
Sign up to save your podcasts
Or


Cloning is back, baby—and with it, our fascination. No longer carrying a thrilling frisson of disaster, it slowly appears to be becoming part of the domestic everyday. With last month’s announcement that his dog Junie is a clone of his late pit bull mix Lua, former NFL quarterback Tom Brady has joined Barbra Streisand and Paris Hilton among others in the celebrity trend of cloning beloved—and sadly deceased—pets. Brady explains that Colossal Biosciences, a company in which he is also an investor, “gave my family a second chance with a clone of our beloved dog.”
This is a booming business: the announcement came at a time when Colossal announced it has acquired Viagen Pets & Equine, a company claiming to have cloned the most animals in the world. While Colossal positions itself as a noble de-extinction project, Viagen appeals purely to emotion, with its glossy website advertising “the science of keeping love alive.”
Across the world, cloning pets is becoming ever more popular, with a price tag of $50,000 for a cat or a dog, increasing to $85,000 for a horse. Viagen describes itself as helping you “continue life’s journey with the companion who changed everything,” apparently forgetting that the one certainty in anyone’s life journey is death. “When love leaves a mark that deep,” it continues, “you deserve a way to hold on.”
Persuasion is on Instagram! Click here to follow us and help us grow our page. And while you’re at it, check out our X and YouTube pages.
These reassurances that cloning simply perpetuates the bond shared between owner and the original pet seem to be working. Although Viagen does not reveal how many animals they have cloned so far, it is certainly in the hundreds, and increasing each year.
The appeal is obvious. Pets are firmly part of many people’s families, and indeed are often given more attention and adoration than inconveniently distant or elderly relatives. Dogs in particular demand a daily routine, with walks and regular mealtimes in addition to vet bills and trips to the grooming salon, providing constant reminders of their loss. My Maltese, Maude, is only four and already I’m dreading the day I come home to find her toys and clothes but no little dog waiting for me. Who wouldn’t want their precious companion back, especially in cute puppy form?
Yet I’m cynical of the promise of pet cloning. It’s simply not true that clones are, in any meaningful sense, the same as the original.
I’m an identical twin—a natural clone. Identical twins are even more similar to each other than a clone is to its DNA donor, because they often share the same upbringing and environment. Yet, as I know first-hand, that doesn’t mean our personalities are the same. If you were planning a party that I couldn’t make and so invited my sister instead, thinking she’d bring the same joie de vivre and bubbly demeanor that I do, you’d be very disappointed when she turned up and sat in the corner as a wry observer—one who, for that matter, is 4cm shorter than the twin you wanted.
Twins are often celebrated as being mirrors of each other’s lives. Stories like the Jim twins of Ohio—who were separated at birth, met at 39, and discovered their lives resembled each other to a spooky degree, including both having first wives named Linda and second wives named Betty—set the tone for how we think about twins. “We even use the same slang,” said Jim Lewis when the brothers were reunited. “A lot of times, I’ll start to say something, and he’ll finish it!”
Thomas J. Bouchard Jr., director of the Minnesota Study of Twins Reared Apart Project, was certainly impressed by the Jims: “When you start to compound the coincidences, they become highly unlikely very quickly. In fact, I’m flabbergasted by some of the similarities.” The study found that “on multiple measures of personality and temperament, occupational and leisure-time interests, and social attitudes, monozygotic twins reared apart are about as similar as are monozygotic twins reared together.”
But this narrative conveniently glosses over a number of things. Linda and Betty were common names at the time, and one of the Jims had married a third wife while the other was still on his second. The idyllic view of “two halves of a whole” has been challenged on a more fundamental level, too. Researchers have argued that because the Minnesota Study of Twins Reared Apart relied on volunteers, it was more likely to attract twins with greater similarity. The study also didn’t publish raw data or full case histories, making it harder to examine the extent of the similarities. A further complication with twin studies is that, while researchers have assumed that identical twins and fraternal twins have the same upbringing, this is not necessarily the case. Identical twins spend so much time together that it is unclear which attributes might be genetic and which environmental or social.
What’s more, epigenetics—the study of how environmental factors influence the way genes are expressed—explains why it’s wrong to expect twins to be truly identical. As individuals in their own right, each twin will necessarily have different experiences, and make different choices about their lifestyles. They will face different stresses, each of which could impact how their genes work, despite having the same genetic makeup. One study found that 35% of identical twin pairs had significant differences in their epigenetic profile—particularly when the twins were older, which suggests that these differences increase over time. This may explain why identical twins tend to have different outcomes in life, contracting different diseases, marrying at different times, and even why they can have different tolerances to pain.
In other words, our DNA, including epigenetics, clearly influences us in many ways; at the same time, we are more than just our genes. Throughout history, philosophers and scientists have debated if nature or nurture is more significant—it seems the answer is that both have an impact.
I’ve always been outgoing while my sister has been more reserved, but it is impossible to untangle to what extent this is genetic and to what extent it’s a result of growing up together, adapting to the other’s strengths and weaknesses. Since a young age we have wanted to be seen as people in our own right instead of two halves of one whole. While there is a cultural expectation that twins are best friends fused together throughout their lives, deidentification with one’s twin seems to be a natural part of becoming an adult. For identical twins, it is a tremendous insult to be told you’re exactly the same as each other—and to be positioned as a replacement for a deceased companion, no matter how adored, would be even worse. Rather, to be respected as an individual is a much-prized gift.
An icon and a clone. (Photo courtesy of the author.)The same respect should be given to cloned pets—to be appreciated for their own personality and habits, not compared with an increasingly distant forerunner. Companies such as Colossal and Viagen promise that the past can be brought back; that a copy is enough to continue the bond with the much-loved original; that death isn’t an end. But to create a new dog and declare it will resolve the owner’s grief does not bring the pet back or undo its loss. An owner will eventually see the tiny differences add up, and the pet will feel it isn’t good enough without understanding why. Like Fitzgerald’s Gatsby fighting to change the past, pretending that a replacement is the same is futile.
Subscribe now
Genetics can be copied, but experiences can’t be—the clone won’t have the same likes or fears of the original, and the owner will be forced to create new memories with them instead of reliving old ones. Would my dog still be Maude if she didn’t recognize the word “brunch,” know all the names of both her friends and mine, or take the Tube with the studied indifference of a daily commuter? While similar experiences would be shared with the new dog, their reaction would depend on their own personality and associations. How could they be the same?
For the sake of both owner and pet, we should accept the finality of death. We feel deep love for those we care about in part because of our fragility; to love anyone is to acknowledge that there will one day be a goodbye. Any new life—whether a clone, twin, or dog—should be loved for its own self.
Leonora Barclay is Head of Podcasts at Persuasion.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
Abigail Marsh is a professor in the Department of Psychology and the Interdisciplinary Neuroscience Program at Georgetown University.
In this week’s conversation, Yascha Mounk and Abigail Marsh explore what to do if a child you know might be psychopathic, whether psychopathy is linked to charisma and success, and how to protect yourself.
Wondering if you or a loved one might be a psychopath? Take the quiz!
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: How should we think about what a psychopath—or do you prefer sociopath—actually is?
Abigail Marsh: That is a great question. There is this weird dichotomy in how people talk about the disorder. The term sociopath is vastly preferred in the media and literature. There are a couple of very famous memoirs that came out about people who were self-described sociopaths. It is not a term that is used anymore by clinicians or researchers. We primarily use the term psychopathy. I do not even use the term psychopath anymore, mainly because it occurred to me that we do not refer to any other people by their disorder anymore. We do not refer to somebody as “a schizophrenic” or “an anorexic” anymore. We have switched to a kind of person-first language, and I have thought that people with psychopathy probably deserve the same.
All that said, psychopathy is a personality disorder. It is not in the Diagnostic and Statistical Manual, the DSM, that is used to define and diagnose other personality disorders, for complicated reasons. It is a clinical disorder, and it is characterized by three main types of traits. One, a bold, fearless, dominant and socially dominant personality. Meanness or callousness is probably the trait people most closely associate with it: lack of remorse, lack of guilt, lack of compassion, etc. Finally, disinhibition, difficulty sticking to a plan, difficulty keeping promises, irresponsible behavior, sometimes very spontaneous, clearly poorly planned antisocial behavior that lands people in trouble.
Mounk: I once had somebody explain to me the difference between a sociopath and a psychopath as being that a psychopath is somebody who has, I guess in your language, all three of those sets of traits, who is disinhibited, who wants to dominate the world, and who ends up doing really bad acts. A sociopath may be somebody who does not have fellow feeling, who does not feel remorse if they have hurt somebody, who does not feel bad about things they have done in the world, but who may be relatively well adjusted, who may think that it is in their interest not to hurt others because they might go to jail or other bad things might happen, but is still lacking that dimension of fellow feeling.
You are saying that there is not this principled distinction between the term sociopath and the term psychopath. Is there a distinction between those two types of people? Would you say that there are some people who have this one set of traits quite strongly but not the other sets of traits, or are they really correlated with each other?
Marsh: Such good questions. That is a really interesting description that you were given about the difference between the two. That is even a new one. I have heard probably 20 different descriptions, including clinicians and researchers describing the two terms and what distinguishes them. You will get many different explanations.
First, it is totally possible for people to be persistently antisocial but not because they lack the capacity for empathy or remorse. I would say that for most people with psychopathy, it is not that they totally lack the capacity for empathy or remorse, but it is very suppressed. They feel these emotions very mildly or very rarely, maybe only with reference to one or two people.
Mounk: That is interesting because I think we do also tend to think about it like it is a switch. Most people have it, and then some people do not have it at all, whereas you are talking about it nearly as a normal distribution. Most people have it to a relatively typical degree, and then a bunch of people have it less than others, but they might still feel it for their mom or for their best friend or for the person they love, but not for anybody else. They might feel it more dimly than others. What do you think is the distribution? Is it something like a normal distribution, or how should we think about this?
Marsh: I think it is certainly a distribution. It has that classic bell-curve shape. It is probably shifted so that most people are on the plenty-of-empathy side of the distribution. The tail of people who experience very little empathy and compassion is a long tail, but a very small number of people are down there. I have worked with people who say they have never experienced anything like what most people describe as empathy or remorse or compassion. They exist, but that is a small percent of people, much smaller than the percent of people who say they wish that they could experience less empathy and compassion for others.
Mounk: Interesting, because that too can be debilitating if you inadvertently step on a bug and you are so empathetic for the natural world that you are terribly guilt-ridden about having inadvertently done that. That also seems like a kind of ill-adjustedness.
Marsh: It can be. Empathy, compassion, and guilt swim together. That is the downside of being a compassionate, empathic person: you often feel guilt about not having done even better, even though you are probably a nicer person than average.
Mounk: Let us get back to empathy and what people who have really strong empathy feel as part of a conversation as well. Sticking with the other end of the spectrum, obviously one reason why we are interested in this topic is that it drives bad behavior. A lot of criminals, people who are serial killers, and perhaps some politicians who do terrible things seem to suffer from some form of psychopathy. What drives bad behavior here? Is it primarily this question of whether you have empathy for others? Or is it whether you have empathy for others plus those other ingredients? What do you think is predictive with the patients you work with? When do you think, oh dear, I’m worried for the people in their life? Is it just how little empathy they have for others, or is it that plus some other thing about them?
Marsh: Each one of these prongs that makes up psychopathy—the bold, socially dominant personality; the low empathy and compassion; and the disinhibition—can drive antisocial behavior for different reasons. I think of them as different kinds of gas or brakes. You can be a very antisocial person who has at least the capacity for empathy and compassion simply because you are so disinhibited: you see something, you want it, you act. Or you may be very bad at controlling your anger. You may have something like intermittent explosive disorder, where it is easy for things to get under your skin and you blow up in a disproportionate way, and afterward you feel awful. Each time this happens you think, I cannot believe I did this again, but it is too hard to control those emotions.
There are also people driven by status goals. People who are very bold and socially dominant often like having higher status than others. Sometimes those goals can drive antisocial acts even though the person has the capacity for empathy but chooses not to deploy it. They may view certain people as unworthy of empathy or compartmentalize their empathy.
There are many people who are functional in their normal social lives. They have a normal family life and long-term friendships—relationships that typically do not emerge in people who are truly, pathologically psychopathic, because it is very hard to maintain friendships when you view others as instrumental means to achieve your goals.
Mounk: At some point they’re gonna figure that out, feel used and won’t want to be your friend.
Marsh: Right, because in people who are highly psychopathic, the “friendship” was only ever based on a desire to get something out of it. It is not a typical way of bonding with others. But there are people who can form close bonds while still being so driven by their need to dominate or to acquire that their empathic capacity gets switched off or suppressed by those goals.
Mounk: I’m trying to think about what this model looks like. One perhaps overly simple way of describing it is that we have desires in the world, and we act in order to serve those desires. Then we have inhibitions. One obvious inhibition is empathy for others: I’m not going to do this because it will upset someone. Is the relationship between desire and inhibition something like one minus the other? How should we think about that relationship?
You can imagine someone who is highly empathetic but has uncontrollable rages and therefore behaves badly, or someone who has very little—perhaps no—compassion for others but is so impassive, so lacking in strong desires, that they calculate it is in their rational self-interest never to break the law or cause drama because they prefer a quiet life of sitting at home watching Netflix.
Marsh: There are plenty of psychopathic people exactly like that who do not particularly care about other people. They just want what they want. The stuff that they want is pretty easy for them to get for whatever reason. Maybe because they have particular cognitive skills and talents that make having a pretty easy life not that big of a challenge for them. I will say the other thing that makes psychopathy such a challenge is that the bold, dominant personality is associated with being fearless, which means being unafraid of being punished. The other thing that keeps people who are psychopathic from behaving in pro-social ways is that they do not fear the punishment that comes from behaving antisocially, so it is very difficult to contain their behavior using threats of jail or other serious punishments. They just want what they want.
Mounk: What is it that makes them unresponsive to that fear of punishment? Is it that they do not care that if they end up in jail for 20 years, they are just as happy in jail as they would be somewhere else? Or is it that they do not have mechanisms of self-control, that they basically fail the marshmallow test? How should we think about what drives that behavior?
Marsh: Many of them would definitely fail the marshmallow test. That is true of people who end up in jail in general. Even in a typical jail or prison, only a minority of people would be considered clinically psychopathic. Close to half, but not most people. Disinhibition and failure to control impulses or difficulty balancing risks and rewards is a major factor that puts people in detention. That is true of people with psychopathy too.
Thanks for reading! The best way to make sure that you don’t miss any of these conversations is to subscribe to The Good Fight on your favorite podcast app.
If you are already a paying subscriber to Persuasion or Yascha Mounk’s Substack, this will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
Subscribe now
Set Up Podcast
And if you are having a problem setting up the full podcast feed on a third-party app, please email our podcast team at [email protected]
They do not like jail. It is not that they want to be there. There does seem to be neurodevelopmental differences in them such that in their very early life they do not fear the punishment ahead of time. They do not like it when they get it, but they do not fear it ahead of time, and they do not do things to avoid the threat or the punishment even if they will not like it when they get it.
