How does your brain decide that some other person has their own inner life? And how does this sometimes go in a different direction where you end up viewing another person more like an object. This is what neuroscientists mean when they talk about dehumanization. It means that your brain is not cranking up its social machinery to understand that the other person has a mind like you do.
We're gonna dive deep.
Into this today with social neuroscientist Lessana Harris, will ask is dehumanization a cause of violence or is it the fuel that keeps it burning? Do people who think of themselves as highly empathic sometimes dehumanize.
More than other people?
And on the flip side, why do we sometimes think that chatbots or robots are people with interior minds? Little children raised with AI grow up to fight for AI rights. Today we're going to dive deep into how your brain sees others. So get ready for a great brain stretch. Welcome to Inner Cosmos with me David Eagleman. I'm a neuroscientist and author at Stanford and in these episodes we sail deeply into our three pound universe to understand how we see the world, and importantly, today, how we see
each other. Every second of your life, your brain is working to figure out something very important in the space around you, what is alive in the sense that it has intentions and feelings and plans. In other words, it has a mind. We are mind detectors, and so whenever you walk into a room, your social antenna are up. You're not conscious of this, but your brain is putting tons of effort into figuring out where there are people.
And you are watching faces, and you're registering postures, and you're listening to the tones of voice, and your brain is building models like okay, I think that person is irritated, and that person is curious, and that person is hiding something, and that person wants approval. This is one of the brain's deepest talents is mind perception. Think of this as your ability to infer an inner movie playing behind someone's eyes. As I said, this is typically done automatically without any
conscious awareness. But you do this all day long with friends and family. You do this with strangers, you do this with people you've never met. And by the way, we don't do this just for other humans. Our neural mechanisms for making this happen. It applies more broadly, so we anthropomorphize, meaning we assign human like minds to non human things. You might treat your car like it has a personality. We certainly do this with robots, and we
root for the animated toys in a Pixar film. You can even watch a film with some animated triangles and circles moving around no words at all, and you end up narrating a story about what the shapes are doing. In all these scenarios, your brain lights up its social machinery, and things around you become characters with motivations. Our brains are always eager to find agency and intention, to find minds that it can predict. So if our brains are so ready to see minds, how do we ever fail
to see them? Now you've heard me talk about this before on the podcast, because history shows us over and over that humans have the capacity to stop seeing other humans as having an entire cosmos going on on the inside.
We can treat other humans as objects. And when you look at any conflict, Let's say we're talking about the communist revolutions in China and Russia, or we're talking about Nazism in Germany or fascism in Italy, or the camer rouge in Cambodia, or the Hutu taking up machetes against their neighbors, the Tutsi in Rwanda. Wherever we look, we see that it is possible for people to look at their neighbors and feel nothing, to feel no tug of empathy,
no sense of shared humanity. Now how does that happen. Well, it's typically helped along by propaganda that trains people to see their neighbors as as vermin as contagion, as something that can be crushed with no moral cost, the way you might treat a bug. So we essentially have this dial in the brain that makes us see a thing in the world as a person within inner life or not.
Today's episode is about that dial.
It's about the neural machinery that supports social cognition and what it means when that machinery gets dialed down. It's about dehumanization, which is when the brain does not engage its social machinery for considering another person's mind.
And as we'll see, a slight.
Dialing down in the brain can have big consequences for our behavior. And we're also going to talk about the mirror image of this anthropomorphism, because in our new landscape, we have all the things around us growing mind like in their behavior. So what happens when the cues that trigger mind person become cheap and ubiquitous. We're going to cover all this and much more today. My guest is
Lasana Harris. He for many years has been at the forefront of mapping all these questions about humanization and dehumanization onto the brain. He's a social neuroscientist and an experimental psychologist at University College London, where he leads the Boundaries of Social Cognition Lab. Lasaana studies how we perceive other minds and how this connects to moral behavior and intergroup conflict and the emerging world of AI agents. Here is Lasana Harris. So, Lasana, let's start with a little bit
of an origin story. What first drew you to the issue of understanding how we perceive other minds?
I think I was always fascinated in how other people experience emotions and how a lack of emotion regulation effect social interactions. Once I got into graduate school, I realized the emotion literature was a mass, and so I had the sort of crisis point. Lots of potential PhD students had where I realized the thing I really wanted to study seemed impossible to study.
At that time.
I happened to take a social cognition class where they talked interestingly about anthropomorphism, some of the very classic hydro and similar research where you had these geometric shapes colliding into each other and chasing each other and people were bringing them to life, and that really fascinated me.
Let me jump in for one second for any listeners who don't know. Heiner and Simmer were two psychologists who made a little movie where there's what was it? It was a triangle in a circle and then another bigger triangle and they're moving around.
The shapes are moving around.
There's no sound, but when viewers watch this, there's a whole story that they on to what's going on.
It looks like a love story.
And then the bigger triangle is a bully who's trying to break up their relationship. And by the way, listen, define for us social cognition.
Sure.
Social cognition is how we consider other people's minds, essentially in the shortcut, so it's the psychological tools that lets us interact with other people. So if I want to have a conversation with someone. In order for that to happen, I have to have some inference about what it is they're thinking. Without that inference, it's near impossible for me
to have that conversation. So this is an ability we have that we use all of the time constantly to not only figure out what other people are thinking, but to also predict what someone might say or do in a given situation. So it's crucial for any kind of social interaction. So what fascinated me about those cartoons from the nineteen forties of shapes moving around is that we would use this ability for something as simple as shapes
sort of acting in funny ways. And that really juxtapose with me when considering processes like dehumanization, where you have this failure to engage these processes. I was struck by how it is that something that seems so prevalent, that happened in such a benign environment wasn't actually coming to bear when people needed it the most. Right, So, when I actually have someone who might be suffering, who might be having some type of negative experience, why was it
that we couldn't seem to get this mechanism going. Yet we watched some shapes running around in the screen, and suddenly we imagine this whole complex story about their lives, right, And so that juxtaposition is really what got me hooked on the stuff that I study.
