language that you would normally see in an online debate. You know, be kind of
- 25:55flippantly like saying, oh, I can see why you might think that. But like
- 25:59obviously it's wrong, but then it would provide all the facts and evidence and it
- 26:02didn't change the effect, it didn't actually undermine the effect because the
- 26:06facts and evidence were so strong that it didn't really undermine it. That
- 26:09doesn't mean to say people hated it. They didn't like after the fact, they thought
- 26:12it was biased, they didn't think it was a good source, but they still changed
- 26:15their mind to the same extent. How does it counter. I mean from what I
- 26:18learned in my basic psychology studies, cognitive bias is you can give me
- 26:22all the facts you want, the math, the language in the
- 26:25universe, and I'm still not going to believe it. What's changed with AI versus
- 26:29I think that's just wrong. Yeah, I mean I think that's the dominant
- 26:32perspective in psychology on this sort of political
- 26:36motivated reasoning. It was a convenient story that one could use to kind
- 26:40of explain why people aren't as reasonable as the people doing the research.
- 26:44Maybe to me there was like. And maybe this goes back to the earlier part.
- 26:47Coming from a small town or whatever. I'm very resistant to the kind of
- 26:51elitism in academia. Like, I don't know what,
- 26:55what the appeal of the idea that people are more reasonable than we thought
- 26:59was, but part of that has to do with the fact that we. It puts
- 27:02less blame in the person for making mistakes. And I think part of the
- 27:06reason, like academically, in terms of the research that we people thought that was
- 27:10that they weren't giving them good enough facts and evidence. They didn't have the
- 27:13tools to really be compelling because they, because the researcher. So if you're going
- 27:17to do an experiment where you're debunking a conspiracy, you have to come up with
- 27:21a conspiracy that you think people believe, then you have to come up with
- 27:24evidence that you think would counteract the reason that people have to hold that.
- 27:28And if there's any mismatch, then you're not going to be very effective. And so,
- 27:31like, if you have this gap between the people who believe something that you're doing
- 27:35research on and the people who are doing the research, the academics, that
- 27:39makes this a problem. Now we can just bridge that gap.
- 27:42Turns out they're not so unreasonable. I want to move to the negative side of
- 27:46AI that you talk about as well, but my last question before I do is,
- 27:49you know, you really framed it as a sport. You know, in politics
- 27:53is red versus blue in the States. Who's the greatest
- 27:57of all time? Is it Wayne Gretzky or Bobby Orr?
- 28:00A sport that people love to have, they love to defend their point of view
- 28:04and they want to get excited and energetic and it gives them a sense of
- 28:08purpose and it gives them a sense of. My currency at this
- 28:11conversation or online is I have something to offer.
- 28:15Do you think that AI, which is obviously a very good debater
- 28:19and a very good organizer and curator and presenter of facts,
- 28:24will it get to the point where it's going to be neutered? Because
- 28:27people still like the gladiator element of the sport where I'm arguing with
- 28:31you. And it's less really about whether my sword is made
- 28:35of platinum or steel. It really just has my ability to swing at
- 28:39captures an element of sport. But what it's missing is sportsmanship. One of the
- 28:43most important parts of sport is learning what is appropriate and what is not
- 28:47appropriate. And there's, I mean, part of that is rules, but part of that is
- 28:50just like, what does it mean to be a sportsman? One of the
- 28:54Kind of like underlying ideas of a democracy is that people debate ideas.
- 28:58And that's great, but it has to be debate under
- 29:01with sportsmanship, which is where people are actively trying to actually get it
- 29:05right. It can't be just for the sake of debate, has to be a
- 29:09debate for the sake of trying to get to the truth. It just so happens
- 29:12that like, there's some research on like actual watching people do debates.