Mounk: You just mentioned a stat by the by that is striking. You said close to half of people in prison, as far as we know, have those psychopathic traits. How does that relate to the general population?
Marsh: Yeah, because psychopathy is a continuum. It is hard to put a clear marker on exactly what proportion of the population has psychopathy because it depends on where you decide to draw the clinical line. Most people estimate it is about one or two percent of the population that has clinically significant levels of psychopathy. That is in the general population. Of course, it is much higher in populations where people are doing enough antisocial things that they have ended up punished for it.
Mounk: It is, to the tune of 25x or 50x, more likely for someone who is psychopathic to end up in jail. Do we know something about what percentage of people with psychopathy will be imprisoned at some point in their lives?
Marsh: We do not know this at all. We do know there is something called the Pareto Principle, the 80-20 rule. For example, 20% of people are responsible for 80% of absenteeism in the workplace. That is true for all antisocial behavior. A small minority of the population is responsible for most antisocial acts. When you look at violent crime, some people have estimated that maybe 1% of the population is responsible for two thirds of all serious violent crime.
We do not know the odds that any particular person with psychopathy will offend over their lifetime. It is definitely higher than the average person. We did recently a big population study of psychopathy, and rates of offending were twice as high to 10 times as high as the general population, depending on the specific antisocial behavior in question. This is a paper that is about to come out. We found that the most common kinds of antisocial behavior that very high psychopathic people engage in, in terms of what they are most likely to have done, is risky, reckless driving.
Driving under the influence, not obeying traffic laws, driving way too fast, et cetera. The vast majority of highly psychopathic people have done that. Somewhat smaller percentages have been convicted or punished for it, but they all do it. A much smaller percent have committed assault, maybe 40%. Smaller proportions have been arrested for it.
Mounk: If we think again of these three different classes of character traits that are necessary according to the DSM to classify someone as psychopathic, are they correlated with each other? If you have one of these three character traits, does that make it more likely that you have the other two as well? Or are we simply saying that 15% of the population have each of those character traits, and the people who end up having psychopathy are the ones who happen to have 0.15 x 0.15 x 0.15, you probably end up with a little bit lower, so you would say 20% of each, and they are not correlated with each other? Does having one of these sets of character traits also drive the other or is it genetically related in some kind of way or related in terms of your upbringing? Or are they independent of each other?
Marsh: The top question is whether these three clusters of traits are correlated with each other. Yes, to some degree. The meanness and callousness and the disinhibition are strongly correlated with each other, maybe 0.4. Not 0.7 or they would be the same thing. The bold, dominant personality less so, as low as 0.15 probably. Although when you do the developmental studies that follow kids over time, that does seem to be the core trait. That is the first trait that we see emerging early in childhood that puts the child at risk for all the other sequela of psychopathic traits.
I will also mention that this triad of traits is not in the DSM diagnosis that is closest to psychopathy, which is antisocial personality disorder. Why they changed the name when they put it in the DSM, who is to say. I am sure you are aware of what a book created by committees the DSM is, and it has some flaws. A lot of these traits are in there, but you will not find this exact triad of traits in the DSM, nor will you find the term psychopathy.
Mounk: I would venture that a term invented by, I believe, your doctoral advisor, Steven Pinker, and a former guest on this podcast has something to do with this, which is a “euphemism treadmill.” If you find a term for something that has negative associations, the term itself will start to be negatively charged over time. Therefore people will say there is something offensive about the term, so let us not go with it. Even if something like psychopathy or sociopathy was invented as a relatively neutral term, over time we associated it with the underlying behaviors. It comes to seem like a slur word. Perhaps someone in the committee said, “We cannot call people that. Let us find a new term.”
Marsh: Yes, Steven Pinker was on my dissertation committee. I should mention he was not my advisor, but I appreciated his fantastic advice then. I appreciate this term, “euphemism treadmill,” which is a huge problem for psychopathy because it is an incredibly common disorder, and it is antisocial behavior. I should reinforce that the psychopathic traits that we see in adults are almost without exception present in adolescents and childhood in people who develop psychopathy. It is a disorder that occurs throughout the lifespan. It is the most common reason that children are referred to psychiatric care in childhood, antisocial behavior problems, not all of which is related to psychopathy, but a big chunk of it is.
Subscribe now
Yet there is very little help for the affected kids and families out there. There is so little in the way of resources or clinical training devoted to helping clinicians help these kids. On top of that, the clinician and researcher community does the population of patients a double-diservice by coming up with 8,000 terms that all mean more or less the same thing. “Sociopathy” was in part a remedy to the stigma that the term “psychopath” had. “Antisocial personality disorder,” same thing. It is, “Let us try to get the stigma away from having a disorder that causes you to do antisocial things all the time.” The reality is you are never going to take the stigma away from the disorder. Of course, if you do antisocial things persistently, that is going to be stigmatized. There is no getting around it.
The way we get rid of the stigma is by finding treatments for it so that people do not continue to perceive this as some sort of a “bad seed” moral disorder that is a permanent stain and we should just lock you up forever. That is how to get rid of the stigma.
Mounk: You mentioned that one of the first things you can diagnose in children is this desire to dominate, and that is the predictor of the other character traits emerging. At what age in a child’s development do we start to see warning signs, and how predictive are they? If you go and visit some relative of yours or the child of some friend, and they have some concerning set of behaviors, at three years old, at seven years old, at 10 years old, at what point can you say, hey, I am not a trained psychologist, but I am starting to be concerned here, and I am not being crazy?
Marsh: It is hard because the age two to three is statistically the most violent period of the lifespan because it is typical for little kids to have terrible tantrums and be quite violent in some cases, biting and doing things that do not happen later in life. It is hard. The thing that you are looking for in kids who are two to three years old, which most parents of kids who develop psychopathic traits will say is when they started noticing something was different, is this fearless personality, which mostly manifests in the kid not seeming scared of things that other kids are scared of: Big dogs, the dark, being alone, very high places, doing things that are unusually risky, and that includes not being afraid of punishments.
Mounk: But some of those might be good traits. How do you know? Where is the line? You want your kid to be happy-go-lucky and relatively fearless. It is a big dog, but it is our friend’s dog. You don’t have to be worried about it. You’re going to be proud if your kid says, hello, big doggy. What makes you know where the line is between having a kid that loves the world and has not had anything bad happen to them, and so they do not have a lot of fear, versus thinking, my God, perhaps I’m raising a little psychopath?
Marsh: This trait by itself is probably not enough to worry you. I will say that being afraid of something and loving something are two different axes so that people who love something will override their fear for it, even if they have a normal amount of fear. Alex Honnold, the free solo climber, is a famous example of this. He is not a fearless guy at all. He talks about how terrified he was before giving his TED Talk. People assume he must be fearless because of the climbs he does. The reality is he loves to climb. It gives him so much joy, he is so driven to do it that that overwhelms the regular amount of fear. That is how I am about dogs. Big dogs, small dogs, barking dogs, whatever. When I was a little kid, I would throw myself at all of them. It was not fearlessness; it was love.
You are looking for fearlessness across the gamut, including when it comes to punishment. For many little kids, having mom or dad get mad at you when you have done something you should not do is enough to make that behavior less likely going forward. Some kids might need a more structured reinforcement system like the “one, two, three magic” system, which I highly recommend, which results in timeout. One, Two, Three Magic is a book about how to use principles of behavioral psychology to help kids manage their behavior in a way that does not involve arguing or yelling or harsh punishment. It is behaviorism. When a child does something disruptive that they should not do, you count them. “That is one.” The first time they do it. “That is two.” The second time they do it, and on three, they have a mandatory timeout. No talking, no arguing, no explaining. They have a timeout. Kids hate timeouts. It is boring. That works.
Mounk: This is true in discussions of criminal law as well. When punishment is certain and immediate, that is the most likely deterrent, and this is applying the same thing to kids. Do not shout at the kid, but make sure you punish the kid.
Marsh: Yes, it should be swift, it should be certain, it does not need to be harsh. It should be annoying. Punishment has gotten a bad name. Punishment is any response that reduces the likelihood of that behavior in the future. When you are a parent, there is no possible way to only use reward and reinforcement to shape your kid’s behavior. Kids will do things they should not do sometimes. Using simple, swift, certain, annoying consequences like a timeout, for most kids that is plenty and that will shape their behavior fine. For kids who have this bold, fearless personality it will not. One of the problems is that most parents have no idea what they are dealing with. They think, I do not understand, I have other kids and this worked fine for them, why is this kid not responding to these mild punishments? Unfortunately, what happens is the parents ratchet up the punishment. The timeout did not work, now I am going to lock you in your room. Locking you in your room did not work, now I am going to start screaming at you. Maybe I am going to use physical punishment. The problem is if the kid does not fear the punishment in advance, none of that is going to work.
Now you have a kid who is fairly fearless and thinks that everybody around them is a jerk. They are being treated unfairly. They are being treated differently than the other kids around them, including their siblings. The reason is their behavior is different, but a little kid is not going to put that together. This is the developmental cycle it leads to. The fearless temperament sometimes is coupled with low affiliative desires. We do not know if those two things are intrinsically linked or if having both is extra bad. Low affiliation means they do not find affection as rewarding as other people.
Parents sometimes try to be respectful of the child and dial back on the affection. Not as many hugs, giving them their space, not snuggling with them because they do not seem to like it as much. That is the opposite of what you should do. Kids do not always know what they need. Kids who are not that sensitive to affection need extra of it. You do not want to lock them into bear hugs while they struggle. You use big warm smiles, lots of affectionate touches because that is what it takes to get through to them.
Mounk: Just to go back to how you are facing this three-year-old and you are trying to figure out, is a three-year-old just an asshole like many three-year-olds are assholes, or is it a special asshole? One thing I have observed with kids, friends and family members, et cetera, is that kids at that age, terrible twos, will behave in horrible ways. They will bite their parents and whatever. They will afterward feel obvious remorse. They will look crestfallen, they hurt mommy or daddy, and they will be hurt by the punishment, and you will see that they are thinking, what did I do? Is the absence of that a reliable indicator at that age? Is that something that even at that young age you can observe, or is that not a key thing to look out for?
Marsh: I would say it is a slightly less reliable indicator because that is something you also might see in a child with autistic traits. You might not see as much of that attention to other people’s sadness or anxiety or stress. If you couple that fearless personality with the lack of seeming to care if you have hurt somebody, it is normal for a two- or three-year-old who one minute is a screaming terror, if mom starts crying or seems sad, to say, oh, mommy, are you okay? Not in the middle of a tantrum necessarily, but most of the time you should see a little of that. That means the child is coming to intrinsically value other people’s welfare.
Subscribe now
That is a normal human trait that should develop. When the people you love are doing well, that should bring you joy automatically. When they are in distress, that should bring you distress automatically. Children need to develop that. Kids with psychopathy do not have that. Other people’s welfare does not have intrinsic value to them.
Mounk: You are starting to speak about how some parents understandably step back from expressing affection for kids who do not seem to value affection in the same way, but that is the wrong thing to do. How much agency is there? If by the age of a two- or three-year-old there is something wrong and you do everything right, you take the kid to a doctor first and a specialist, and they are diagnosed with, this kid is in danger of developing psychopathy or they may have inclination toward that, what can you do and how much of a difference is that going to make? If kids are born without that sense of empathy for others, is that something you can give them through treatments and therapy and the right behavioralism, or is the best you can hope for to manage that condition so that the outcomes are not as terrible as they often turn out to be?
Marsh: It is treatable. You are not going to bring a kid’s personality from one end of the spectrum to the other. If you are a parent of a child who is three, four, five years old and you are noticing punishment does not seem to do anything to your kid’s behavior when you punish them consistently, swiftly, and not too harshly for their behavior, they do not seem to care, they do not seem affectionate, they are misbehaving more than their peers, they are lying more often than usual, they might be starting to steal things, what do you do? Most people start out taking their kid to a pediatrician. It is likely the pediatrician will have no clue what is going on. They have not been trained to assess behavior problems in children, but hopefully they will assess. This is something. I co-founded an organization called the Society for the Prevention of Disorders of Aggression to try to work on some of these problems, but the problems are big. Hopefully they will recommend a child psychologist or social worker to assess the child next.
They may or may not know how to assess these problems in kids. They may, even if they recognize that your child probably qualifies for a diagnosis of conduct disorder, which is the child equivalent of psychopathy, or at least some of it is, or oppositional defiant disorder. Those are the major disorders names that we give kids who look like they are on a trajectory toward persistent antisocial behavior. They may deliberately give you a different diagnosis because there is fear of stigma, there is fear that families will reject these diagnoses. There is a misperception among many clinicians that they are untreatable. They may want to give a diagnosis like autism because resources will come your way if you have that diagnosis. They may give ADHD because it is a light definition and it is not offensive. Maybe even depression. They may decide the reason that this kid seems cold and unemotional is because they are depressed. It is common that parents get a wrong diagnosis. Even if you do get a right diagnosis, the odds of finding somebody trained to use a therapy that works is not high.
There is therapy that works. PCIT, Parent-Child Interaction Therapy, is therapy where the parents are trained in techniques that will bring out the best in the child, that will over time eliminate their antisocial behavior, show the kid warmth, more warmth than you think you need to show them, to develop a bond between you and the child so they care about you and they care what you want and they care what you think.
Mounk: Without going into too much detail, what kind of behaviors are there? What can you do as a parent? You were saying one thing is to show even more warmth to your kid. Most parents give a lot of warmth to most of their kids, right?
Marsh: I am sure you are familiar with the pop psych book about love languages, which has a kernel of truth. Sometimes parents fail to realize that the way they are showing their children love is not as interpretable to the child as they might think. For example, taking your kid to swim lessons and piano lessons and doctor’s appointments and PCA stuff is love. That is very much love. It does not feel like it to the kid. Being taken to the doctor is not feeling love. Making sure their clothes are clean and that they have had breakfast and get out the door on time are things loving parents do, but to a kid it feels like being harassed and nagged.
One of the things that PCIT entails, and many good behavioral therapy programs involve, is something called floor time, where you set aside five to ten minutes every day with your child to engage in behaviors that show that you care about your child, that you like them, that you approve of them, and that you feel positively about them. They involve things like, while your child is playing on the floor, imitating what they are doing, playing alongside them, narrating what they are doing, and positively commenting on what they are doing.
It feels simple and dumb, but in reality that is a moment when you are not nagging your child, you are not hustling them from one thing to another, you are not telling them what to do, you are not asking them questions, which many parents interpret as a sign of care, which it is, but to a child being endlessly peppered with questions—what did you do in school today? What did you have for lunch? Who did you sit next to?—feels like an interrogation. Floor time is an isolated time when you are showing approval and positive regard. It is amazing how much just ten minutes of that a day will do. A seven-minute workout does not seem like it should be enough to improve health, but it does. Ten minutes of floor time can do wonders for a child.
That is the baseline. Then there is a separate set of behaviors that parents engage in that help their child learn to avoid bad behaviors and adopt good ones through reinforcement.