So tell us about dehumanization and give us an example of that.
Sure, So, dehumanization is one of those psychological terms where everybody has an idea what it is, but psychologists tend to think of it a little differently. So for most people, dehumanization this horrible thing that only happens in cases of human atrocities, and we've never been able to get any evidence that's actually the case. Right, we can't go in the lab and do unethical things to people to know
that dehumanization is present. So for a long time there was all of this theoretical philosophical work about dehumanization and human horribleness, but as psychologists, we tended to define dehumanization a lot more simply. We basically said, when you encounter other people, you tend to spontaneously get these social cognitive processes going, right. You tend to spontaneously think about what's going on in their minds in order to interact with them,
to understand them. Potentially to have empathy towards them. Dehumanization then was a case where you didn't get this process going in the presence of another person. So you encounter another person and instead of that person triggering this psychological process, there was an absence of it. So it was a very simple definition of humanization, what we like to call not your grandpa's definition of dehumanization.
Right.
It's the definition that is detached from the human atrocities, and it makes it much more of an everyday process. And a lot of the work we've done in the last twenty years or so really has established it as an everyday process which can be used to do horrible things, right, but so can lots of other psychological processes.
And so one of the things that you did is you looked at what was happening in social psychology, and then you asked, what is happening in the brain?
So tell us what you found when you did those studies.
Sure, so, in the brain, there's this large network of regions that supports us figuring out what's going on in other people's minds. These brain regions are mainly in the neocortex, that's the more recently evolved parts of the brain. So the parts of the brain that separates us from other species, and what you tend to see is that these brain regions are tonically active, which means if I suddenly took a picture of your brain, now, I'd probably see activation in that large network.
What we did is.
We had people look at other dehumanized targets. So these are groups in society that you wouldn't typically humanize, that is, you would treat as more like an object than a human being, homeless people being one example. And what we found was that this network was not engaged when our participants just looked at pictures of these people. And I was shocking to us again because shapes running around in a screen can sort of trigger some of this engagement as well.
And so the.
Brain research was the place we first discovered this effect because we found this failure of these networks to engage. Now, at the time when we first made that discovery, almost every study in social neuroscience got these brain regions coming on right because whenever I stuck someone in an MRI machine and I showed them a person, I had them think about a person, I had them think about themselves,
you would see this network lighting up. But for these targets, right, like homeless people, you didn't quite see the same pattern of activation.
So just give us a slight bit more color on the experiment. People are in the scanner. You're showing them pictures of other humans. But some of the humans were let's say, athletes or successful businessmen, some were homeless people or drug addicts. And so give us a sense of what you presented and what you saw.
Yeah, that's exactly right.
So participants would lie there, we'd flash these pictures up, these what we call stereotypical representations of social groups, So a picture where you look at it and you know exactly who it is you're looking at, be it an athlete, a college student, a business person, or a homeless person. And what we would find is that for all of the other social groups, the ones that aren't typically dehumanized,
you'd get engagement of this network of brain regions. But when it came to the traditionally dehumanized social groups, that is, the homeless people or the drug addicts, you wouldn't see this naturally engaging in quite the same way. So this absence of engagement is where we've sort of picked up in this dehumanization response.
And So when I first read this work, first of all, I thought it was amazing for several reasons. One is that this is different than simply disliking somebody. This is actually the networks that understand that person as a human. These are diminished.
They're absolutely right. This is not dislike or simple prejudice. For instance, So let's take prejudice as an example. Let's say I hate a particular racial outgroup. I would see an activation in the brain that's correlated with sort of a threat response, because you tend to feel threatened by the groups data.
I'm not the old group, right.
People that belong to a different political ideology, have different opinions, come from a different racial or ethnic group. You tend to feel threatened by them, and there's a very clear brain response in those cases. That's a typical prejudice response. This is not that. This is not just a dislike response or prejudice response. This is a failure to process as a person, because when you see people, you tend to turn on these networks so you can understand something
about their minds. And that's really important, right, because interacting with a person is very different from interacting with an object. So if I see a table, I'm not wondering what the table's thinking. I'm not wondering what does the table think about me. I'm not guessing at the intentions of the table. I'm not doing any of the stuff I do with other human beings. But even if I see a stranger, I'm going to have those thoughts, right because
I might have to interact with that person. That person might be a potential friend someday in the future, and ouri they may have information I need. So when you see other people, we tend to always engage these networks because it's useful. It gives us information that facilitates any interaction I may have with them. So to see human beings and to not even switch these processes someone was stunning for us as well.
How do you think it happens that we can go from engaging these networks with other people to not doing it anymore. For example, let's say in Nazi Germany, when people had Jewish neighbors and friends and then things changed, or in Rwanda where you had intermarriage and friendships between the Hutu and the Tutsi, and then there was lots of propaganda about the Tutsi being like talkroaches and then these things, presumably these networks turned down.
How does that happen?
Yeah, that's the million dollar question, isn't it. I think, as you mentioned, propaganda does play a role in those cases of genocide and human atrocities, because what the propaganda does is it puts the idea in your head that this person isn't quite a human being in the way that I am a human being, right, And so you might have interacted with them before, but now the propaganda suggests to you that there's something about their fundamental nature
that's just different. And it turns out that we still have a lot of beliefs about other human beings not being human. This is something that has been kicking around science for centuries, right. So there's a lot of what you can call of dustbin science, where scientists were trying to prove right, members of different groups weren't quite human like the human beings of interest, right, people like themselves.