- 29:16It generally is not a good form, like a pitch debate,
- 29:19because the person is explicitly in a context where they have to win the debate
- 29:23and not in the context where you're learning from the other person to try to
- 29:26figure out what the truth is. Using debate as a mechanism to figure out what
- 29:29actually is true. So I think that's part of the problem. I think it is
- 29:32the case that they find entertainment from the partisan
- 29:35backlash. But I think it also goes both ways. I mean, people don't like it.
- 29:39I mean, in general, the poisonous political atmosphere is not something
- 29:42that most people view positively, even if there are contexts in which it
- 29:46might be fun to engage in a bit of debate. But I think we've gone
- 29:49so far beyond that. That's so infrequent that it's not really a consideration even. But
- 29:52those in power know that it's not even debate. It seems like I'm after.
- 29:56I just want to dictate what I believe the truth is and I want to
- 30:00overpower you with that. Even those in power that you would hope
- 30:03would have some decorum, some desire to be fact
- 30:07checked, very often don't they just come out with these statements that really
- 30:11are as far from the truth as possible. Are we just being
- 30:14as a society? Is this an element of submission
- 30:18where in many cases democracy is giving away to dictatorship? I
- 30:22think so. What we want to be is debating the
- 30:25accuracy and the truth of what the political elites are saying. What they
- 30:29want us to do is to debate who's right or who's wrong
- 30:33amongst each other. That's the other side. You should hate the other side. And the
- 30:36debate is the battle between the two sides of this political
- 30:39thing where really it's the people and the leaders and it's. We have to be
- 30:43assessing if they're doing what they ought to be doing and they're doing things
- 30:47that are consistent with, with what the people who voted for them are actually want
- 30:50them to do. And that is not remotely what we're seeing. The
- 30:54political leaders are doing what they can to hide what they're doing and convince
- 30:58people to like focus on all other parts Putting
- 31:01troops in cities, ostensibly, is about safety
- 31:05within the city, where it's just obviously not the impetus for that. It's
- 31:09about political control. None of us benefit from this.
- 31:12It's the people in power who benefit from it. The whole idea in the states
- 31:16is that there's this battle between parties, and that's all the
- 31:20fighting is between the Democrats, Republicans, but really it's just between the
- 31:24extremely wealthy elite and everybody else. And do you think
- 31:27when you talked about earlier, the negative and positive
- 31:31and the positive you found within the research, but the negative seemed more
- 31:35intuitively in your heart, do you think there's a future for democracy?
- 31:38I think part of the reason why the US Is
- 31:42still where it is is that people are generally fine. You know what I mean?
- 31:46It's still a very wealthy country, but eventually that doesn't
- 31:49happen anymore. You can stretch the truth far enough
- 31:53until you have to pay the piper. It depends on how extended
- 31:57that window is of future for democracy. But I think there's a possible that there
- 32:01is some rock bottom that we hit, and maybe we aren't there yet, and maybe
- 32:04that rock bottom is in for 10 or 15 years, but there's a rock bottom
- 32:07that we can crawl back out of. And this is the ebbs and flows of
- 32:10history. I don't know, maybe the ebbs and flows are now more rapid than they
- 32:14were in the past, but I still feel underlying, some optimism for
- 32:17the future of humanity. While I have you, I also want to shift gears and
- 32:21talk about the negative side of AI. And you talk about AI
- 32:25swarms, for example, that could flood the web with fake
- 32:28consensus. And talk to me a little bit about what your research is doing in
- 32:32terms of how AI in the wrong hands are in
- 32:36fact putting even stronger strings on society in terms of them becoming
- 32:39puppets. It allows you to automate good things and bad things. I
- 32:43mean, we've done stuff where we use AI to give people good facts and evidence.
- 32:46You can do the opposite. In fact, we have done the opposite. You can use
- 32:49AI to convince people to believe conspiracies, and that works too.
- 32:53And so it's just another component of a larger problem, which is
- 32:57our information ecosystem is becoming extremely complex and hard
- 33:01to distinguish truth from falsity. It is absolutely
- 33:04imperative to teach people the skills that they need. Psychological
- 33:08skills, critical thinking, learning how to think analytically and question things.