Mounk: Interesting. In the best case, you can take a kid that has a tendency toward psychopathy and avert the worst kinds of outcomes. Tell us about how the life of a kid like that is likely to go. Then tell us what happens if you do not intervene in that kind of way and how the problems of misbehaving when you are three years old turn into the problems of misbehaving when you are 13 or 23.
Marsh: I will say that parents of kids with these problems are blamed and shamed routinely for their kid having this callous, remorseless personality. It sometimes does not help to say the solution is parenting management training, which reinforces the idea that you must be a really bad parent. That is not true. In general, the kids who develop these traits who are raised by parents who get them into some kind of therapy, or sometimes just figure out a way to manage their behavior, end up as kids who are not going to be the warmest, most cuddly people in the world.
They usually have learned good habits and strategies to get what they want that do not involve overtly breaking rules or hurting other people. They may end up in professions that attract extroverted people that are high risk, high reward, and offer a lot of status. There is evidence that jobs in finance and banking might attract a higher proportion of people with psychopathy than teaching, for example, just to name a quite different job.
Mounk: This is one of the questions that my producer Leo tasked me with asking, which is that there is a stereotype in our culture that certain kinds of successful people may be more likely to have traits of psychopathy, whether that is banking or surgeons or CEOs. Is there something to that, that having psychopathic traits makes you less likely to be successful, but that among some subsection of highly successful people the share of psychopaths is higher? Politics?
Marsh: My sense is that there is not a ton of great data on this. There is at least one very good study that found that among executive-level people in the business world, the proportion of people estimated to have clinically significant psychopathy, that top one or two percent, is about four percent. So double the regular population, maybe four times the regular population, but that is also 96 percent of them who do not have clinical psychopathy.
Mounk: Among the prison population you said it is close to 50%, and among these people it is 4%.It is a little elevated, but not enough to look at your cousin who is a CEO and say, you’re probably a psychopath; that is why you’re a CEO.
Marsh: People who go into high-risk, high-reward professions—surgery, politics, business—are more likely to be men for various reasons. Men score a little higher in psychopathy than women on average. There are more men than women at the very high end and more women than men at the very low end, but the average psychopathy scores of men and women are not that different. Professions that attract more men will have a higher level of psychopathy because they have more men. Even if you correct for that, there is probably a small effect.
A lot of those professions attract people who are high in the boldness trait that is characteristic of psychopathy: boldness, social dominance. That is sometimes called the adaptive arm of psychopathy or the adaptive feature of psychopathy that may help explain why some level of these traits persists in the population over evolution.
Mounk: Right. Risk-taking, bold, fearless. Someone who is a rock star performing for 150,000 people is going to be more fearless than the average person, but that is not necessarily a bad thing. It would explain some of that statistical correlation.
Marsh: Most personality variation is what makes the world go round. It is what makes life colorful. Most personality traits have a wide range that can be adaptive in particular contexts. Some contexts make it more adaptive to be introverted and some extroverted. That is true for psychopathic traits until you get to the very high end. That is not adaptive in almost any context because you burn bridges with everybody everywhere by mistreating them and using them for instrumental purposes. Bernie Madoff is an example of someone who, based on what is available biography-wise, was probably a psychopath. He was using people to meet his instrumental goals.
Mounk: What is the evidence for that? From my understanding of it, at the beginning he was running a relatively normal investment business. Then he made promises that were bigger than they should have been. Then he thought, damn, in order to keep this going, I have to take in more money, and it turned into a Ponzi scheme.
I do not know the details of the case, so I am probably missing elements of it. That seems like something an average asshole might end up in. He did not set off saying, I’m going to perpetuate the biggest financial fraud in history. He thought, I’m going to invest these people’s money and make a lot of profit. Then he thought, shit, either I go bankrupt now or I fudge the numbers. Ten years later, he’s in way over his head. What element of what he did indicates that there was more to it than that?
Marsh: It is a good question. I will say it has been a while since I dug into his biography. What you are looking for is never one thing, because you are right. It is plausible that someone leveraging the money they have to invest and trying to build up a bigger portfolio could get in over their head, not because they are trying to use people, but because they were trying to do the right thing and ended up getting people bankrupt. That could happen.
Information you look for to distinguish between someone who did one bad thing, even if it is a big bad thing, and someone who is truly psychopathic is patterns of behavior throughout different parts of their life. Here I am not talking about Bernie Madoff in particular because I am not confident enough about my memory of his biography. If you are looking at someone’s life to assess psychopathy on the basis of some antisocial thing they have done, you first want to look at their relationships over time. Do they have evidence of real friendships, friendships based on mutual affection and reciprocally benefiting one another? Romantic relationships, relationships with their children. Do those seem to be characterized by sincere affection? When something bad happens to someone you love, are you truly upset about it?
You want to look over time biographically. For people who are psychopathic, there is almost always evidence in early childhood that their behavior was difficult to control. For example, children who were sent to special schools that were trying to manage their behavior might indicate their behavior was unmanageable to their parents. It is not the only reason kids get sent to special schools, but it is a characteristic thing you might see. Was there evidence that they engaged in juvenile delinquency? Were they underperforming in school?
Those sorts of biographical details are helpful to look at. It is rare that people who are doing fraudulent, dishonest things in the business realm are not also doing things like that in the personal realm, maybe in romantic relationships. You want to look at multiple prongs to get an overall portrait of someone who does not care about other people’s welfare and is using other people to get what they want.
Mounk: We have another set of questions from Leo. One part of this is whether it is true that psychopaths are often particularly charismatic, that they have an ability to manipulate people that goes beyond that of the average person. Relatedly, we have the holidays coming up. If you go to your Christmas gathering or whatever your favorite holiday may be, and your cousin’s significant other that they are bringing over for the first time seems to have behaviors that may be indicative of psychopathy, how do you think through whether you are talking yourself into a rage and need to touch grass, or whether you have to be concerned for the well-being of your relative?
Marsh: This is what is confounding about psychopathy: how hard it is to detect in a casual interaction. It is impossible. I do not think any psychopathy researcher says there is any tell in casual conversation that reliably indicates someone is psychopathic because people who are psychopathic are shapeshifters. This is not true of everybody, but most of the psychopathic people I have worked with will say that they mask in most interactions. They present the face to the people they are interacting with that they think those people want to see. That often results in their being some of the most charming, likable, friendly people you will ever interact with. They do not seem fake. Most of us have too much confidence in our ability to detect deception in regular life. We are not good at it. Everybody I know who works with psychopathy has found themselves duped at least once.
You are out of luck when it comes to detecting psychopathy in your cousin’s new boyfriend unless you know someone who knows him and can say, that guy is not trustworthy. He talks a big game, and half the stuff he says is not true, and it is not clear why he was making it up. He has had six different jobs in the last two years, and nobody knows why. He cheated on his last girlfriend. That is the kind of biography you are looking for.
Whether people who are psychopathic are usually charismatic is hard to say. There is not good data on this. They certainly report that they are. They will say, yes, I am. That is not satisfying. It is possible that if you want to manipulate people, being charismatic is a good tool to do it. Many people who are psychopathic may develop a more charismatic demeanor in an effort to do that. It may also be the case that it is people who are very psychopathic and charismatic who cause trouble. Those are the ones you do not detect.
Mounk: Right. It is an availability bias where if there are a hundred people with psychopathy, the ninety who are not particularly charismatic are less successful, less likely to become celebrities, less likely to become elected office holders. The subset of psychopaths we think about when we consider the condition are the ones who are charismatic. That is a very good point.
Marsh: That is possible. There is not as much good data as I would like on whether having a conversation with the average person who is high in psychopathy makes them come across as more charismatic. What they come across as is a genuinely likable person who does not seem particularly antisocial. I have done a little research on this, not published, that people tend to conflate how nice a person is, how good a person is, with whether they personally like them. For example, we are overwhelmingly likely to assume that political figures on the opposite side of the aisle from ourselves are psychopathic. That is why I always urge people: just because this person has politics you do not like, whether they are a national politician or a voter, that is not a sign they are psychopathic. You do not like them, and I get that, but they may think they are doing the right thing sincerely and honestly.
Mounk: That is true for ascriptions of psychopathy. It is broadly true for ascriptions of moral evil. It is tempting to think that my way of seeing the world politically is so obviously true that anyone who disagrees with me must be stupid or a bad person. That is an assumption I see people make all the time. One of the things you get from attending focus groups and listening to people of very different political points of view is that they seem genuinely to be struggling to think about what is right. In a different way, one of the things you get from reading the most brilliant political thinkers of the past, who lived in different moments and therefore had different political assumptions and different values, is that there are many well-intentioned, smart people who have views that are very different from those that I have. That is the core to beginning to have any productive form of politics.
Marsh: Exactly. People can do things that you dislike, have beliefs that you dislike, and be a sincerely well-intentioned person who is trying to be moral and is being moral according to what they view as moral. That is important when trying to explain what psychopathy is and is not. It is not everybody who disagrees with you politically.
I have had more than one conversation with academics who say, well, I will tell you where you can find all the people with psychopathy. They are on the Republican side of the aisle. That is not how it works. The most famous social psychology experiment of all time, the Milgram study, was designed to dispel the idea that Germany was a nation full of psychopaths. The belief at the time was that you could never have had such a thing happen in Germany as what happened in World War II unless German people were unusually callous and heartless and cruel. That was not true at all.
Mounk: Well, obviously there is the historical point that it is hard to understand why Germans in 1925 were not psychopaths, and in 1943 they were all psychopaths, and by 1970 they were back to having normal psychologies. It is not a convincing explanation. You spend a lot of time with psychopaths. You spend time studying them in research setups where there are layers of confidentiality so that you elicit honest responses. Tell us what it is like, as best you can tell from speaking with those people, to be a psychopath. How do people think about their own condition? How do people think about the world?
Marsh: One thing I should mention is that there are a couple members of the organization I created who have been diagnosed with psychopathy. They have written well-known memoirs in the last couple of years. One is Patric Gagne and the other is Amy Thomas. I have valued them as colleagues. Part of the interactions I have with people with psychopathy is as colleagues. I come at this from different angles. The first work I did in people with clinical psychopathy, or callous unemotional traits, which is the euphemism we use for children, was in my postdoctoral research when we were doing brain imaging research trying to understand the neural basis of psychopathy. We interviewed kids from all over the greater Washington, D.C. area and interviewed their parents to get a better sense of what made these kids tick.
One of the interesting discrepancies we found was that there was a wide range in how aware the kids were that there was anything different about them. This is true of adults with not only psychopathy but conditions like bipolar disorder or stroke. Some people recognize that there is something different about them, and some people think there is nothing different about them and that everybody else is just like them. That makes a big difference when it comes to treatment. If someone has no idea there is anything different about them, it is hard to convince them to get treated.
The kids who recognized there was something different about them had reasonably good insight about the fact that they found themselves doing antisocial things repeatedly that other kids did not do. Some of them said they wished they knew how to stop because they could tell that it was not working out for them in the long term.
Mounk: They did not say, I wish I knew how to stop because it makes mommy and daddy sad. They said things like, I realize I keep getting punished for it, or, nobody seems to like me. What is it that bothers them?
Marsh: They say, this is not the life I want. I have talked to adults who say the same thing. They say, this is not the life I want. I want to have friends. I want to have a regular job. I want to do well in my life.
For many of the kids I worked with, they had been kicked out of multiple schools. This was not the life they wanted. Many of them had an image of what a good life looks like, and that was not it for almost anybody. Some of them did not care. There was a subset of kids who did not care. They found school boring. They did not see the point in doing well. Getting kicked out of multiple schools would hopefully end with no school, and they would have been happy. People who are psychopathic and in their teens or twenties are the least likely to appreciate that they are the source of their own problems and want to change. As time goes on, it is more common for people to say, I’m the problem. It’s me, to paraphrase Taylor Swift, and to have some sense that they could learn how to do better. Some of them do learn how to do better.
Some of the people I have talked to, out of a desire to not lose a romantic partner they realize has been a lucky outcome for them, have said things like, I was a way worse person before I started dating this woman. I do not want to lose her. I tried to act like a nice person for a year and a half, and then it started to click, and I started to want to be a nice person. It is almost like they gave themselves therapy, exactly what a therapist would tell you to do. You need to start acting like a nice person. Do the things a nice person would do. Learn how rewarding and beneficial that is, and eventually it becomes a habit and feels natural.
Mounk: When they succeed at that, do they now genuinely love their partner and genuinely have remorse when they act in ways that upset their partner? Or is it that they have, in a way Aristotle would have described, developed a habit of acting in a certain way until it becomes an automatism, and they have gotten out of the habit of acting in ways that upset or hurt the partner? But deep down, when the partner is upset, it still does not create the wrenching feeling that you and I have when we have done something to make somebody sad.
Marsh: I have talked to people for whom both are true. Some are one and some are the other. I have talked to people who said that the only time in their life they have felt something that approached love was when something bad happening to this person made them feel emotional, and they felt bad about it in a way they had never felt about anybody else. I think there are some people with psychopathy for whom there is a lingering latent capacity to experience something akin to real love. It just takes a lot. That person has to be very nice and a kind, affectionate partner. Other people never approached what I would call love, but there was mutual understanding and reciprocity in the relationship. It was a stable relationship where both people were glad they were in it. This person was motivated to keep up their end of the bargain, something like loyalty. I have talked to people who are psychopathic who say that over time they learned to feel something like loyalty, even if not love in a more emotional way. Are these people the norm? It is hard to say, maybe not. They give some cause for optimism about what is possible.
Other people with psychopathy who are less insightful will say, I just want to get what I want. All of my relationships with people are aimed at getting what I want and figuring out what they want from me. I give them what they want to the degree that I have to in order to get what I need. I do not like my job. I do not like people. I take a lot of drugs to feel something, to feel some sense of enjoyment. I engage in risky behavior like reckless driving to feel something. This is a crappy life, but this is the way I am.
Mounk: How honest and open are people about the lack of remorse for bad actions or some of the things they have done? I understand this is a context in which they know you are not going to report what they say to the police. What are the things they would tell you in that context? What are some of the shocking things people have told you?
Marsh: I cannot tell the details. Pretty shocking things. I have been shaken by some of the things I have heard in conversations. I wish I could tell the details, but I cannot. The research that we do is protected by a certificate of confidentiality from the NIH that protects us from having to reveal what we are told under subpoena, because we would never be able to do this research if the people we work with thought that we were going to reveal what they told us. We are mandatory reporters if they tell us about something they are going to do, if they are planning on hurting somebody else or themselves, or if there is any indicator of child or elder abuse. We have to report that. We tell them that. We say, if you tell us anything about these kinds of items, we are mandatory reporters, just like many professionals are. If it is something they did in the past, we would never be able to do this research if we could not guarantee that.
Mounk: If they have confessed to terrible crimes in this context, what do they sound like? Do they tell that to you the way I would say, oh yeah, today for lunch I went to a Japanese restaurant? What is the tone?