And so these ideas are really old ideas that have stuck with us as a society and as a civilization. And so once the propaganda makes it salient that that particular group might not quite be human in the way that you're human It's not impossible to think that you can now switch off the network. But that's the sort of kind of answer where we have no evidence. The kinds of stuff we do have evidence for suggests that you switch off these networks because it serves a particular function.
It gives you a particular benefit in a moment that.
You want to enjoy.
So let me give you an example of some of these benefits. Let's take the case of the homeless person. Let's say you live in a big city and you're on your way to work, and you're rushing there. You're worried about the important meeting that you have, and you come across a homeless person sitting at the side of the road. Chances are that's not the first homeless person
you've seen today. If you stop and fell sorry for that homeless person and every other such homeless person you see, you would probably not get to work.
Right.
The amount of human suffering in the world is overwhelming, So in order to focus on the tasks at hand and do the things we need to do, we sometimes shut out that suffering because it's useful for us, right, It's useful for us not to be delayed as we rush about our daily lives. That's not a novel finding. We've known that since the nineteen seventies. Research and the bystander effect, for instance, demonstrated that. So the buyside under
a fact. There's a really fun set of studies where, in the most famous case, they brought in a bunch of theology students, people practicing to be pastors and priests, and they had them think about the Good Samaritan and come up with a samon around the theme of the Good Samaritan. So they're thinking about helping people, and then they tell them, oh, we're sorry. This room booking we have with the experiment, we lost it, so we have
to switch you to another room across campus. But we're running out of time, so could you hurry over there. And these theology students hurrying out the building came across someone lying on the street who seemed to need help, and very few of them stopped to help this person, even though they were thinking about helping as they were doing it, because again, it would have been inconvenient in that moment to help, right, you're sort of focused on
your task at hand, which is rushing across campus. And so I think that's one of the reasons you might dehumanize someone because it saves a particular function. Reason you might do it is because if I start thinking about you as a human being, that brings with it a bunch of moral obligations, and I think this is what might happen in the genocide cases.
So when I.
See you as a human being, there's a bunch of rules that come with how I treat human beings. And one rule says that I empathize with you if you're suffering, right, I try to help you if I can. That's what human beings tend to do with other human beings. It's in our very nature. In fact, some researchers think it's why we evolve to be the species that we are, because we're so helpful to other human beings. But I may not have the capacity to be empathic towards you, right,
Maybe I'm emotionally taxed, maybe I'm drained. And this is the sort of evidence we see when you look at medical professionals, for instance, Right, they're dealing with suffering all day. If they felt terrible for each suffering person they encountered, again, they wouldn't get through the day, And so dehumanizing some and shutting down that part of the brain allows me to not have to expand those resources and empathize with you,
and I can go about my business. And now the morality isn't salient, So I don't have to feel guilty about not helping you, Right, I don't have to worry about the fact that I might see myself as a terrible person.
Always, Yeah, you might think of.
Me as a terrible person because I didn't even process you as a person to begin with, right, I didn't get those parts of my brain going. And so we have a lot of evidence for those kinds of explanations for why dehumanization happens. But again, we can't study it in the genoci context because that's an ethical so we do things like liquid medical care professionals look at cases with homeless people, that sort of stuff.
This argues what you have argued is that social cognition is actually quite flexible, and what are some of the other key factors that dial it down.
Another key factor is that feeling of negative moral feelings. Right, So if I feel like I've done something horrible to someone, or I feel like my group has done something horrible to another group, and I don't want to feel that guilt. An easy way of getting rid of it is by saying, well, those people didn't really suffer because they're not quite human
in the same way. And so you often see with our groups who are subject to a lot of suffering, narratives pop up around their capacity to endure that suffering, for instance.
Right, so you.
Get these stories about some types of groups being superhuman, for instance, and having a great capacity to deal with pain and suffering. So the actual suffering the experiencing isn't that bad because they're not quite human. Right, So if it will me going through what they went through, I would feel horrible. But because they're not quite human the way I am, it's not really that bad. So I don't have to feel that terrible about the horrible thing
that I did. So I can sort of dehumanize as a way to protect myself in the face of evidence that I haven't been a great person, for instance.
From your perspective, what are the main roots the main psychological roots to dehumanization. Is it things like unfamiliarity, or is it perceived immorality, or is it the borders of our own in groups?
I think they are multiple roots to dehumanization because it's a very useful tool for a range of situations. So these days we've been doing some interesting work. I'm going into caring domains beyond just medical professionals, and we're looking at parents and their kids, for instance. And so imagine you have a five year old and you've already given you a five year old some treat for the afternoon, and they've come back to you begging for more, tears
streaming down their face. What that kid is doing is tugging at your empathic strings, right, You're trying to get you to feel sorry for them, so you do what they want you to do. If you want to be a responsible parent, you have to somehow shut that out right and find a way to stick to your guns and tell them that's the limit for today. And so doing that means you have to ignore their perceived suffering, right because the tears streaming down their face makes it
seem like they're suffering. And so having a flexibility, having a tool that lets us shut that out and say nope, that's final, allows you to stick to your goal in the particular case of taking care of that kid. So we've been interested in cases where dehumanization needs the benefits. For instance, right, because I've shut out your mind, it allows me to do some stuff that I wouldn't have otherwise been able to do because now I'm not quite
processing you in the same way. That's exactly what we see with domadical professionals who are actually.
Trained interview human beings like that. Right.
They're taught to view the body as that machinery, a biological machinery, and the medico curriculum really pounds that in, and so they learn, both sort of explicitly and just true practice that they need to shut these people suffering out. And shutting out the suffering requires you to shortsake at these social cognitive processes.
I want to return to this question about, let's say, on a larger political cultural level, what are the things that make it easy to dehumanize another groups? So the examples I gave you know, I'm unfamiliar with them, I just don't know enough about them, or I perceive them as immoral in some way, or I have boundaries to my group, you know, my in group and my outgroup, and they are clearly on the other side of the outgroup.