- 33:12What is a good source, what's a bad source, stuff like that. Digital literacy, AI
- 33:16literacy. All these things need to be taught, of course, to children as
- 33:19they're growing up. But we need to teach everybody because these
- 33:23are evolving technologies and we are losing the battle right now. Do you think we're
- 33:27teaching the right things to children growing up? No. I mean, if we were, then
- 33:31we wouldn't be in the predicament that we're in. I think there's a lot of
- 33:34improvement that needs to be made. I think educators often talk
- 33:38about teaching critical thinking. That's good. But
- 33:42think about how much we teach math versus how much we teach critical thinking. Like
- 33:45basic skills of like how to question things. You know, it's
- 33:49not close. And there are countries that do better. Finland, for
- 33:53example, has real classes that are about navigating the
- 33:56digital information ecosystem. And they are far more resistant
- 34:00to misinformation in Finland than we are in the States.
- 34:04The problem is that it doesn't benefit a political
- 34:07oligarchic society with very wealthy political elites that are trying to control
- 34:11things, to have an educated masses, not just educated with
- 34:15knowledge, but educated for how to think and to question things. And where do you
- 34:18see this, the scams? I mean, I just did a study with
- 34:22rbc, an interview, just talking about how vulnerable
- 34:25seniors are to. To cyber threats and bad actors.
- 34:29I mean, is that something you're looking at as well, that how
- 34:33it's becoming so much more sophisticated and how we
- 34:36convince people and how they're under attack? Yeah, I mean the scams are
- 34:40becoming more sophisticated. In some cases they're even like will
- 34:44do things where they are explicitly unsophisticated. Like they'll have the email, will have
- 34:48errors, but that's done in a sophisticated way so that the people who click on
- 34:51it and go through and then they have to do all this kind of stuff
- 34:53to try to get their like bank details. They don't notice the errors. And so
- 34:57they're more likely to not notice that it's a scam. And they. It's all very,
- 35:00very complicated. It's. With everything else, this is less that
- 35:04you can trust. But we don't. Our underlying psychology has not changed.
- 35:08Humans are inherently trusting species. If someone tells you something,
- 35:12you don't just immediately reject that as being untrue. Like it's. It
- 35:16would be kind of like normatively inappropriate to just completely question everything that someone says
- 35:19to you. And so you need to teach people actual literal tools and
- 35:23skills, give them the tricks that they need to try to identify
- 35:27these sorts of scams and falses on the Internet. You have to put as much
- 35:30effort into teaching people how to fix them as the people are
- 35:34putting effort into creating the new scams. In the first place. And there's a gigantic
- 35:37imbalance there. Speaking of teaching, one of your courses at Cornell
- 35:41is called the Science of Human Stupidity. Talk to me about
- 35:45to what end. People, I think, sign up for the course because they
- 35:49want to understand other people's stupidity. Because everyone thinks that they're smart
- 35:53and everyone else is dumb or not. I mean in general sense. And what I
- 35:56want them to come away with is for them to understand the nature of their
- 35:59own stupidity more, which is where are the places that we make mistakes? Where are
- 36:03our own blind spots? How are we overconfident? The whole
- 36:07idea of the course is for people to engage in more self
- 36:10reflection and to have the tools to essentially become
- 36:15better critical thinkers. People that are more discerning in the world. And
- 36:18discerning doesn't just mean, oh, I know what the true and the false
- 36:22things are that I'm seeing. It means I'm aware of