Marsh: Casual is the best way to describe it. Yes, this is something that I have done, because there is always a reason. One of the things that people with psychopathy are good at is externalizing blame. It was not, I did it because I wanted to. It was, they put me in this position. The classic line of abusive partners in domestic settings is, they made me do it. They pushed me to it. Domestic violence is something that people with psychopathy are more prone to. It is never your fault if you are psychopathic. Everything is somebody else’s fault. You get explanations like, they put me in this position. They told me they were going to tell on me, so I had to do what I had to do.
I have worked with people who explain what they do through a moral framework they created, where a “good person” will only physically hurt people who do racist things. I only hurt people who say or do things that are racist. It is an interesting moral framework because it channels their antisocial impulses in ways that are, in their view, relatively prosocial.
Mounk: That is interesting. There is some research suggesting that people with dark triad personality traits are attracted to political extremes because those are a cover or an excuse for engaging in these kinds of behaviors. The explanation would be the other way around because those extremes give an excuse for saying, most Americans are terrible human beings. That allows someone to engage in wanton cruelty online, participate in cancellation campaigns, or other kinds of things. It is a way to live out this desire within a socially sanctioned sphere.
Marsh: I have not talked about politics much with most of the participants we brought through, but it fits with what we know. We know that people who are very antisocial online are more likely to have psychopathic personality traits. It is not that you become a different person when you go online. That would have been a reasonable hypothesis, but it seems that people’s online behavior is a reflection of their real personality, even if they put on a more prosocial front in in-person settings and act under the cloak of anonymity online.
Mounk: I know that you also are interested in altruism, and we have talked about it a little earlier in the conversation. It strikes me from what we have talked about so far that the problem of altruism is somewhat ill-posed in philosophy and in biology. The basic paradox of altruism is why we evolve to have altruism. You could expect people who serve their own interests to be the most successful at passing on genes to the next generation. Over time, altruism would be selected out in a population. There are various ways of trying to explain that, group-level selection versus individual-level selection, et cetera. It strikes me that what you are describing is that if you have somebody who is not altruistic at all, who has no sense of care for others, the life outcomes end up being poor. Perhaps some amount of altruism is necessary for success, which seems paradoxical but gives a more straightforward explanation for why altruism has evolved as a human trait.
Marsh: Absolutely. I think that is a good way of putting it. To me, one of the strongest arguments for the idea that humans on average have the capacity for altruism is the existence of psychopathy. Psychopathy fundamentally boils down to not caring about other people’s welfare for its own sake. That is the trait that is important.
Mounk: One way of being skeptical about the existence of altruism is to say, well, people seem like they care about others all the time, but because you can get things out of seeming like you care for others, perhaps deep down we are not altruistic. You can make the same argument about pets. Your dog seems like he loves you, but he has realized this is a smart way of getting food and does not care about you. What you are saying is that if that were what everybody was like, everybody would suffer from psychopathy. That is what this would be, but we clearly see in the population that psychopathy is rare. Interesting.
Marsh: I will say that a cynical, untrusting personality tends to go along with being relatively untrustworthy and callous yourself. When I talk about my research with altruism in general audience settings and people ask questions that reveal fundamental disbelief in the possibility that people could be altruistic, I usually give an answer along the lines of, this question does not say much about what people are like as a whole. There is a spectrum of altruistic capacities in people, but your question has told me a lot about you.
Every so often, if they push me, because some people are very cynical about whether it is possible for people to genuinely care about other people, many of those people do not have that experience themselves. They do not care about other people. Some of them may live lives that are ordinary enough or prosocial enough that it has not occurred to them that there is something unusual about them. They think they are like everybody else. “We are all like this to some degree.”
Mounk: If anybody in the audience now thinks, oh my God, I am skeptical about altruism. Am I a psychopath?, how can people self-test whether they are or are not?
Marsh: Yes. You can go to my organization’s website, disordersofaggression.org, and there is a link you can click on for some screeners for psychopathy, totally anonymous.
Mounk: What kind of screeners or questions do you ask?
Marsh: The psychopathy screener we have for adults is called the TriPM, the Triarchic Psychopathy Measure, and it assesses those three bins of traits we asked about before. There are questions along the lines of, I have the ability not to let my emotions dictate my behavior, I like to do risky and dangerous things for fun, I am more competent and capable than the average person. It is questions that get at all the traits we have been talking about.
Mounk: I realize the pull of psychopathy is such that we have gone away from altruism again. One of the questions to ask about both of these traits is whether it is nature or nurture. Is there something in the constellation of the particular genes that a child has that makes them more likely to display these psychopathic traits, or on the other hand to be an ultra altruist, somebody who is particularly altruistic? Or is it something about the early childhood environment that triggers this in some way?
Marsh: I would say the best estimates are that about half the variation in these personality traits is genetic, which is true for most personality traits. The estimates always end up about there. About half the variation is genetic factors and about half is environmental factors, although not in a simple way, as you know, like shared environments. The socioeconomic status of your family and what kind of school you went to and whether your mom worked are none of the things that are terribly predictive of life outcomes in general, mostly because different kids respond differently to the same environmental variable. They do not affect every kid the same way, and you cannot make simple predictions. There is a strong genetic component to psychopathy, which I think is hopefully a relief to parents of kids who have these traits, that it is not like you caused the kid to have these traits.
Because there is a big non-genetic component, there is the capacity for environmental variables to help treat them. That is the good and the bad news.
Mounk: Sometimes the idea of genetic traits seems harsh or mean, but in a sense it is not. If we said this is all about upbringing, then if somebody has psychopathy it is probably that the mom was not loving enough or whatever sexist tropes you can immediately bring out. If there is a strong genetic component, and it is a polygenic trait, it is not one gene where you have one thing that is passed down and because of this gene you are a psychopath. It is a combination of many genes, and some of those combinations are more likely to result in these traits. You can have one sibling that does not have any inclination toward that at all and another sibling that does by random combination of genes. In a sense that is a less harsh view of the world, or at least a less harsh view of the responsibility of parents.
Marsh: I completely agree. Parents give themselves more credit and blame than they deserve. Children are agents of their upbringing in a way that is not clear in parenting books. Children elicit parenting from parents. They are not passive receptacles of things done to them. Children are also fairly robust. This is not an invitation for people to treat their children harshly. Your child will be attracted to certain kinds of outcomes, and they will want certain kinds of behavior from parents in ways that shape you in a complicated dance. Your job as a parent is not to shape your child to turn out some particular way.
There is the analogy of the gardener and the carpenter by the psychologist Alison Gopnik that I like. Parents think of themselves as carpenters too often, that they are trying to build a chair, and if they do this and they do that, then the chair will have a nice sturdy frame and strong back. That is not it. Throw away the wood. You are a gardener. Your child is a seed that could be a geranium seed or a marigold seed or a zinnia seed, and you have no idea what they are going to be. Your job is to try to help foster and let that seed grow as best you can. They are going to turn out the way they are going to turn out to some degree because genetics are powerful.
Mounk: That is a lovely analogy. I have a dear friend who is an avid listener to this podcast who studied sociology and used to ideologically believe that most outcomes are due to culture and social factors. Now, as she has children, she told me recently, I was wrong. Kids just come out one way or the other. Bringing them up in the right way, nurturing them, is important, but they are different from each other, and it is not because you treat one one way or the other another way.
In the rest of this conversation, Yascha and Abigail discuss what to do if someone in your life may be psychopathic. This part of the conversation is reserved for paying subscribers…
Polarization is at an all-time high. It can feel daunting—perhaps even misguided—to engage in meaningful dialogue with those holding starkly different views. What does it mean to champion pluralism in such a moment? Persuasion’s new series on the future of pluralism, generously supported by the Arthur Vining Davis Foundations, features essays and podcast interviews that make the case for civic dialogue and highlight inspiring examples of it in practice. You can find past installments here.
Political polarization tempts all of us to collectivize our enemies, blinding us to distinctions among those we consider opponents. Conflict over the university has not been immune to this tendency. Conservative proponents of viewpoint diversity sometimes describe humanities professors as an undifferentiated mass of illiberal bullies. Some on the left return the favor. Johns Hopkins professor Lisa Siraganian’s widely-read essays in Academe and the Chronicle of Higher Education, for example, describe the campaign for viewpoint diversity as “a MAGA plot.” By her lights, viewpoint diversity is the duplicitous rhetorical strategy of a conspiracy of politicians, plutocrats, and far-right professors determined to burn the university down.
To be sure, some on the right do want to burn the university down. But critics like Siraganian fail to distinguish between scorched earth right-wing culture warriors and the broad, politically ecumenical group of scholars—including some of her own colleagues—who want to bring more conservative perspectives into the curriculum in order to build the university up. This group isn’t carrying water for the Trump administration or trying to undermine academic freedom. They recognize, instead, that the quality of university teaching and research has suffered because of the neglect of conservative ideas, and are working to reintroduce such ideas to campus while respecting the best traditions of the university.
We consider ourselves members of this group, and have started an initiative to bring more conservative thought into college classrooms without imposing anything on anyone. Last spring, the two of us—a professor at a liberal arts college, Claremont McKenna, and a fellow at a center-right think tank, the American Enterprise Institute (AEI)—partnered to invite a politically mixed band of professors and think-tank scholars who teach courses on conservatism to help interested academics develop their own courses on conservative thought. The workshop attracted an ideologically heterogenous group of faculty, none of whom were enemies of academic freedom or looking to push a narrow political agenda. In fact, perhaps the only thing we collectively agreed upon is that the conservative intellectual tradition is worth teaching.
How did it go? If we had been looking to advance a MAGA plot, we would have had to count it a failure. While it’s true that conservatism—with its characteristic appreciation for the value of our civilizational inheritance, the fragility of moral and political order, the limits of human understanding, and the perils of revolution—furnishes students with critical tools to assess various left-wing enthusiasms, those same ideas provide students with critical distance on the MAGA right.
Giving students an occasion to discover the divergence between the depth of the conservative intellectual tradition and the shallowness of the contemporary online right is one of the major attractions of teaching conservatism for one of our session leaders, Emory’s Frank Lechner. Lechner said that one typical student “had expected a class dealing with the likes of Tucker Carlson and Candace Owens” but instead “got something much more interesting and serious.”
One reason is that conservatives are often, paradoxically, devoted to the liberal order. Partly because liberalism is our most important tradition, American conservatives are often consumed with preserving what is best in it, not burning it down like the online right. That also means many conservative intellectuals are hard to put into a tidy philosophical box. UNC’s Rita Koganzon, another of our seminar leaders, reported that teaching conservative intellectuals reveals “that no really good writer is merely conservative, just as none is simply a liberal. They all have strange, heterodox, often politically uncategorizable ideas.”
Conservative thinkers sometimes even embrace fashionable liberal causes for conservative reasons. For example, the agrarian conservatism of novelist Wendell Berry, an influential figure among young traditionalists, has driven him to engage in both anti-war and environmental activism. Conservatives can even embrace revolutionary change when they conclude that our institutions are so broken that there is nothing left to conserve. Hence, our workshop also exposed faculty to the work of Patrick Deneen, author, most recently, of the book Regime Change.
Subscribe now
Indeed, in studying the breadth and intellectual history of conservatism, students invariably notice that it is not a narrow orthodoxy but a tradition enriched by deep internal tensions. While conservatives may share a common interest in preserving what is best in their inheritance, they often disagree over how to do so or even which inheritance should be conserved. A session with Stanford’s Jennifer Burns detailed one such tension in that history: The long-running intellectual and personal feud between the libertarian atheist Ayn Rand and the highbrow Catholic William F. Buckley. AEI’s Matthew Continetti, meanwhile, highlighted another: the hard fought tug-of-war between neoconservative intellectuals and the GOP’s more populist base.
Getting beyond a cookie-cutter vision of conservatism helps students see that what is most interesting in this tradition are not its reflections on the narrow question of how we should vote but its insights on the broad question of how we should live. Another of our speakers, Eitan Hersh, the founder of Tuft’s Center for Expanding Viewpoints in Higher Education, discovered how much studying conservative intellectuals meant to young people trying to sort their own lives out. This should be no surprise, since the conservative intellectual tradition addresses questions especially relevant to young people coming of age: Should I get married? Watch pornography? Have casual sex? Have children? Believe in God?
While conservative intellectuals have no monopoly on such questions, they have given answers that resonate with the intuitions of millions of Americans. Without engagement with this important strand of our intellectual inheritance, what Yuval Levin calls “civic self-knowledge” is impossible. No role could be more suitable to the university than to help the young people of our troubled country acquire such self-awareness. Even students who reject conservative arguments will be better citizens for understanding why they do and why others do not.
In other words, reintroducing intellectual conservatism to our universities will temper, not accelerate, our polarization. Today’s partisan spirit is a reductionist one: tell me if so-and-so is left or right, a squish or a solid, a friend or an enemy, and that’s all I need to know. The spirit of the university should be the opposite: it should lead one to follow the arguments implied by diverse political stances to the strange, profound, hard-to-answer questions with which every human being must wrestle in determining how to live.
Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal democratic values wherever they are threatened, please consider becoming a paying subscriber today!
In this way, a greater presence of conservative thought in the classroom can help universities do their essential educational work better. It might also help them recover some of the good will they have lost. Public trust has been declining in part because Americans believe the university has become too sectarian. And while this view is pervasive among Republicans, large numbers of Democrats share it as well. According to a recent Gallup survey, nearly a third of Democrats say that universities do a “fair or poor job” of “exposing students to a wide range of opinions and viewpoints.”
These concerns are not unreasonable. Very few university courses focus on American conservative intellectuals. As the political theorist Mark Lilla observed back in 2009: “A look at the online catalogs of our major universities [reveals] plenty of courses on identity politics and postcolonialism, nary a one on conservative political thought.” Research also shows that professors generally fail to teach a wide range of scholarly perspectives on the issues that most divide us. Even at the best universities, too many students graduate without being assigned a single article or book by any author who might even plausibly belong to this broad tradition.
The result is that even the most well-read academics are often innocent of viewpoints that contradict their milieu. Fairly late in life, for example, the NYU psychologist Jonathan Haidt stumbled upon conservatism by accident as he was browsing in a used bookstore. He started reading Jerry Muller’s excellent introductory essay in his anthology on conservative thought. By the time he hit the third page, he had to sit down right there in the bookstore to keep reading—the experience, as Haidt puts it, “literally floored” him. “I began to see [that conservatives] had attained a crucial insight into the sociology of morality that I had never encountered before.”
Unless students are assigned more authors like Muller, public trust in the university is likely to continue its descent for two reasons. First, a growing diploma divide is realigning our partisan loyalties, a trend that may well leave the next generation of Republicans even more estranged from higher education. Second, because younger cohorts of academics are more progressive on average than their aging peers, the university is poised to become increasingly dominated by leftists. As the political scientist Steven Teles noted, “a further drift leftward among the professoriate is already baked in.”
Without reform, the present partisan rift over higher education is on track to widen into a chasm. Such a future will likely include further rollbacks of tenure protections and more ham-handed legislative attempts to dictate what faculty teach. And more Americans will fail to benefit from higher education because they get the message that college isn’t welcoming for people like them.