What are the important things and what else do you see is the roots to dehumanization.
I think in the end to group space, we have to be careful because we have these ideas about dehumanization doing a lot of heavy lifting when we don't have a lot of scientific evidence for it. And I think that's because there are multiple processes operating. So if there's a group that I'm unfamiliar with, yes, unfamiliarity might promote dehumanization, but unfamiliarity can also permit a desire to learn more about this particular group, right, So it can go both ways.
I think what sort of helps is when you have a combination of these processes coming together, then dehumanization can be useful. So let's say that this outgroup you have in the past has done some wrong to you, so you know that they're established as a particular threat. Here, dehumanization is not going to be helpful because if someone is threatening you, you kind of want to know their intentions, right, You want to know what it is that they have
in store for you. So dehumanization on its own isn't going to do the work. Now, let's assume that the same our group that you view is threatening, you're able to do something about it. So now I've done some horrible thing to them to mitigate the potential threat. When I reflect on that behavior, I might feel terrible. Now, dehumanization becomes useful to shut out those negative feelings I
have around the thing I did to the group. So I often tell people I don't think dehumanization motivates things like political violence. I think we have other psychological mechanisms that are much better at getting us to be cruel and violent to other human beings. But I think dehumanization is able to sustain it, right, because what it does is it shuts out the suffering of the group that you've now done something horrible too, so it lets the
violence keep going. So for me, dehumanization is really problematic when you're already in a bad place. So if there's already animosity between groups and you get political violence going, now I worry about dehumanism because the horse has already left the stable, so to speak, and dehumanization is going to keep.
That violence going.
So I don't see it as the motivator.
Right.
The analogy I often use is, I don't think dehumanization is going to cause you to pick up a machete and go to your neighbor's front door. But when you're in the process of kicking that door in dehumanization might be useful, right to shut out their mind when you're doing the horror black We have no evidence for that scientifically. But then afterwards you have to live with yourself, and here's where dehumanization is particularly useful, because it allows you
to live with the things that you've done. Because you've now viewed the other person as not quite human, morality goes out the window. So in these cases of integroup conflict, we tend to focus on dehumanization because it has a bad reputation and it is playing a role, but I feel like it can often mask some of these other psychological processes like threat, for instance, which is really what
drives a lot of the violence and the animosity. But they were hand in hand, right Our job as psychologists is picking these things apart, but in reality, they often occur in sequence, they co occur, and that's the part of it that we're working actively to figure out.
I totally agree with your intuition on this, because you know, when I look at a squirrel that is dehumanized, for me, it's more like an object.
But I have no desire to harm that squirrel. But if the squirrel were.
Rabid and charging at me, and I felt a threat and so on, that might lead me to feel like I need to do something in a way that I wouldn't feel terrible about.
So I agree with you on that.
You have other studies about putting a price on people and what that does in terms of these networks that are involved in dehumanization.
Tell us about that.
Yeah, So, once we did the classic dehumanization, we're GRAVI loked at it in a group context. I really wanted some evidence of this being more of an everyday thing. So I didn't just I didn't imagine dehumanization just evolving and functioning because we have to deal with our groups. I thought it might occur in cases where you didn't just need to process people's minds for a host of reasons, and one reason might be because you're actually outcome dependent
in someone. So what we did is we looked at a labor market context, and to do this we used essentially a fantasy football league.
So the idea really came.
From an interview I heard an NFL player give many years ago, where he was injured and he couldn't play, and he received all of this horrible abuse from fans online because he couldn't play, and a lot of the fans were abusing him because their fantasy teams would suffer, right because he wasn't going to make any points for them that particular week. And he said that he felt
very dehumanized. Now, no one's going to shed a tear for a very highly paid professional athlete saying that they're dehumanized, But it did suggest that there's something about being in a labor market that might promote that kind of outlet. So what we did in our studies we essentially created a fantasy league, a fantasy time estimation league. We didn't have athletes available. We took regular people off the street and sort of put them in a league where they
had to guess different intervals of time. We then brought other people in, gave them some money and had them purchase some of these people to be on their team. And then these people they purchased, these players would go out and compete, and if they won, they want money for the owners, just as it works in sports leagues or any other type of labor market. And what we found is that the owners, the ones who had purchased.
These players were dehumanizing the players, but just the ones that they had purchased, not all of them, right, because the ones they had purchased they are now outcome dependent on. What matters for them winning money is that they get these guesses correct, just like that football player.
Right.
What matters is that you go out there and you perform. I don't care what's in your head. Is win me any money? I don't care what your intentions are. All that matters is that you're competent in the thing that I'm paying you for, essentially, And so we found some evidence that in these labor market context right, people become sort of cogs in a machine. Now, again, there's not
a lot of novelty there. People like Marks have been saying that for quite some time, but it's fun to get some brain imaging evidence for it as well.
Do you find that certain people are more or less prone to dehumanization of others?
Absolutely, and it goes in the opposite way you might think. So the people who believe that they are very good people and they never do anything wrong, those are the ones most likely to dehumanize others, because those are the ones that have this need to protect this idea of
themselves that they are a morally good person. If you accept that sometimes you're good and sometimes you're not, then you don't tend to dehumanize as much because you're not as invested in protecting this self image you have of being being a morally good person. So it's actually an ironic effect whether people that think of themselves as being very good people are the most likely to dehumanize.
At Lisnawa Resarch.
You know, this is really interesting because so I did a study years ago where I was looking at in groups and outgroups and essentially you're looking at hands getting stabbed on the screen, and they all have a label Christian, Jewish, Muslim, Hindu, Scientologist, atheists, and depending on your in group, you have a larger response. And let's summarize this as the pain matrix to your own in group hand getting stabbed then any of your outgroups.