- 36:26the cases in which I believe something that is false or that might not be
- 36:29true. You are more calibrated and you have stronger, what we refer to as
- 36:33metacognition. You're thinking more about the way that you think and that's what you
- 36:37need to be a more reasonable person in the world. And we kind of just
- 36:39like need that to happen for everybody. If we add a bit more
- 36:43reasonableness, that will shift the way that people are engaging with content, maybe change
- 36:47the incentive somewhat and maybe get us in a better place. You know,
- 36:50you're young person making a real mark in this world and
- 36:54what's next? There's a lot of this stuff that we're talking about is
- 36:58giving people tools to be more reasonable in the world. I guess it goes back
- 37:01to, you know, my dad saying, you know, what's this for? At the end of
- 37:05the station? To some extent what this is for is to actually produce that
- 37:08benefit in the actual world. And some of that is, you know, things like this,
- 37:12whatever, just telling people the stories and giving them the data, doing
- 37:15a book or something. But some of it is actually like building tools,
- 37:19using things like artificial intelligence that people can use. We do have one
- 37:23website that's called debunkbot.com people can go and they have a conversation about
- 37:27conspiracies just to kind of test that out. But we're trying to do more versions
- 37:30of that sort of thing that are more accessible to people. I
- 37:34have no illusions that I'm going to save democracy, but if I can, mostly what
- 37:37I do is enough to convince themselves that I'm doing something
- 37:41so that I can go to sleep at night. You know, I always end my
- 37:44podcast with my three takeaways. And I actually
- 37:48love the way you just ended your sentence. I mean, what's this
- 37:51for? It's a great lesson in life for anybody to really
- 37:55talk about that. To what end? What am I doing this for? What
- 37:58matters? How is it going to materially impact people? And if
- 38:02we did follow the work you're doing with some of your colleagues and spend a
- 38:06little bit more time thinking versus just absorbing and
- 38:09sharing, it might be the bounce back that democracy needs. The social
- 38:13media algorithms and the political climate are
- 38:17marching step and step. I no longer need a megaphone on the back of a
- 38:20train to convince people in the. In the Rust Belt that I can do a
- 38:24better thing. That with these algorithms, I can be incredibly
- 38:27impactful at firing all their emotions for
- 38:31the wrong reasons as opposed to bringing them the truth. I
- 38:34love the sense of AI. It's about automation, good or bad, and it really
- 38:38comes down to the word protecting us at times. It can provide us
- 38:42with the tools we need so that we understand really what the truth is. And
- 38:45there's going to be times where we need protection from AI because it's
- 38:49going to get more and more sophisticated at trying to either scam us
- 38:53or change how we think, feel and behave. And again, the last
- 38:57thing is, which often comes through is just how sometimes that single
- 39:01conversation, in your case, you know, third year university
- 39:04and somebody said you have an aptitude for an individual like, you
- 39:09probably spent a lot of time cracking up the class and having fun and
- 39:12partying and just deciding what they wanted to do, realize that they
- 39:16might have a different purpose or calling in life. And I hope that wherever he
- 39:20is, looking down, he's looking at you, saying, that was a conversation well spent,
- 39:23as I think this conversation today has been one well spent. So I really
- 39:27appreciate you joining me on Chatter that Matters. That's my
- 39:31distinct pleasure. Thanks, Tony.
- 39:39Now, let's hear from the brilliant mind of Milos Stojadinovic,
- 39:43our cybersecurity expert from rbc.