To avoid this outcome, universities should work self-consciously to widen their intellectual horizons and build back the broad-based public support they enjoyed not so long ago. Their best response to the charge that they have become sectarian centers of leftist indoctrination is not guild-like defensiveness, but curricular reforms that make that stereotype less true. Such changes are the most authentic and enriching way for universities to respond to their present political travails: by using the resources of a serious but marginalized tradition of thought to deepen their internal intellectual life.
Jon A. Shields is a professor of government at Claremont McKenna College.
Benjamin Storey is a senior fellow in Social, Cultural, and Constitutional Studies at the American Enterprise Institute (AEI).
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
Bookstack at Persuasion is delivered to you every other Sunday at 6am EST. Grab a coffee, settle onto the couch, and see your weekend out with a deep dive into the best new book releases…
This week, Richard talks to Shadi Hamid, author of The Case for American Power (Simon & Schuster).
Bookstack is a production of American Purpose at Persuasion.
Host: Richard Aldous
Producer: Bhaasita Athani
Follow Persuasion on X, LinkedIn, Instagram, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below. Bookstack listeners can take advantage of a 20% offer for paid Persuasion membership, with access to special columns and bonus Good Fight with Yascha Mounk podcast content.
20% off for Bookstack listeners!
Sven Beckert is Laird Bell Professor of History at Harvard University. His most recent book is Capitalism: A Global History.
In this week’s conversation, Yascha Mounk and Sven Beckert explore the origins of capitalism, how this triggered the Industrial Revolution, and whether today we’re in late stage capitalism.
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: The scale of your ambition in this book is monumental. It really is to tell a one-volume history of global capitalism. What do you think are the misconceptions that smart readers, smart thinkers, and smart listeners to this podcast are likely to have about capitalism before they read your book?
Sven Beckert: That is an excellent question, Yascha. I think there may be three misconceptions about capitalism that are not universal, but broadly shared. One is that you can understand capitalism purely on an entirely abstract level, that somehow there is a way to understand the mechanisms of how capitalist economies work and to identify the essence of these mechanisms and leave it at that.
While it is important to think about what capitalism is on the most general level, the book is fundamentally a history of capitalism. It is a history of really existing capitalism, a history of capitalism in action. What you see very rapidly when you start researching that history is that, while the general characteristics of capitalism are universal to capitalism in all places and at all times, capitalism changes drastically over the past 500 or 1,000 years.
You can only understand capitalism from a historical perspective. This is the first common misunderstanding about capitalism: that it can be understood purely from an abstract perspective, when it really needs to be understood historically.
That is why we do history, because we believe that we can understand the world in ways that are fundamentally different and perhaps even better than understanding it from a purely timeless and abstract perspective.
The second misconception about capitalism that is very pronounced is that capitalism can be understood by looking only at European history, and perhaps including the history of the United States as well. The rest of the world is often seen as not really mattering to the history of capitalism. If it matters at all, it is because the rest of the world can see its own future by looking at the European or North American experience.
My book fundamentally disagrees with that. It argues that capitalism can only be understood from a global perspective.
The third misconception is that many people who think about capitalism are drawn to industry—to the big machines, the textile industry, steelmaking, car manufacturing, and whatever came later. They are also very much drawn to understanding capitalism from the perspective of cities, because cities, of course, are also a product of the capitalist revolution as it unfolded in the past couple of centuries.
Mounk: You write a book about the history of the American railroads or you write a book about the history of Manchester.
Beckert: Exactly—Manchester, or look at the Pittsburgh Steelers, or you write a history of the Ruhr, or of Alsace and the cotton industry there. Of course, this is very important to the history of capitalism. Do not misunderstand me. Much of the history of capitalism actually unfolded in agriculture, and it unfolded in the countryside, where most of the world’s people lived until very recently.
This is the third common misunderstanding of the history of capitalism: to see it as entirely an urban and industrial story, and in that way to exclude much of the human population and much of planet Earth from the narrative of the history of capitalism.
Mounk: I want to come back to some of those assumptions over the course of our conversation. It is going to be hard to summarize a thousand-page book in one podcast conversation, but tell us a little bit about what the accurate history of capitalism is. Some of the textbooks about capitalism or economics are probably going to start with the invention of the spinning jenny, the invention of the steam engine, or perhaps the first railroads in England in the 18th or 19th century.
You start in a number of different places and much earlier on, and you say that there are “islands of proto-capitalism” that arise long before the textbooks usually start the story. What is significant about those islands of capitalism? What makes them meaningful precursors of the system we are talking about and living in today?
Beckert: The starting point of the book, and I think of any kind of productive thinking about capitalism, is that capitalism is not the natural order of things. Organizing economic life along capitalist logic is a revolutionary departure from much of human history. We need to start by seeing the world we live in not as natural or as the kind of economic logic that animated economic life throughout human history, but as a very particular moment in human history.
Once we accept that, we need to think about the questions of where it starts, when it starts, and how it starts. There are many explanations for that question. Some people have argued that it is something about Protestantism that made people more capitalist. Others have argued that it is a peculiar kind of class relation in the southern English countryside that created the conditions for the transition to capitalism.
All of these approaches tend toward seeking a small kernel, a small essential element from which capitalism eventually sprung. Some people see it in the Industrial Revolution. Once modern machine production came about, they argue, this is when capitalism began in the late 18th or early 19th century. I think this is completely mistaken.
The Industrial Revolution, the machine production itself, is the most meaningful offspring of capitalism, but not the beginning of capitalism. I argue, and I show in the book, that capitalism really begins with the first people who used a capitalist logic in organizing their economic lives. These people are merchants—people who did not become rich because they plundered others or forced them to work for them, but also not people who engaged only in subsistence agriculture and lived off the land.
These are people who had capital, invested capital in trade, and hoped to generate more capital that they could then invest in further business enterprises. I see merchants as the crucial originators of the capitalist revolution. These merchants have been around for a very long time. This is hardly something novel. But they have always been marginal to economic life on Earth. There have been merchant communities for a very long time, but they have been very marginal to economic life. I begin the story of capitalism with these merchant communities, which can be found in many different parts of the world.
Mounk: Some of these merchant communities are in places where you might expect them to be in a more Western-centric reading—Venice and Genoa and those kinds of commercial republics that were very prosperous in the late Middle Ages in Italy and Europe. Others are much farther afield, in India, China, and elsewhere.
Beckert: Everybody knows about the Medici. Everybody knows about the Fuggers. Of course, they mattered. It is not that they did not matter. But if we look around the world, we find merchant communities in many different parts of the world. We find them in West Africa. We find them in China. We find them in India.
The book actually starts in a most unlikely location: the port of Aden in modern-day Yemen. I trace the history of this merchant community in this town in the 10th, 11th, and 12th centuries. You see these merchants already being quite modern actors. They are recognizable to us. We understand what they do even now, 900 years later. But when they were alive and practiced their trade, they were total outsiders. They were very marginal to how much of humanity lived at that particular moment in time.
Mounk: Those merchants existed well before that, right? What is it that makes those merchants proto-capitalist in a way that a trader a thousand years earlier in China or in the Roman Republic or elsewhere would not have been?
Beckert: You could legitimately write about all of them. I just had to start at some point in time. These people who employed a capitalist logic in their economic lives have existed for a very long time. I start around the year 1000 because at this point there were many more of them. I also write about them because I need to show, and I do show, how out of these communities modern capitalism emerges.
There were capitalists in the 11th century, and there were capitalists certainly in the 6th and 9th centuries, but there was no capitalism. We have capitalists without capitalism, but it is these capitalists who eventually build something that is more recognizable as capitalism.
Mounk: How does that happen? For now, we have these islands of people operating by capitalist logic, able to accumulate very significant wealth, but they are disconnected from each other. There is no real connection, for the most part, between those traders in Venice and the ones in China and so on. Is the next phase that they start to be in communication with each other and begin building connections among themselves?
Beckert: This is the very logic of their existence. They always connect places to one another. In that way, they are very different from all other economic actors, such as a feudal lord in the French countryside, who does not necessarily connect to people economically elsewhere in the world. These merchants always connect to one another, and they engage in astounding amounts of long-distance trade. They connect the Arabian Peninsula with the western coast of India. They connect the city of Cairo with what is now southern France. The West African merchants connect the forest belt of West Africa all the way into Europe. They build astoundingly extensive economic connections across large spaces.
In the 15th and 16th centuries, something happened that had never happened before. This occurred within a small subset of these merchant communities—the merchant communities of Europe—who before then were neither particularly rich nor particularly well-connected, nor particularly capable of building institutions in which to embed their economic lives.
Thanks for reading! The best way to make sure that you don’t miss any of these conversations is to subscribe to The Good Fight on your favorite podcast app.
If you are already a paying subscriber to Persuasion or Yascha Mounk’s Substack, this will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
Subscribe now
Set Up Podcast
And if you are having a problem setting up the full podcast feed on a third-party app, please email our podcast team at [email protected]
In the 15th century, they sought a way around the very powerful merchant communities of the Arab world and the Middle East in order to directly access trade with India, which was the center of the global economy, along with China. At the same time, European states, which were very splintered and fought many wars with one another, sought new resources to fund these wars and to buy the weapons needed to battle one another.
A kind of coalition emerged between these European states and this particular group of merchants—namely, European merchants. This coalition propelled them first into the Atlantic world, off the coast of Africa, to places like the Canary Islands, Cabo Verde, and other such places. Eventually, by 1492, it propelled them across the Atlantic all the way to the Americas.
Mounk: When you go back to Amerigo Vespucci and others, they are, in a way, an expression of this coalition between merchant families with deep experience in navigating the seas, building ships, and all those things. They were financed by the monarchies—in this case, by the Spanish monarchy—in order to undertake those voyages.
Beckert: There is a kind of merging of the interests of these expansive merchant communities and the interests of the expanding European states. What these merchants then do, in coalition with the states, is build all kinds of new islands of capital. Cabo Verde, for example, becomes a kind of plantation economy. Potosí becomes the source of much of the silver that eventually propels the global economy. The Caribbean islands, first the island of Barbados, become a kind of plantation economy dedicated to the production of sugar.
This is something world-historically new because these places are now further islands of capital, similar to those that had existed for several centuries before, but these new islands of capital are completely dominated by the logic of the merchants. They become the first truly capitalist societies because everything in these societies is oriented around the logic of investing capital, accumulating capital, and making more capital. This is the moment in which the capitalist revolution is taking off.
As you notice, this is taking off in a globally connected world. It is not just about Europe. It is not just about Florence. It is about connecting all of these places to one another, creating the dynamic of the capitalist revolution at this particular moment in time. What is truly novel is that much of the capital the merchants had invested in the past was invested in long-distance trade. They bought cheaply and sold dearly. They bought spices in India and sold them for large sums in Cairo or Amsterdam or elsewhere.
Now their capital goes into organizing production to produce more of the things they want to trade. For the first time ever, they come to dominate production. They are the ones who organize the production of the commodities they sell on global markets. That is totally novel, and this sets into motion the logic of the capitalist revolution that, at first very slowly but by the 19th century very rapidly, spreads all around the world.
Mounk: We are on the cusp of a mechanical revolution, of what we would usually call the Industrial Revolution. Before we get there, there is not a bright line dividing the period before the Industrial Revolution and after it. There is technological progress throughout the 16th, 17th, and 18th centuries.
When we place ourselves just at the cusp of the invention of the spinning jenny, the steam engine, and all those innovations, to what extent is the system you are describing capitalist? In what ways is this meaningfully a capitalist system? What elements that we would think of today as the obvious ingredients of a capitalist system—and I take your point that there is not one capitalist system, as it has changed and transformed over time—were not yet present at this historical juncture?
Beckert: That is an excellent question. As I said earlier, capitalism does not emerge at one historical moment fully formed. It slowly expands. Capitalism is not so much a thing as it is a process, and it takes centuries to create the world in which we live today. In a way, capitalism even today is still in the process of its own emergence. There is still the spread of capitalist logic into new spheres of human life.
Before the Industrial Revolution, much of economic life on planet Earth still followed a totally different logic. Capitalism, in some ways, is expansive. It is very important to world history, but it is still quite marginal to the economic life of most people living on planet Earth.
Mounk: How is the economic life of most people determined? It is feudalism. For people who are not historians, who are used to living in a world made by capitalism, what does it concretely mean to live in a world that is not shaped by capitalism?
Beckert: This is the problem. We live in capitalism like fish in water, so it is very difficult to see it. The capitalist logic is a radical departure from previous human history. The vast majority of people on planet Earth, well into the 19th century, lived in subsistence-oriented economies in which they produced for themselves and maybe for local exchange. For example, I grow a lot of wheat, you grow a lot of corn, and I exchange some of my wheat for your corn.
Mounk: There’s still people living like that in the United States, but I think it’s spelled W-E-E-D, not W-H-E-A-T.
Beckert: That is perhaps true. The capitalist revolution is ongoing. This is how the vast majority of people on planet Earth lived. Many people also lived as dependents of rulers or religious authorities. These rulers or religious authorities forced the peasants to give up some of the things they produced. They might even have accumulated a lot of wealth. If you travel through Europe today and see all these castles and beautiful churches and other buildings, this is partly because these institutions were able to dispossess people. They took some of the product that others produced and consumed it.
Subscribe now
This is completely different from modern capitalism. In modern capitalism, people who have capital invest that capital productively, not just to show off their wealth, but to enable further accumulation of capital. If you look at the world before 1900, you see that this capitalist logic has moved into areas of economic life that were previously not dominated by it. For example, a vast plantation sector had emerged in the Americas, in which huge quantities of sugar, coffee, rice, indigo, and later cotton were produced for European consumers.
This production followed a capitalist economic logic. In the European countryside, or to some degree in the North American countryside as well, people manufactured goods under the auspices of merchants. The merchants controlled these producers, acquired these goods through market exchanges, and sold them into long-distance markets. This occurred in many different parts of the world.
The stunning thing about this world is that there was not much technical innovation. Often we think of capitalism as being driven by technical innovation, but we do not see much of that here. There were no great gains in productivity at this time. What we do see is a huge redistribution of the wealth of the world from certain areas into others. For example, the beneficiaries of the radical recasting of production in the Caribbean were not the Africans forced to work on these plantations, nor the Indigenous people of Latin America, but rather the European merchants who accumulated great wealth by dominating this form of production.
This is the world we have before the Industrial Revolution. There are not great productivity gains. Economic growth is very slow, but the capitalist logic has now infused some parts of economic life on Earth. It is still somewhat on the margins, but less so than in the year 1200, when the merchants of the Arab world were dominant.
Mounk: That is an interesting implication of the way you tell the story. If you think capitalism is about the Industrial Revolution in Manchester and all of those things, then there are big debates about how quickly and to what extent it improved the living standards of average workers in Britain. I studied history as an undergraduate in Cambridge, and the first paper I read in my first term there was 18th- and 19th-century British social and economic history.