And this was true across everybody, including atheists, by the way, who care more when they see atheist hands getting stabbed. But we also did questionnaires about how people saw themselves in terms of their empathy, and we actually found something very similar, which is that the people who felt they were more empathic actually had a larger difference between their in group recons response and their outgroup response. This very
low level neural response. So one interpretation that we considered was that maybe when they're being asked the question about empathy, they are thinking about their in group.
You know, how would you.
Feel if you saw someone twist their ankle and fall off the sidewalk. Maybe they're just thinking about their own in group, and that's why they rate themselves as empathic. But this is very interesting what you're saying, because you found something similar here that people's responses seem to run counter to what's actually happening in their brains.
Absolutely, and I think you've seen evidence of this in other places as well.
Right.
So I have a collaborator that studies compassion. He gets a very similar result, Right, the people that identify as the most compassionate and the least likely to engage in compassion. You see it also with things like racism. Right, there's a really interesting study where people that say, well, I don't like black people, for instance, I'm the ones that show huge implicit bias scores. This is a measure of
people sort of subtle racist beliefs. But the ones that say they love black people and they have a bunch of friends and they egalitarian, they're the ones that tend to show these bigger differences. So I think it speaks to a commonality about some of these psychological processes and how they function.
Right.
That is, if you view yourself in a particular way, you're invested in protecting that view of yourself and that might obscure some of these other biases that you might have. Now, saying that has become very unpopular because it seems to be an attack against people that who are very positive views of themselves. But the goal here isn't necessarily to criticize those people. It's simply to make people aware that their self perceptions have an impact in the psychological processes
that they're able to then display towards others. And so it's really important, I think, when we think about ourselves to be honest with ourselves, right, and to realize that we're just here and we're going to have good sides and bad sides and that's okay, right.
Yeah.
So is it possible that the neural responses we're seeing, at least in some cases, are rivalries between different networks, some of which are saying, Hey, I'm actually feeling this way, but I perceive myself that way and I like to present myself socially that way. And what we're seeing is a big response because of this conflict.
Quite possibly, quite possibly, it's hard to rule that out. I think of it in a very straightforward sense, and the fact that what's really important to us as human beings, one of the many goals we have is to maintain a positive self image and a positive view of ourselves. If you don't have a positive view of yourself, you tend to slip into mental illness, depression, those sort of issues, and so it's really important we maintain that, and there
are lots of ways we can do it. We do it through groups, right, if we belong to groups that are prestigious. Those prestigious groups allow us to sort of for any negativity about ourselves, which is why everyone puts on their team's football t shirt when they're winning, right, And so you want that positivity that comes with it. This desire we have to maintain this positive view of our selves can now get in the way, right, So it gets in the way of our typical psychological functioning.
So if I've done something horrible. I want to feel better about myself. An easy way to do that might be to engage in some of these processes we've been talking about. So I always give the anecdote of let's imagine that you have a house that you would like to sell. So you're living in this house, you decide you need to move.
You want to sell.
It till happens your best friend is a realtor. Now they're your best friends, so you know that they're a terrible realtor. They never get asking price, they always complain about it, They worry about their job security because of it. Do you go to your friend and say, can you sell my house?
Now?
If you're thinking about the friendship and the way your friend might view you, you might say, Okay, I'm going to go talk to them to sell my house, knowing you'll make a huge loss. But if you focus on the fact that I need to make a profit from selling this house so I can buy the next one, you'll probably keep that away from your friend.
Right, you might go.
Elsewhere that doesn't make you a terrible place or a terrible friend. Right, you're doing the thing that sort of sensible and rational. But if you hold this view of yourself as being a good friend, right, that can eat you inside. Right, you're probably going to be led as straight. And so oftentimes these self perceptions we have leaders astray lead us into paths that are suboptimal for ourselves but also people around us as well.
I want to return to one point before we move on to the next thing, which is you mentioned that dehumanization may not be sufficient for groups of people to take violent action against their neighbors. What are the psychological factors that are involved in that, because threat obviously is one.
But you know, when I think about dehumanization, probably for many of us, we think about things like just as one example, there's this famous picture of a Nazi soldier who's got his rifle aimed at a woman holding her baby, and he's about to execute them both, and the woman and baby are obviously terrified he's about to shoot them. His colleague took the photograph and he proudly sent this photograph back to his family. And the only way that
sort of behavior is possible is with total dehumanization. But clearly the woman and the baby are not a threat to him. So what are the psychological factors that allow for that sort of violence.
Yeah, I think it's hard because there's a timeline here, a timeline element that we never really talk about. So it could be that that soldier who has taken the picture and his friend have already gotten to the point where they've dehumanized the enemy. And this is something that we see a lot in police forces, in militaries, where as part of your training, much like the medical professionals, you're taught to view the enemy in a particular sort of way, again, because it makes it easier to engage
in these types of behaviors. Now that doesn't mean it's motivating you to shoot the person, right, but it means that upon shooting the person, you don't have the negativity that would typically come from it. And so sorting out that timeline is one of the challenges that we haven't quite cracked in the psychology literature. Which process comes first
in that task? Gape there's a wonderful book written by a historian Browning is his last name, where he interviewed a bunch of people at Nazi dead squads and he basically asks them, why did you do some of the horrible things that you did? And for a lot of them, their story starts with threats from the Nazi regime where they had to enlist and they had to engage in these behaviors or their own families were threatened, so they got into these behaviors initially for their own sort of
self protection and preservation. It's essentially because they were threatened. Once they got into the behaviors and they started engaging in these behaviors, now they needed a mechanism to keep going, to keep doing these behaviors, and that's where dehumanization is handy.
Now.