- 39:47Just give us a sense. I know you're quite a modest guy, but you're
- 39:51very senior. I would argue RBC is one of the best in
- 39:55the world at trying to counter this. So just give us a sense of what
- 39:58you do for a living, and then I can ask you a couple questions. My
- 40:01background is largely in what we call adversary emulation, or perhaps
- 40:05a little bit more informally, offensive security. So I have
- 40:08built the majority of my career breaking into governments organizations,
- 40:13defense contractors, you Name it. And using my skills of offense to inform
- 40:17how we should design adequate and appropriate defense. So by attacking
- 40:20organizations and institutes, we understand, understand where are their gaps, where are their
- 40:24weaknesses in their environment, across people, process and technology and
- 40:28implementing that change. On the other side, I have a. Threat hunting team which
- 40:31sets the hypothesis that there are threats inside of our environment. There are threats that
- 40:35are targeting our clients. They actively try to hunt for those threats inside of our
- 40:38environment and in our digital channels. And lastly, I have a threat modeling team
- 40:42that integrates the thought process of security into the products and the solutions that
- 40:46we build. So we don't think about security as an afterthought. We think about it
- 40:49as something that's part of the journey. Do you ever think how much you could
- 40:51have made on the other side? Yeah, I mean, you sound like my little brother,
- 40:54because you ask me that question all the time. But look, the reality is, is
- 40:58that ethics is a large part in what we play, right? So in my role,
- 41:01I really have one thing. If you strip all my technical
- 41:05skill sets away, I have integrity. If I compromise my integrity, I obviously can't work
- 41:09in this industry. Not to mention there's top secret clearances and all the other stuff
- 41:12that happens. But I have a deep desire to want to help people and protect
- 41:16people. I don't believe in victimizing people. I've seen friends and family around me
- 41:19become victims of cyber crime or traditional crime. Like most
- 41:23people in life, I've been a victim of crime as well. So it's just.
- 41:27It's not in me. Right. It's just one of those things I think you're either
- 41:29born to do or not do. So paint a picture for us what
- 41:33is really going on out there, because obviously RBC
- 41:37and all of your clients are all massive targets. So
- 41:40just open our minds to the. Seriousness of the threat. 15 years ago or so,
- 41:44if you wanted to target a company or target, you know, someone of wealth, you
- 41:48had to, like, know all this stuff about how to target them, right? You had
- 41:50to know how to do reconnaissance and how you might get inside of the environment,
- 41:54and how do you persist in that environment and how you move around it and
- 41:57how you get access to the data. And that has become commoditized
- 42:01in many different ways. So these days, if you fancy the
- 42:04idea of being a ransomware threat actor and holding organizations for ransom,
- 42:08you don't have to know how to write ransomware that's sold as a service. Now
- 42:11on underground markets, you can get 24 by 7 support, where you can
- 42:15literally talk to someone 24 hours a day, seven days a week to get support
- 42:18for your malware. Malware, It's a big business model because what happens is the people
- 42:22who are authoring the malware, the ransomware, are taking a cut off of
- 42:25anything that the ransomware their clients effectively
- 42:29are implanting. And then on the other hand, you have, for example, threat
- 42:32actors that focus on initial access, brokering, so they may
- 42:37breach organizations or even people of wealth into their
- 42:40systems and they just sell that access to the highest bidder. So
- 42:44if you, you don't know how to compromise and you don't also know how to
- 42:47write ransomware, but you've got maybe a sliver of that skill set if you
- 42:51have the funds to get started, this is how you can kind of approach that.
- 42:54And that's why we're seeing the scale grow up massively. This is organized, it's
- 42:58sophisticated, and they're making tons of money. I think we opened up the conversation
- 43:02with this being a $10.5 trillion problem globally. And
- 43:06there's also increasing nation state involvement and support in this. Right. It's
- 43:09not unknown, for example, that if you live in a country that's non allied
- 43:13with another country and you target their citizens or their corporations, you have virtually
- 43:17no chance of facing any consequence. And some countries actually even
- 43:21encourage that. Is it fair to. Because the other thing I was reading about is
- 43:24that some of those leaders actually also take a cut. $10 trillion
- 43:28is a lot of money to launder. Increasingly, a lot of these funds are moving
- 43:32towards cryptocurrencies. Right. That don't are. They're not part of the same
- 43:35standard regulation. But these people can leverage that wealth to still, you
- 43:39know, do what they need to do, purchase the things that they want to purchase,
- 43:42and that kind of leaves them largely outside of the regulated space that the rest
- 43:46of the world operates in. Once again,
- 43:50a special thanks to RBC for supporting Chatter that matters.
- 43:53It's Tony Chapman. Thanks for listening and let's chat soon.