The first two weeks of my university education were all about relatively detailed, to me at 18, arcane articles about whether the standard of living of the average English worker had gone up from 1800 to 1850, and what evidence we have about that from increasing height or records about how much they were able to eat and all of those things. Broadly speaking, if you start the story in 1750 or 1800, you then have this incredibly rapid improvement in the standard of living. It is clear that living standards have vastly improved and transformed—initially in Britain and other parts of Europe and North America, and now across the world.
You associate capitalism with those improvements. Part of the implication of your story is to be more skeptical about that, to say that you can perhaps decouple those two things to some extent—that you start having this capitalist or proto-capitalist revolution well before you get those jumps in productivity.
How connected are those? Is it the case that they are really unconnected, or is it the case that what you see with those forms of redistribution in the 16th, 17th, and 18th centuries is the accumulation of capital that may be extractive and involve a lot of violence and injustices, but that actually allows people in the north of England to invest in new technologies and factories and all of those things? Is that what leads to the mechanical progress that results in rising living standards? Take us through the next stage of this history and tell us how connected these two things are, or are not, in your mind.
Beckert: They are deeply connected. You are right, and this is important to emphasize, that in the long term, looking at this problem from the perspective of the last 500 or 1,000 years, capitalism has led to the most astounding increase in human productivity and the most astounding economic growth in world history. There is no other form of economic life that has even remotely produced anything like this.
It is impossible to explain this rapid increase in human productivity and economic output by starting only with the moment in which it happens. You cannot understand the Industrial Revolution—which occurred in the late 18th and early 19th centuries—without understanding the earlier history of capitalism that made it possible. As I said before, capitalism was not born with the Industrial Revolution. The Industrial Revolution is the most important offspring of the capitalist revolution. It is a result of the capitalist revolution, and therefore we need to think about it in longer terms.
The Industrial Revolution begins principally with the recasting of the production of cotton textiles in British factories. Everything about the possibility of the emergence of this new industry is linked to the global connections that had been created in the decades or centuries before. It is connected to this earlier form of capitalism.
For example, why did the British want cotton textiles? Because for centuries they had traded with India and had acquired very high-quality cotton textiles from South Asia, which they also used to trade for enslaved workers on the African continent. There was already a market for cotton textiles before the British ever started producing them themselves. There was an incentive to replace imports with local production.
A second factor was the essentially unlimited supply of raw cotton that could feed these machines. Why was there such an unlimited supply of raw cotton? Because it was possible to force enslaved Africans to grow this cotton on another continent. British agriculture could remain as it was because the cotton was not sourced locally—it was sourced elsewhere.
Furthermore, because Britain and other European powers had already expanded their influence into many different parts of the world, they had the ability to dominate trade in the cotton textiles that came out of British factories. By the 1850s, India was already the single most important market for British cotton textiles.
The Industrial Revolution is a fundamental event in human history and in the history of capitalism, but it cannot be explained in isolation. It must be embedded within a broader global story—one that reaches much further back in time.
Mounk: That is a nice way of putting it, that some of the ways in which people place the Industrial Revolution at the heart of the story posit it as both cause and effect. They are trying to explain a change out of a change, in a way that is a little puzzling. There is always the question of what the prehistory is. Some of those answers involve capital accumulation; some involve chance invention. Some involve a very slow increase in the background conditions that make such change possible until a threshold is suddenly hit. But it can be mysterious where this threshold comes from.
Your story offers a cause of the effect that is sufficiently different, but it seems somewhat less puzzling. What impact does the Industrial Revolution then have? If you are talking about capitalism as a process rather than as a thing that is unchanging, then the rapid change in the mode of production that comes with mechanization and those inventions propels that process forward in a very significant way. Tell us what that looks like in the 19th and early 20th centuries.
Beckert: The Industrial Revolution at first unfolds in a very small part of the British Isles, principally in Lancashire, but also in some parts of Scotland. It is a tiny part of the world, and it remains concentrated in this tiny part of the world for one generation, maybe almost two. It then spreads—the cotton textile industry spreads to Belgium, to France, to Prussia, to Saxony, to Italy, to the young United States. It spreads to Egypt, to Mexico, and to various other parts of the world.
What is really key, however, is not that people invented a new way of making cotton textiles and then produced ever more of them. That alone would not have been an industrial revolution. What happened is that this ability to increase production, to increase economic output, and to vastly increase human productivity spread to other industries.
By the 1830s, the cotton industry in Great Britain is still very large and remains an important economic sector, but it is no longer the core of the British economy. The focus moves to other things such as coal mining, iron and steel production, and then the railroads, which are a huge sink of capital investment but also a market for coal, steel, and machinery.
Subscribe now
The real core of the Industrial Revolution is not the sudden increase in productivity or the recasting of production in one particular industry, namely cotton, but the fact that the revolution becomes a permanent condition. It never stops, and it does not stop to this day. There is always something new. It begins with railroads, steel, iron, and machinery, but by the later 19th century it includes more science-based industries such as chemicals and electrical machinery.
The Industrial Revolution also spreads to other parts of the world in more substantial ways. Both the United States and Germany become much more important to the Industrial Revolution, and even more important than Great Britain itself. With the Industrial Revolution, as we can now see in retrospect, this is the beginning of a state of constant revolution in the technologies of production, in productivity, and in total economic output.
Mounk: That permanent revolution can be thought of as an expansion of capitalism across a number of different dimensions. There is geography: it starts in one part of northern England and then spreads to other areas. In terms of industry, it starts with cotton and then moves to steel, coal, and the railroads. It can also be understood in terms of realms of human life and activity.
To name one relatively contemporary example, dating was not really part of a capitalist process, but now most people meet each other on online dating apps for which they pay subscriptions. In some ways, a realm of human life that had been outside of the capitalist sphere has now been more firmly subjected to market forces.
Beckert: Both processes happen. There is an increase in productivity and more or less constant economic growth. Of course, there are economic crises, but if looked at in the long term, it is astounding economic growth.
There is also the process in which the logic of capital that we spoke about at the start of this conversation—the one we saw in merchant communities in West Africa and the Arabian Peninsula, in Florence, and elsewhere—was once very marginal to economic life on Earth. This logic spreads at first slowly, but then in the wake of the Industrial Revolution ever more rapidly, into all kinds of forms of life.
People throughout much of human history produced the food that they ate and manufactured many of the things they consumed, such as clothing and shoes. As we all know, that became ever less prevalent, and ever more parts of economic life, but also life more broadly, became subsumed under this logic. When it comes to dating, it also becomes subject to investments of capital for the returns on that capital and for its growth. That is clearly the animating force of the capitalist revolution.
Mounk: What is the role of finance in all of this? Something like the Monte dei Paschi Bank in Siena was founded a very long time ago. When we were talking about the conquest of the Americas, you needed substantial capital in order to invest in those sea voyages. That capital was provided by monarchs, who often needed money that they either raised from their subjects or borrowed from rich people, even at the time.
The role of finance, and the primacy of finance in the capitalist system, really accelerates in the second half of the 20th century. As you describe it, it is now one of the characteristic features of this stage, this particular shape of capitalism. Tell us a little bit about how finance slowly grows as an element of this process and then becomes so dominant over the course of the last decades.
Beckert: That is an excellent and very important question. In some ways, the world we live in today with regard to finance has more in common with the world of the very early history of capitalism than with the moment of the Industrial Revolution that we just discussed.
Merchant capital and associated finance capital, including the banks you mentioned, were absolutely at the core of the early capitalist revolution. The East India Company, for example, which was the most important company worldwide for a very long time, was closely associated with these accumulations of capital among merchants and bankers in cities such as Amsterdam, London, and elsewhere.
International trade required significant capital investments, and the only people in the pre-modern or even medieval world who possessed these kinds of accumulations of capital were merchants and their associated financiers—the bankers.
With the Industrial Revolution, something changed quite fundamentally. For the first time, people could invest in production even if they had limited capital resources. Starting a cotton mill was not a particularly expensive endeavor. For the first time in world history, people could accumulate capital and become rich not from trade or banking primarily, but from investing in production itself. This early production, in various industries at various times, was extremely profitable. People earned very high returns on their capital.
Some individuals who started out as skilled workers in the cotton industry could accumulate so much capital that by the end of their lives they owned their own cotton mills. They were rich in certain ways. This ability to accumulate large amounts of capital in production itself continued well into the 20th century. Think of Henry Ford. Henry Ford did not like bankers or finance. He financed his expanded production entirely from retained earnings.
This marked the rise of the industrial capitalist, and merchants became less central to global capitalism. However, this has now shifted again. By the 1970s, and especially in recent decades, finance capital and merchants have become much more important again to global capitalism. In this way, the system connects back to the very early history of capitalism. The period during which industrial capital held the leading role in global capitalism, in retrospect, was quite short—just a hundred or a hundred and fifty years or so.
Mounk: I take the point about industrial capital, that this is not, by and large, how people make their fortunes today. I am sure there are some people who have made their fortunes in industry and manufacturing. There are probably a good number of them in China and in Vietnam and in places where much of the world’s production happens. But those are not the people at the very top of the Forbes list of the richest people.
I am trying to understand how different the situation today is from back then. It is striking that the very richest people in the world today, by and large, are not people who have inherited money. They are not people who went to Wall Street at the age of 21 or 22 after getting an economics degree at Harvard or Stanford. They are people who founded their own companies. Of course, they borrowed money to found those companies and had investors.
I do not know enough about those earlier titans of industry to know whether they did the same, but I would imagine that at some point Henry Ford took out a loan from a bank or took investments from other people in his company as well. When you think about how Elon Musk has made his money or how Mark Zuckerberg has made theirs, is the role of finance so much more central to their story than it was in the rise of Henry Ford?
Beckert: I do not know how to measure that exactly. There is a lot of inequality in capitalism, as we know from Thomas Piketty, and this inequality has sharpened in the past few decades. A capitalist economic order, unlike a feudal economic order, always contains within it the possibility of social mobility. At all moments in time, we can see in a capitalist economy that people with modest resources have struck it rich.
In the 18th and into the 19th century, merchants and financiers were clearly the central actors within the capitalist economy. They were the ones who struck it rich. They were the ones who dominated production. They organized textile production in the countryside or sugar production on the plantations. This changes in the 19th century, when mechanics and people with technical skills begin to create dynamic enterprises and accumulate significant amounts of capital in industrial production. They become the central actors within the capitalist economy.
Today, there is still production. You are absolutely right about China and Vietnam; that is absolutely true. But think about the textile industry today or the shoemaking industry. You have a few globally active brands that are, broadly speaking, agglomerations of finance capital or merchant capital, and you have hundreds of thousands of producers in all corners of the world who actually make these shoes. The power relation is very clear. The large brands, which often do not produce anything themselves but simply organize production, dominate the industry. The manufacturers themselves are quite powerless.
This does not mean that people do not make money producing things today. They do make a great deal of money. But the relative power, if we think about capitalism as a system in which there are merchants, industrialists, and bankers, has moved toward merchant capital and finance capital in recent years and away from the General Motors and Henry Fords of the world.
Mounk: That is really interesting. I am struck by the fact that many people who write about the history of capitalism or who are critics of capitalism are attracted to the term late capitalism or late-stage capitalism. It is not a term that, so far as I know, you use, or at least use very much. Is there a kind of naive assumption buried in that term that, in a nearly neo-Marxist way, capitalism is about to founder and about to end, that you are skeptical of?
How do you think about this widespread talk about late-stage capitalism? Is there any way of knowing whether we are toward the end of this remarkable story, in the middle of it, or only at the very beginning? Will the next thousand years of human history continue to be driven by this process playing out?
Beckert: We do not know. The future will, for certain, be full of surprises, for better or for worse. You are absolutely right that I do not use the term late-stage capitalism. People who have written about capitalism have been describing its late or latest stages for more than 150 years. Predictions about the end of capitalism have been made since at least the mid-19th century, and they have always been described as imminent. None of that has happened.
You could argue that the full force of the capitalist revolution only occurred after the first predictions of its immediate demise. Having looked at the history of capitalism globally and across a thousand years, I see a great deal of energy and an ongoing ability to innovate in this system of organizing economic activity, even today.
Capitalism is historical. It has a history. It is not the natural state of organizing economic life on planet Earth. Therefore, it has a beginning. As any historian will tell you, anything that has a beginning also has an end. Presumably, at some point, another form of organizing economic activity will emerge. I do not know when that will be or how it will unfold. Capitalism, for better or worse, has shown itself to be enormously dynamic.
The secret to capitalism’s dynamism has been that it is generally undogmatic. It has changed drastically in the course of the past 500 years. Things once seen as absolutely essential to capitalism have become completely marginal a few decades later. Capitalism has always connected and reconnected different forms of labor, political organization, capital, and territorial structure. It has never been dogmatically linked to just one configuration of these elements. Presumably, that will remain the case in the future.
There is one caveat. The capitalist revolution has drawn on the free gifts of nature. It has drawn tremendous energy from capturing natural resources, the productivity of the land, the unpaid labor of women, and fossil fuels. There are limits to that, as far as we know. This might become a more systemic constraint on the further expansion of capitalism.
People called it late-stage capitalism in the 1960s, and now we are more than sixty years beyond that. What are we supposed to call it now—post–late-stage capitalism? That does not make sense to me. Capitalism shifts its shape drastically. The capitalism of today looks drastically different from that of the 1960s, which looks drastically different from that of the 1920s, which in turn looks drastically different from that of the 1850s. That is all true.
But the basic logic of the organization of economic life has remained the same. So when does early stop and late start? I do not think we can tell. Future generations of scholars, a few hundred years from now, will know.
Mounk: If there is one prediction you are willing to make, it is that if we still have the capitalist system in 60 years, or 600 years, or 6,000 years, it will be very different from today. The one in 60 years will be different from the one in 600 years. Whether we will still call it capitalist, and whether it will still meaningfully be the same system, we cannot predict.
I am interested to return to some of the assumptions that you started by rebutting about capitalism and to the language you have come back to several times regarding what is at stake in denaturalizing capitalism. This is something that always struck me when I was taking various classes in graduate school and as an undergraduate. I am sometimes struck by the way my students think about it as well.
When scholars in history and sociology, and to some extent in political science, aim to demonstrate that something that appears natural is not natural, they sometimes describe it as a social construct. I am both sympathetic to that enterprise and worried that it can confuse people in certain ways. This is a side note. You have not talked about social constructs, and I think the term itself can cause particular confusion. Students often think that if money is a social construct, it somehow means it is fake or does not exist.
The example I often give is that tea is a social construct. The English word tea includes chamomile tea, whereas the French word thé does not include chamomile tea because that would be infusion or tisane. This shows that the term is a social construct; it differs depending on the cultural and social context. It does not mean that tea is not real or that it is fake.
To return to the idea of denaturalizing capitalism, one question I have is: to what extent is this surprising? Among thoughtful, educated people, how many truly believe that capitalism is a kind of God-given system with fixed attributes that never change over time? And perhaps more importantly, what follows from denaturalizing it?