A lot of people in those interviews also said, well, I always miss my shot, right, so I never aimed at the people. I always shot outswhere, knowing that in a firing squad somebody would probably shoot them didn't have to be me. So not everyone was going directly to this place where you're now dehumanizing. But in these contexts, dehumanization is very useful because it allows you to keep engaging in behaviors that you may have started because you
felt threatened. But now that you've already engaged in them, you need a way of living with yourself. There's a time course element where dehumanization comes later. It's not the motive right at the front. That may just seem like
an academic distinction, but I think it's really important. And you think about prosecuting people for war crimes, for instance, where they view dehumanization as demotive, but you can't really get evidence of that because people in the Dutch squad say I didn't do it because I dehumanize them.
I felt like I have no choice.
Well, let's take something like, for example, in America and in many places in the world, we happen to be in a very polarized era and there is violence that keeps cropping up. This violence doesn't happen because somebody feels that they're being recruited and the government is threatening their family. This is happening for other sorts of reasons. I often think about in group and out group issues. But what do you see as the main psychological dribes that allow violence.
We agree that dehumanization might be a later piece.
I think part of it is identity. So people in a particular group may feel that they have to engage in certain behaviors to be a good member of that group, and that's one thing we can't ignore. So sociologists have done some interesting work looking at propaganda and the themes in propaganda, Right, the kinds of things that propaganda's talk about. Dehumanization is present, but it's not the most popular thing.
What ends up happening is a lot more talk about threat, obviously how threatening the other group are, but also a lot of talk about what you do as a good member of this group. Right, if you are really a good member of this group, these are the kinds of behaviors you engage in. So I feel like a lot of the political violence you see in the US is motivated by identity, right, people believing that as a good member of this group, these are the kinds of things
I should do. So, for instance, why would I sign up to an organization like ICE. It's not because someone has forced me to do it. It's because I feel like, as a good American, there's this real problem in the country around immigrants, and I should do something about it if I'm able to. And one way I can do that is joining this organization which is working towards.
Addressing this problem.
So those are identity issues that get you signed up in the first place. Now, when you're out in the field and you're happening to be engaging in violence because you're caught in that particular situation where violence becomes necessary. Maybe you feel threatened at a riot, for instance, how do you make sense of that. That's when I think dehumanization is handy. So now I have to explain why I've done these things that my identity says I must do.
It's because they're not quite people in the way that way people. And so you often see in the propagandis rhetoric dehumanization occurring, but strong messages around identity that really gives you information about what a good member of this group does, and strong messaging around the threat that the art group presents as well, right, which potentially motivates you to feel like you have to defend yourself open your country.
So I think the multiple psychological factors that go into these kinds of political violence, and I don't think there's one that we can point at and say it's the most critical one, because they're all having a role at different points in the cascade.
Okay, now I want to switch gears to talk about anthropomorphization, which is the flip side of dehumanization. We often attribute minds to things that don't have them, like our pets or characters in the Pixar film or the moving shapes that we talked about earlier. Why what's going on with anthromorphization and what does this have to do with dehumanization.
Yeah, so that juxtaposition between the two is what I find really fascinating. So anthwer momorphism occurs because it's so useful to get these mechanisms going, Right, Like, when I think about what's going on inside of your head, it allows me to explain your behavior. Right, I can attribute it to your personality, or your psychological mood or your emotional state. Right, I can say you threw that chair across the room because you were upset, and that gives
me a reason to explain your potentially erratic behavior. But it also lets me predict your behavior. It lets me say what you're likely to do the next time you're in a situation where you potentially get into that psychological state. Right, you're the kind of person that flings chairs about. That's really useful, and we want to use that when we encounter things in the world that we might want to
explain and predict the behavior of as well. So I think you saw a lot of anthropomorphism throughout human history. When it came to things like weather. Right, if there's suddenly a storm or a drought, we usually pray to the race God so that there's more rain, because it helps us explain and predict the occurrence of the weather. And we still do it right. We anthropomorphize hurricanes and give them names and talk about their behavior as if
they were people because it's a handy explanatory mechanism. So we have this psychological process in our head that gives us explanations and allows us to predict.
Stuff that's really handy to you.
So if I'm interacting with my pet, of course I'm going to infer a mind there one because the pet actually has a mind, And pets are a gray area for us for exactly this reason, because there is a mind there. But even when there isn't, for instance, when I'm talking to chat GPT, right, it's better for me to think of it like a person and sort of triggered the social cognitive processes because I can better understand it and predict what it might do. And that's really important.
As I move around the will interacting with stuff that seem to have minds of their own.
And so you recently wrote a review where you looked at the way we've you humans and the way we view AI agents, and what do you find there and what's interesting and surprising given the increasing presence of AI in our lives.
Yeah, one of the early in my career philosopher once gave me a challenge that he said, if you're anthropomorphizing, don't you still know the thing isn't actually a human? And it turns out he's right. When the brain anthropomorphise, it tends to use a slightly separate network so that you can still maintain that distinction. Right, because even when you're anthropomorphizing, you're still not treating it quite like a
human being. So I could play against a computer in chess, and if the computer is beating me, I could unplug the computer. Right, I could do something that I wouldn't do it another human being. Right, if a human being is beating me, I'm not going to upset the chess board. That's horrible. That's going to damage my potential reputation because when I'm interacting with the human, I'm also worrying about
what does this human think about me. When I'm anthropomorphizing the computer, I don't care what it thinks about me, Right, that's irrelevant because it's not a person. It's not going to tell anybody that I'm a horrible loser. And so there are these such a differences that occuld for things that are human and things that aren't human that the brain still preserves as it's anthropomorphizing. So it's using some of the mechanisms, some of the social cognition, but not
all of it. If it were a complete overlapping processes, we wouldn't be able to tell the difference, and then there'd be much more talk about the rights of pets and the rights of AI. But we don't have big conversations about the rights of AI because we don't quite
see them as human beings. Despite the fact I can have an hour long conversation with Alexa and tell her about all of my problems and asks for advice, right as I might with another human being, at the end of the day, the brain knows the difference and preserves it. The worry for me with your question about the prevalence of AI is whether that distinction is eventually going to
go away. So will future generations of humans who have now been interacting with AI as long as they've been interacting with people suddenly lose that distinction, right, and suddenly will we see a rising cases of people fighting for AI rights right. That to me is fascinating, very sci fi stuff, but a potential possibility given how plastic de brain is and how it's willing to adapt or circumstances.
So these are the kinds of things that I think about in the AI ethics space that I don't see tons of conversations about.
It strikes me that maybe part of the difference in the way we humanize another person and humanize CHETCHPT might have to do with all the consequences that we might envelop in this, like, oh, it's another person, so it's a repeated game where we're going back and forth and their legal consequences if I kill my chest opponent as
opposed to unplug the computer and so on. So as long as those other consequences remain the same as in I'll still get busted for hurting a human, but I won't for a computer, then you know, maybe maybe we won't come to ANTHROPOMORPHISEI in the same way.
You have great faith. You have great faith in the law. I have less faith in its ability to drive behavior. I think reputation is a big one, right, So I think, for instance, why do you why are you polite to strangers? There's no reason for you to be polite to a stranger. We live in a world of eight billion people. Chances are that stranger will never see you again, We'll never know who your friends are, won't be able to tell people that you were impolite to them, therefore affect your reputation.
But we evolved in very small groups of human beings where reputation was paramount, right, it really mattered, and so we learned to treat other people nicely whilst there were damaging consequences for our reputation. That has somehow held on in our brain, and we still have this belief that when we encounter another human being we have to treat them in a particular sort of way to manage potential damage to our reputation. That's not a concern we have
with AI, for instance. But what could change is if AI now goes about spreading information about your reputation. So if every time my interaction with Alexa was recorded in her memory bank somewhere, and the next time somebody used Alexa it said, you know that guy David Eagelman, He.
Was really mean to me.
I think we will start caring about, right, what Alexa thinks about us, And so I don't know if the legal changes will drive us to have the neural changes, but I think it's an issue of what is the technology capable of and how it's being used that's going to drive these changes.
Well, great, and this is a great segue to something.
You have argued that legal systems should explicitly account for dehumanizations. So what's an example of one policy that you would like to see Given your expertise in the neuroscience, what's something you would like to see changed in our legal system.
I think I've spent a lot of time talking about the stories that we tell with in a society and the things that are allowed to be said. So we have a censorship system, right that doesn't allow us to go on television or make media that does particular things. Usually it's a wrong swearing and sexuality, things coming from our Protestant past that are holdovers. Nobody cares about those things today in quite the same way. I think we're
focused on the wrong stuff. I think we should be focusing on stories that promote dehumanization of particular groups because what it does is it facilitates any kind of violence that might be occurring against them. So if you say a particular group is eating dogs and cats, for instance, that's a dehumanizing image you've put out of that group.
That's very dangerous to me. And that's the kind of stuff I wish legal systems would pay more attention to, because those stories, whether you believe it or not, whether you endorse them or not, they get into your heads, and your brain is this powerful machine that's taking in all of this information and holding it relevant. When you're in a situation where that information is potentially useful, that is, it can facilitate dehumanization of that group, it's going to
kick in, and that's what I really worry about. So I really am one who promotes not necessarily a lack of freedom of speech per se, but the kind of careful monitoring that we use for domains that quite frankly were relevant three hundred years ago that I think aren't relevant today. So I would like to see less attention to swear words, for instance, and more attention to dehumanizing rhetoric about groups. Because even though someone says, oh Wes
said it ingest, it's a joke. Your brain doesn't register that it makes that association, and that association can pop out at any point. It's convenient to influence your psychological processing. So that's one of the sort of policy legal things that I've been sort of promoting for for quite some time. And that doesn't mean you take away the ability to
say those things from people. You at least give people the option to know that this content they might consume does contain this kind of messaging, so they can make a conscious choice whether they want to be exposed.
To that or not.
And we do that already with our rating system, right we say this is for mature audiences because it has sexually explicit language or scenes in it. Why can't we do it with things that promote negative stereotypes about groups, for instance, or has them in a dehumanized like And this was my pet peeve with Disney for the longest time, and then recently I noticed that Disney started doing that
right on some of their older programming. They now put up a warning saying there's a bunch of stereotypics representations in this programming. And so as the consumer, I can make a choice. I can say I want to expose myself to that or I don't, And I think that's a very simple change we can make that would have enormous consequences, positive consequence.
Now switching from legal system to individuals. So if a listener wants to reduce the amount of dialing down on other people that they're doing, what are some practical take home lessons that they can take away from this conversation.
I think depends on the contacts that they're in, but there are lots of strategies. And the case of homeless people, I always tell people just make eye contact, because that's the first thing you'll notice that people don't do right, and looking at someone's eyes gives you a lot of information about what might be going on inside of their heads. So if you simply looked at them in the face, that right there is going to make it less likely
that you will shortsake these processes. If you're rave enough, have a conversation, ask them a question, ask them what they were planning to do today, Right, Like, those very simple things that don't seem to matter much actually trigger these processes.
In a very rich sort of way.