Beckert: It is really important to think about that. Let me start by saying that I do not think capitalism is a social construct. Capitalism is very real. It is the most powerful process impinging upon our lives today, as well as upon the lives of our parents and grandparents before us. It is a reality, and we need to understand it as such.
The problem is that any social reality powerful within your own life and time is the hardest to see. In some ways, when we look back at Stone Age economies, or even at the feudal economy of 12th-century France, we immediately see that they are strange and different from the world we live in today. Therefore, we are quite capable of understanding them because they so clearly call for explanation.
Mounk: They would not have seemed strange to the people living in a feudal order, in the same way that capitalism does not feel strange to us.
Beckert: Exactly. We take the prevalence of markets, the fact that things can be commodified, and the fact that most of us sell our labor power in markets for wages. We take that to be how the world works and perhaps how it has always worked, but it has not. It is a radical departure from previous human history. We need to start by saying that this is not natural, not common to all of human life. It is a radical, even revolutionary, departure from the organization of economic life in the long past.
That has real implications for how we think about the present as well. Once you see that this is a world that is not natural—a world that is created, a world created by human agency, by people making certain decisions, political decisions, daily life decisions, social decisions—you see that they shape that world. If we can see that the world we live in today, the world of contemporary capitalism, is a world created by us, then we can also see that we are not merely its victims.
It is us—not individually, but collectively. People have different levels of power in creating this world, but it remains a human-created order. Therefore, this order can also be changed in ways that are more conducive to our shared interests. We are not merely victims of this order; we are also the creators of this world.
Sometimes it feels almost like artificial intelligence, as if it is out there and coming to dominate all of us. The logic of capitalism has somewhat similar characteristics to the logic of artificial intelligence. But we created artificial intelligence just as we created capitalism, and therefore we can change it in ways that are conducive to making life better.
Capitalism has the logic of needing ever more access to ecological resources, and we know that this threatens our very existence on planet Earth. Yet we are not powerless in the face of that. We can act. It also helps to see, as I said earlier, that there is not just one capitalism. Capitalism has taken very different forms at different moments in time, and it is entirely imaginable to have a different kind of capitalism in the future.
Mounk: It is this connection that I was curious about because I can see how one proposition seems to naturally flow from the other. If you think that capitalism is somehow natural or God-given in an ever-changing form, then it seems to imply that there is really nothing we can do to change or regulate it. Whereas if it is the product of these complex and contingent historical processes, it seems to invite agency. It seems to invite the thought that perhaps we can reshape it today.
It is not clear to me how strongly that inference follows from the premise. The way you describe the history of capitalism, there is both a lot of human agency involved and very little human agency involved. It is the product of the endeavors, ambitions, and decisions of millions of people across geographies and across a millennium of time. But the Medici in Italy, the factory owners in England, and the traders in China were not aiming to produce or shape the capitalist system. They were acting on their own motives and incentives in their particular moment, and they created this world.
Over the course of the 19th and 20th centuries, we have seen the ability to shape capitalist institutions in important ways. We have welfare states today in North America, Western Europe, and many other parts of the world that certainly shape capitalism. That is a very good thing. It makes the system much more humane and sustainable and allows us to harness its energy in a much better way for human good. I share some of those underlying goals.
However, one might also think that although the system is contingent, man-made, and the product of these complicated processes, it is also an incredibly powerful machine. It is a process that is, in some ways, a perpetual motion engine, pushing forward in a powerful way that we might not be able to reshape. It might not be a matter of the will of one politician, one citizen, or even all of us collectively to fundamentally reshape the system.
If the United States or Germany were to try to reshape the system in a very radical way that was not the most productive use of economic resources at the time, those countries might be outcompeted by others, and that experiment would fail. This may be one mechanism that shows why the system could be less subject to human agency than this emphasis on denaturalization might imply.
Beckert: That is an excellent point. I do not want to suggest that we just have to sit at our desks and invent a different kind of world, and that by our willpower we will suddenly bring about this world. I want us to see that much of economics has the tendency to emphasize laws that are allegedly timeless in capitalist development, as though the best we can do is read them properly and implement them properly, which will then make us wealthier collectively as a result. I do not think this is the case.
The capitalism of the 1960s, with strong welfare states and a significant presence of the state in regulating economic activities and even owning firms, is drastically different from the capitalism of the 1920s or from the capitalism in which we live today. Power is radically unevenly distributed, and certain actors have far more power in shaping our world and our economic lives than others.
To cite the most unlikely example, in an earlier moment of capitalist history that we discussed, slavery—the plantation economies of the Caribbean and the American South—was extremely important. Many people, including European observers who were not adherents, believed the world would be better off without slavery but concluded that for the capitalist economy to thrive, slave labor was unfortunately necessary. Yet slavery came to an end, and it did so very significantly because the least powerful, least educated, poorest people on planet Earth—the sugar workers of Saint-Domingue—staged a revolution that eventually led to the abolition of slavery. This was one of the most significant shifts in the structure of capitalism ever.
The same is true of the welfare state. Industrial workers, who were often poor and without much political power, gained the franchise only in the 20th century in most parts of the world. They were the ones who pushed for the expansion of the welfare state and for the recasting of capitalism. So even though there is a radically unequal distribution of power in shaping the world in which we live, in the long history of capitalism, people without access to many resources have still made a huge difference in its structure.
I think we can make a difference. The entire book, as you noted, is written as an actor-centric history. It is not a story about something that fell out of the sky, to which we are subject as passive participants in abstract economic laws. I try to show that merchants, industrialists, workers, enslaved people, and those working in the countryside all acted upon these forces and created the world in which we live today. That is the main takeaway of the book.
In the rest of this conversation, Yascha and Sven discuss why capitalism developed the way it did, and whether we should be optimistic about this economic system. This part of the conversation is reserved for paying subscribers…
There is a lot that I love about America. Thanksgiving is one of the things I love the most.
More than any other date on the calendar, Thanksgiving reminds us what we supporters of liberty are really fighting for.
“More than the Fourth of July?” you might ask. “More than the date on which we commemorate American democracy, freedom, and independence?”
Yes, more than the Fourth of July.
Democracy, freedom—politics, too. These are not ends in and of themselves. They are a vehicle for delivering human happiness and flourishing. That goal is what we’re fighting for.
The notion of a free society is abstract. Thanksgiving celebrates abundance, and abundance is tangible. You can taste it. Smell it. Hear it. The turkey and mashed potatoes on your plate, the chatter with loved ones, whom you’re free to visit—these are the fruits of a free society.
Subscribe now
President Abraham Lincoln formally consecrated Thanksgiving as a national holiday in the middle of the Civil War. His proclamation is full of vivid imagery of America’s bounty. Lincoln opens by recalling:
The year that is drawing towards its close, has been filled with the blessings of fruitful fields and healthful skies.
He continues:
Needful diversions of wealth and of strength from the fields of peaceful industry to the national defence, have not arrested the plough, the shuttle or the ship; the axe has enlarged the borders of our settlements, and the mines, as well of iron and coal as of the precious metals, have yielded even more abundantly than heretofore.
Speaking of abundance, you may have heard the story of Boris Yeltsin’s visit to an American supermarket.
It was September 1989. Yeltsin had been installed as a member of the Supreme Soviet of the USSR just a few months earlier. He took a tour of Johnson Space Center. I’m sure he was impressed, but the Soviet Union had rockets and even an imitation space shuttle too.
But the Soviet Union did not have abundance like America did.
On the way back from the space center, Yeltsin asked to visit a supermarket. The store was given only minutes’ notice that a senior Soviet dignitary would be dropping in.
Yeltsin wandered the aisles, amazed at the variety and volume of products, colorful popsicles and frozen treats; plentiful produce, all painting a stark contrast to dreary communist scarcity. Yeltsin’s associates and biographers describe the transformative impact of the grocery pit stop. It further convinced the future Russian president of the brokenness of the Soviet system.
The store Yeltsin visited was a medium-sized outpost of Randalls—a Texas grocery chain. In America, there are and were far more palatial supermarkets. One wonders how Yeltsin might have reacted to the vastness of a Costco or the opulence of an Erewhon.
But in September 1989, an unremarkable local supermarket was sufficient to capture Yeltsin’s imagination. Eight weeks after his visit, that Randalls location was undoubtedly crowded with Americans stocking their kitchens for Thanksgiving.
Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal democratic values wherever they are threatened, please consider becoming a paying subscriber today!
Years before Yeltsin’s trip to Texas, I had my own “American supermarket moment”—actually several—when I experienced, viscerally, the difference between free and unfree.
As a top chess player, I enjoyed privileges my neighbors in the Soviet Union could not even dream of.
I got to travel to Western democracies to compete. Assuming they even had the means to travel, a Soviet citizen could count themselves as extremely fortunate if they got approval for a holiday in Bulgaria. I was playing chess in places like New York, London, Paris, and Amsterdam.
Of course, that freedom was conditional: Before I turned 20, the KGB would not allow my mother to travel abroad with me to competitions, essentially holding her in the USSR as human collateral to discourage defection. (Remember what I said about the ability to freely visit loved ones?)
The only chess title I never won was the under-16 championship. I tried, first in 1976, and then again the following year. The competition took place in northern France. My mother, always an essential, supportive presence at my matches back home, was replaced by a state-appointed coach. I was a 13-year-old kid from Baku—from the fringes of the Soviet empire—with only a stranger as my company in a strange land. It was a deeply uncomfortable experience, and I just couldn’t manage the win.
Still, the Soviet authorities could not hold a blindfold over my eyes. I, too, was exposed to the fruits of a free society, and from a much younger age than Boris Yeltsin.
In France, I met a family who would cross the border into Belgium each morning to buy milk. They were confident there would be milk readily available to buy. That shattered my sense of travel as a deeply proscribed activity. No commissar had to sign off on that family’s daily excursion. Free societies enable freedom of movement.
Four years later, in 1980, I was in Italy, on my way back from my first chess olympiad in Malta. During a long layover, I caught The Empire Strikes Back. It was my first experience with American cinema that hadn’t been filtered by a censor and a dub. What a drag to leave the theater knowing I was returning to a country where the Darth Vaders ran the show. Free societies foster creative expression; unfree ones force it underground or wipe it away altogether.
In 1991, in the final days of the Soviet Union, I spoke at a rally—as it happens, in support of Boris Yeltsin’s presidential campaign (we had higher hopes for him back then). I stood on Manezh Square, just beyond the Kremlin walls. I was confident; I’d been in the public eye for some time. But, standing in front of hundreds of thousands of people, I briefly found myself at a loss for words.
When I collected myself, I recounted how I’d recently been to America. There, I saw long lines in only one place: Disneyland. I told the crowd that we ought to have a country where people queued for amusement park rides and not for food.
Free societies deliver abundance.
These days, there is a lot of doom and gloom about the United States across the political spectrum. I am not talking about America’s current democratic and institutional crisis, which is indeed deathly serious. I am referring to the short-sighted ideological decay that is increasingly popular with radicals of all stripes: on the right, the perception of America as sinful, deviant, and overly tolerant; on the left, the view that America is criminal, colonial, illegitimate.
The exhortation to make America great again implies that America is no longer great. Far-left self-flagellation tells us it never was.
Americans would do well to discard these self-destructive narratives. It may be hard to describe what lofty concepts like democracy and freedom really mean, but you can see the rewards of those concepts all around you if you’re willing to open your eyes.
If Abraham Lincoln could find time for gratitude in the middle of a deadly Civil War, Americans today can too. If Yeltsin could be so impressed by a grocery store many Americans might consider average, then you have something to be thankful for. I’ll dispense with the caveat that America isn’t perfect (what country is?). If you are thankful for something, then you have something you can fight for.
Happy Thanksgiving.
Garry Kasparov, the 13th World Chess Champion, is an author, speaker, and chairman of the Renew Democracy Initiative, the publisher of The Next Move.
A version of this essay originally appeared in The Next Move.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
One of the main tactics of the second Trump administration has been the frequent use of threats and legal action against people who, for whatever reason, have attracted the ire of the government.
The latest such example is the Pentagon’s investigation into Senator Mark Kelly for a video in which he urged members of the military to refuse to obey illegal orders. In the video, published online earlier this month, Kelly, along with five other lawmakers who are veterans of the military or intelligence services, said the following:
This administration is pitting our uniformed military and intelligence community professionals against American citizens. … Right now, the threats to our Constitution aren’t just coming from abroad, but from right here at home. Our laws are clear. You can refuse illegal orders.
On November 20, Trump took to social media claiming that these words were “SEDITIOUS BEHAVIOR, punishable by DEATH!” He later rowed back on the idea that he was calling for the death penalty, but continued to insist that the words amounted to sedition. On November 25, the Department of War posted a memo on X in which Secretary Pete Hegseth ordered Navy Secretary John Phelan to investigate “potentially unlawful comments” made by Kelly. This came a day after the Pentagon posted another statement about “serious allegations of misconduct” it had received against Kelly, going on to say: “All servicemembers are reminded that they have a legal obligation under the UCMJ [the Uniformed Code of Military Justice] to obey lawful orders and that orders are presumed to be lawful.”
There are several things to say about this latest flare-up.
First, the lawmakers’ video did not appear in a vacuum. On multiple fronts—the ICE raids against illegal immigrants in multiple American cities, the strikes in the Caribbean Sea against ships carrying alleged Venezuelan narcoterrorists, the government’s use of force to quell protests in cities like Los Angeles and Portland—the Trump administration is enacting policies that require the direct participation of military and intelligence personnel, and which many observers argue are illegal. The situation is complicated by the fact that some of the government’s actions (such as the domestic deployment of troops) have been challenged by judges, whereas others (such as the strikes against civilians accused of narcoterrorism) have not yet made it to the courts. Yet these cases were surely on the minds of the lawmakers who made the video. The administration’s current actions are the clear context of the remarks.
Subscribe now
Second, the language used in the video is extremely vague. The lawmakers speak in broad terms about the requirement that the military and intelligence communities uphold the Constitution (which is trivially true), and make the aforementioned (extremely vague) allusion to the government’s current actions. In part, this vagueness is probably because the legality of many of these actions is at least contestable, however tenuously. On those matters which have so far come before the courts, the judges have often ruled against the government. Still, members of the military and intelligence communities would want to be absolutely certain that an order is illegal before disobeying it. This would require the illegality of the order to be clear on its face, per the “manifestly unlawful” test. But the way the system works, courts usually determine whether the actions of the Trump administration are legal over the course of weeks or months, and they don’t always rule in the way legal observers predict. The kind of situation in which blatantly illegal orders come down the chain of command, providing an officer an unambiguous opportunity to disobey them, is quite rare. This might explain why the lawmakers were unwilling to single out any specific action for criticism in their video.
Third, most Americans would unambiguously have the right to express such a message thanks to the First Amendment. But the rules are different for military and intelligence personnel. Kevin Carroll, a lawyer and retired military intelligence colonel who served as senior counselor to the Secretary of Homeland Security during the first Trump administration, told me via email that “unlike a civilian, a servicemember can be court-martialed for disrespecting a superior officer, or using ‘contemptuous words’ against elected officials.” But it would be hard to argue that this is what Kelly did: He “just accurately cited military regulations. Most every commander and every judge advocate in the military has made some version of Kelly’s remarks to the troops in training or before a deployment.”