In another context, and let's take the political violence context, where things are very polarized. I think another very simple thing you could do is instead of listening to other people's opinions or points of view, is finding the commonalities right, figuring out, well, what's the same thing that we have
in common. For instance, we're all Americans. We have a lot more in common with other human beings than we are different from them, and identifying these commonalities is often quite powerful for shifting how our brain process is people. And I'll give you an example of a case where
I think this actually happens. So a few years ago, we did a project with a charity in the UK called the Museum of Homelessness, and what they did was this wonderful performance are piece where they got a bunch of stories from homeless people about their daily lives, and they got objects that these people donated, And one object I really remember was a pack of cigarettes, and the guy who donated it said, oh, this is so meaningful to me because I need a coffee and a cigarette
to start the day. And that resonates with lots of people who also have a similar experience, and so considering that person has a similar experience to you is sufficient to now trigger processing of them in a way that's very, very different. So in the polarization context, I often encourage people to look for the similarities if you're finding with
members of your own family. While that's in some sense easy to resolve because you have so many similarities you could talk about instead rather than the differences and the stuff that you know is going to cause friction. And then in some cases, I don't think we want to get rid of the dehumanization. In the care context, for instance, I think the dehumanization is very useful. I think what we want to do is be aware of when we're
dehumanizing so it doesn't spiral out of control. So if you're a physician and you're seeing a patient and you're checking up on their treatment, that's when dehumanization is not relevant, right. You want to sort of care about them as a human being in that context, so you can ensure that whatever it is you're prescribing them is actually benefiting their psychological experience as a human being. You're not just treating them as a number on a sheet or a person
with a particular disease or ailment. When you're now operating on the operating table, dehumanization is useful there, right, because they're the broken machine. Analogy actually helps you get the task done. So just being more aware of the context in which we might be dehumanizing bass not, I think is powerful as.
Well excellent if you think forward ten or twenty years, what would you like to understand the most about these issues of what causes us to turn on or turn off onnderstanding someone else's mind.
The timeline is a big one for me, especially in the context of violence, Like I really feel like that's the next not we have to crack to really understand how these processes interact and how they can facilitate or inhibit violence. I think that's really crucial, especially in the time that we're living in. So for dehumanization, that's the big.
One for me.
The other one, which is not as big but also very interesting, is really trying to understand something about how it is that we regulate these processes. So I've given you a bunch of reasons that you might regulate it, but we don't have a lot of evidence in these heart circumstances where there's active violence or genocide occurring, political violence,
any of that stuff. So knowing how it is that some people are able not to engage in these behaviors when all of the forces are pushing you to doing it. It's really important for providing us some strategies that might help help the majority of people who fall victim to these psychological processes. So those are the two big ones for me that we're trying to work on in the lab. They're very, very difficult, of course, because again you can't do this work ethically quite well.
Right on, our.
Lab studies feel like toy studies compared to what happens out in the real will.
You talked about showing people.
Pictures of people getting their arms smashed, Right, if you did that in the real will and you actually witnessed that, they'd probably be so much more happening. And we're always going to be constrained in that way. But I think there are ways and methodologies who are developing to get around some of these hurdles where we can still do the research in an ethical way and answer some of these very important questions.
That was my interview with social neuroscientist Lasana Harris. We focused on this single superpower of the human brain, which is the ability to see minds in others. We have this machinery that lets us infer intentions, and that's how we predict behavior, and we build trust and we coordinate. All of that depends on this neural infrastructure that models other people as beings with interiority. This allows us to look at a face and hear a voice, and those
things become a portal into an imagined inner world. Now, Lasauna's work puts a spotlight on the fact that sometimes these networks fail to engage in the presence of another human being. When his participants viewed images of stigmatized social groups like drug addicts or the homeless, the usual mind perception machinery cranks way down. So you can see this in our everyday cognition, and you also see the issue
writ larger in history's worst crimes. The brain can turn this dial, and once we see this, we start noticing the logic that makes it possible. Part of it is just triage. Human suffering is infinite, and your bandwidth is finite. If you were to fully simulate the inner world of every person that you passed by, you would collapse under the weight of this. So the brain conserves and keeps
on trucking past most of it. And by the way, as we talked about if you are in a profession where you get repeated exposure to suffering, that makes full empathy psychologically very expensive. So some professions like surgery train a style of perception that focuses on bodies as systems and solvable mechanisms, because you have to dial those networks down to get the job done. Another part of the logic of dialing these networks around has to do with
moral self protection. If you see someone as fully human and that brings moral obligations, then dampening your mind perception can reduce your guilt and internal conflict. Lasana and I both talked about the irony in our studies that people who strongly view themselves as morally good or highly empathic sometimes show stronger patterns of dehumanization, and his interpretation was that the psyche has to defend its self image. And finally, one of Lasana's key points is about sequencing in time.
In his model, violent action gets driven by things like threat and in group out group identity issues, and dehumanization can sustain the violence the exact timeline of when each
process enters this cascade. This really matters for science and society because it points to different sorts of interventions is the important part about reducing threat perception, or reshaping stories about identity, or changing our media environments, or training awareness around the moments when mind perception starts to get dialed down. And we also talked about the flip side, which is anthropomorphism.
And this is timely because we are surrounding ourselves with synthetic entities that emit the cues that our social brains have evolved for. So Lasauna asks whether children who grow up with AI agents and robots will have a shift in their human machine distinction and whether that will lead to new moral intuitions and new political movements and new fights over rights. Finally, the question raised by today's podcast is what do you do as a single person walking
around in this enormous social world. Lasana offered some simple suggestions more eye contact, more conversation, more small acknowledgments that activate the mind perception machinery. As I've talked about in several episodes, search for commonalities as a way to reshape how the brain categorizes people. There are contexts where dampening mind perception can serve a function, but you don't want that to become your default stance. So the big picture
is this. Your brain is a three pound universe that constantly builds models of the world, and one of its most consequential models is its model of other minds.
Now, that model can.
Be richly detailed, or it can be thin, or it can be absent. Also, it can be projected onto pets, onto storms, onto chatbots, and it can be withdrawn from groups of people. And I think that if we want to go mining for the mother load of morality, much of it lives right there in.
The fidelity of the.
Model that you build of someone else's inner world. Go to eagleman dot com slash podcast for more information and to find further reading. Join the weekly discussions on my substack, and check out and subscribe to Inner Cosmos on YouTube for videos of each episode and to leave comments Until next time. I'm David Eagleman, and this is inner Cosmos.