Even if Kelly had breached the Uniformed Code of Military Justice, the fact that he is retired should shield him from prosecution. As Carroll told me, “a court would reasonably find that this power does not apply to alleged crimes committed after retirement, only to crimes committed by officers on active duty about which fresh evidence is subsequently discovered, newly implicating a retired officer in past misconduct.”
Fourth, the government had a point when it argued that “orders are presumed to be lawful.” As David French writes in The New York Times: “You can’t fight a war… if every soldier acts as an independent legal check on every order he or she receives. Individual service members don’t have sufficient knowledge or information to make those kinds of judgments. When time is of the essence and lives are on the line, your first impulse must be to do as you’re told.”
As such, the vague and uncontroversial appeal to constitutional principles made by Kelly and others in their video does not clarify the very real stakes of applying these principles in the real world. The duty to disobey illegal orders has been given special emphasis in the military ever since the infamous My Lai massacre of 1968, in which Lieutenant William Calley and his soldiers murdered hundreds of unarmed Vietnamese civilians. But as one retired lieutenant colonel told me via email, things get tricky when the orders originate from much higher up—from cabinet members, or indeed the president. In such cases, it may be “the responsibility of senators, members of Congress, the courts, and the American electorate”—not individual troops—to judge the legality of the orders they receive.
Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal democratic values wherever they are threatened, please consider becoming a paying subscriber today!
Fifth, the government’s response to the video has been predictably hysterical and autocratic. There should be a strong norm against levelling threats against military veterans, especially if they are sitting members of Congress. To give the government the power to court martial any elected official who strongly criticizes the president under suspicion of treason would obviously create an enormous chilling effect. The Pentagon’s investigation of Kelly violates that important norm in an extreme way.
Furthermore, as countless observers have pointed out, the idea that Kelly is guilty of sedition is laughable. Kelly was right when he described the administration’s response as blatant intimidation. It matches the characteristic tactic of the government in recent months: Go after prominent figures who resist or rebuke you, find every loophole, exploit every misstep, clog up the courts, put your enemies through hell fighting to clear their name, and mobilize your supporters to harass them—all in the name of enacting a policy platform that is, at best, erratic, and at worst, illegal and unconstitutional.
And here lies the difficulty. Kelly and co. were well within their rights to publish that video. If—like me—you are deeply worried about a presidency which is blatantly overstepping the boundaries of its rightful authority, you should applaud their courage in reminding servicemen and women that they have duties to the Constitution which take priority over their loyalty to particular commanding officers. Anyone familiar with the 1964 Kubrick classic Dr Strangelove will recall the iconic moment in which one of Peter Sellers’ characters, Group Captain Mandrake, attempts to resist an insane general who wants to launch unilateral nuclear strikes against Russia. Yet the world we live in rarely permits such moments of clarity: Ours is a world in which orders are legally dubious rather than blatantly insane, and come from the top of the chain of command, rather than from rogue generals.
There may come a time when members of the military will be faced with orders that are clearly illegal. They will then have to make a judgement about how to act, possibly in the course of minutes or seconds. That is not a long time to balance the necessity to (in French’s words) do as they are told against the necessity to (in Kelly’s words) obey the Constitution. And it will be all the harder for them to do so in a principled manner if they have legitimate reason to fear that any hesitation may lead to the most powerful man in the world calling for their execution.
Luke Hallam is senior editor at Persuasion.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
Nine years ago, then-venture capitalist and lawyer J.D. Vance published his memoir Hillbilly Elegy. A vivid and thoughtful retelling of his life growing up in a poor Appalachian expatriate family in post-Industrial Ohio, the book was in many ways a typical American rags to riches story. Yet Vance also tried to offer insight into the unique challenges and problems that face Appalachia in particular, and rural America more broadly.
As someone who grew up in Appalachia and has lived there most of my life (with only a brief pause to attend the University of Virginia for my doctorate), Hillbilly Elegy has loomed large in my cultural imagination, and that of my neighbors, for some time. Before he was known as “J.D. Vance the heir apparent to the MAGA movement,” he was “J.D. Vance the traitor to his people,” who had written a book airing the entire region’s dirty laundry. (To the more generous, he was “J.D. Vance who is not really from here but wrote an okay book about us.”)
Since his rise in politics, all but the most ardent supporters of President Trump seem to have adopted the more cynical view of the book, regardless of what they thought when it was first published.
The truth that few people like to admit is that Vance captured many of the problems that plague rural America. Over the course of the book, he lays bare the devastating effects of job loss, the drug epidemic, generational poverty, and broken families. All of these problems stem in no small part from the economic decline that has beset the rural communities of this land. What’s more, he outlined how many of the people in his post-industrial town did not really want to work hard even when a job opened up; he revealed that many struggling families, rather than prudently spending what little money they had, spent it in frivolous binges; and he showed how the communities of Appalachia can often be far too insular and opposed to necessary change.
In short, Vance showed that for all the incredible warmth, honesty, and hard work that define the people of rural America, there are particular faults in the local culture that perpetuate real problems.
Vance correctly locates the source of many of these problems in the crushed spirit of Appalachia. As Professor Arlie Hochschild recently captured in her account of visiting my home county in Kentucky, many rural Americans feel as though they have lost all sense of belonging and meaning in the modern world. Most long, more than anything, for the restoration of that meaning, while others are so spiritually crushed that they turn to drugs or TikTok to fill the void.
Over the past decade, Vance has transformed his diagnosis of the problems in his home state into a political project. From his vantage place as a senator and now as vice president, he has, with characteristic smarm and verve, sought to articulate a program rooted in a traditionalist nationalism that seeks to revitalize Appalachia and the country beyond.
Being an unusually well-read and thoughtful politician, Vance’s nationalism is not the screaming, reactionary variety one finds at far-right political rallies. He articulates a clearly-defined ideal of nationhood, according to which a nation is more than a simple geographic territory, or even a project dedicated (as the Founding Fathers might put it) to a rights-bearing creed. In Vance’s account of the nation, it is shared history and traditions that really define who we are and who we must be going forward. He openly disdains abstract ideals like life, liberty, and the pursuit of happiness in favor of residential permanence and shared cultural norms.
Subscribe now
Vance’s vision has several policy implications. First, he wishes to revitalize American manufacturing. We are a nation, in his view, of workers, and to lose our tradition of manual labor is to lose a large part of what it means to be American. Secondly, those with strong hereditary claims to citizenship must receive first priority over those who arrived more recently. This cannot be achieved without cracking down on immigration.
Yet while Vance accurately diagnoses many of rural America’s problems, his solutions come up short.
The collapse of rural towns, small industrial cities, and remote farms has coincided with the decline of local cultures. A local identity brings with it pride of place and a certain willingness to live with the disadvantages endemic to the location. When people feel that their locality serves a purpose—that it is embedded within a larger whole—they are willing to tolerate or even embrace its remoteness, slower pace of life, and faulty infrastructure. Rural Americans once thrived on a belief that for all their region’s faults, they were the backbone of the nation.
But Vance’s concept of the nation does not restore this sense of local pride. Instead, it substitutes a globalized vision of tradition for a local one.
The proof surrounds me every day in my native Central Appalachia. As a child, most local businesses in my neighborhood seemed to identify first and foremost with East Kentucky. Many bore names like “commonwealth insurance” or “mountain music.” Though people in the region were patriotic, the primary emotional attachment was regional and not national.
Yet since Vance—and MAGA more broadly—have encouraged a strong dose of nationalism in red states, this has changed. Now when I cruise down the highway I am greeted by “Patriot RV” or “American Laundry and Cleaners.” This is a subtle but telling shift.
Subscribe now
MAGA-style nationalism has proven itself incapable of stemming the tide of rural decline. The past ten years have not seen a rejuvenation of rural America, nor do there seem to be any indications that the economic trends of the last half century are starting to reverse in these regions. Instead, rural communities are finding themselves sinking further and further into economic stagnation. Their population continues to decline at a rapid rate and rural voters feel just as spiritually displaced as ever. In short, as nationalism has risen to the heights of American political power, rural communities have continued to sink to the bottom of the economic abyss.
This is not a new phenomenon. Years before modern globalism emerged, nationalism was already eroding local cultures and institutions. You could even say that the nationalism that Vance praises—embodied by figures such as Theodore Roosevelt and Otto Van Bismarck—paved the way for the internationalist outlook he so openly disdains. By undermining the power of local community, nationalism weakened the most natural and deeply rooted ties individuals posses. After such ties have been cut, the ethical argument propounded by numerous religions and philosophies—that all humans are one community regardless of place—was able to exert its influence without substantive opposition.
Champions of rural America must reject reactionary nationalist attempts to rewind the clock back to the 1950s. They need a totally new solution to rural malaise—one that combines the localism of the past with the values of the open society that will likely dominate the coming century. This is not an easy task; it asks us to combine two things that have not historically gone well together. Yet it is the only real hope for rural America.
How might it come about?
The first step is for government and civil society to rejuvenate local cultural institutions. For instance, in my native Eastern Kentucky we should work hard to make local newspapers a strong cultural force again. We should restore historical buildings and landscapes, beautifying cheap utilitarian constructions so that they fit with the vernacular culture, and preserve local environments. We should refurbish the folksy brick buildings native to this region, tearing down or renovating eyesores, and conserving the stunning Appalachian Mountains that are the physical home of my people. Perhaps most importantly, we should invest in local art, music, and culture. This could mean cultivating a serious appreciation for bluegrass music, Appalachian literature, and local history. Learning to appreciate the culture of one’s home is a guaranteed path to restoring a sense of purpose to the locality.
Persuasion is also the home of American Purpose, Francis Fukuyama’s blog, and the Bookstack podcast! To receive all of this great content, simply click on “Email preferences” below and make sure you toggle on the relevant buttons.
Email preferences
Furthermore, rural America needs a robust middle class. The characteristic values of modernity—pluralism, innovation, equality, freedom—have not always taken easy hold in the remote regions of this country. They are the values of a middle class that can afford to worry about such things. To cultivate a strong rural middle class will be no easy feat—it requires rethinking the basic workings of our economy, which has come to favor centralized power over place-based needs. Yet this is what regions like Appalachia need if they are to adapt to a modern economy.
All of this requires investment in the sort of infrastructure it will take to update rural America. This includes roads and airports, but also technological infrastructure like fiber optic cables. Remote, idyllic towns with a unique culture are uncommonly well-suited to hosting those who wish to do remote work while raising a family. Likewise, industries that America will never feel comfortable farming out to other countries—like nuclear energy and weapons production—could naturally be housed in rural locations.
Rural America is in serious trouble. Vice President Vance deserves a great deal of credit for being one of the earliest and most eloquent voices to sound the alarm. Yet Vance and the nationalist MAGA movement offer a solution that is every bit as bad as the problem—one that offers no hope for the rejuvenation of a struggling people. My neighbors, friends, and coworkers in rural America need a politics that restores localism while updating Appalachia for the era of technological modernity. We have no use for the nightmare of nationalism, which will undoubtedly kill the patient before it cures the disease.
Jeffery Tyler Syck is an Assistant Professor of Politics and the Director of the Center for Public Service at the University of Pikeville in his native Kentucky.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
Nate Soares is president of the Machine Intelligence Research Institute and co-author, with Eliezer Yudkowsky, of If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All. He has been working in the field for over a decade, after previous experience at Microsoft and Google.
In this week’s conversation, Yascha Mounk and Nate Soares explore why AI is harder to control than traditional software, what happens when machines develop motivations, and at what point humans can no longer contain the potential catastrophe.
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: You have just written a book that is already on the New York Times bestseller list with a cheery title. I am used to writing books with not-so-cheery titles, but this one takes the cake. If Anybody Builds It, Everyone Dies. What do you mean by that?
Nate Soares: I mean that if anyone builds artificial superintelligence, which is to say an artificial intelligence that is better than the best human at every mental task, then the most likely outcome of the creation of that technology is literally everybody on the planet dying.
Mounk: That is a scary and stark thesis. I really enjoyed—if that is the right vocabulary, the right verb to use in the context of this conversation—the way that you laid out in the book your reasoning about what AI is and why that makes it so incredibly hard to control. Perhaps we can go through that step by step. One of the key premises you make is that AI is grown rather than built. What do you mean by that?
Soares: AI is not like traditional software. In traditional software, when the software behaves in a way the creators didn’t intend, they can debug it and track it down to some line of code or some interaction or some piece of the software that they wrote that was having some interaction they didn’t understand. Then they can say, oh, whoops, I understand it now. They can usually fix it and get the software to behave how they want. Modern AI is nothing like that.
When AIs threaten reporters with blackmail and ruin, when AIs try to avoid being shut down—and these are cases we’ve already seen, the former in Sydney a few years ago and the latter in somewhat contrived lab conditions—the creators can’t read through and find some misbehaving line of code and say, whoops, somebody set threaten reporters to true. Let’s set that to false. The way modern AI is made is people assemble huge computers, huge amounts of computers. They assemble huge amounts of data and there’s a process for tuning basically every number inside those computers according to every unit of data.
They’ll be tuning these little knobs, essentially trillions of knobs, trillions of times. Humans understand the process that does the tuning. At the end of that tuning process, the machine can carry on a conversation. Nobody knows quite why. We understand a device that runs through twiddling knobs, and if those knobs are at just the right setting that you get to after a year of twiddling, the machine’s talking.
We don’t quite understand how it’s talking. If it starts saying things nobody ever intended it to say, like threatening a reporter, we can’t figure out exactly why. This is a very different paradigm, and it leads to these AIs acting in ways nobody asked for.
Mounk: The contrast to a traditional computer program here is really helpful and instructive. I don’t have a computer program background at all, but I did once take CS50 as a remote student, the famous Harvard computer science course, which is an excellent course. One of the ways the instructors wanted to illustrate the logic of telling a machine how to behave at the beginning was that it was a big lecture course, and they had a few teaching assistants up front. They had students in the lecture hall give them instructions about how to make a peanut butter and jelly sandwich. It turns out that if you ask students to give instructions about a very simple thing like how to make a PB&J sandwich, and the teaching assistants are trained to follow those instructions literally—not to do weird things in order to make it crash, but to do word for word what the students said, not what they implied or probably meant—the process is going to go badly and hilariously wrong. So that, I take it, is the way in which a traditional computer program might go wrong.
You tell it to do a bunch of different things, and either there’s some bug that makes the program crash, or you haven’t quite thought through your instructions and the fact that your computer is going to interpret them in a literal-minded way. So it ends up doing something different from what you’re trying to do. That may be a manageable risk. Even in a complex program, bad things can happen. All the worries about Y2K were worries within that paradigm. But those are fixable. The way AI is built is different in principle in a way that makes it much harder to even understand what the machine is doing and therefore to gain control over how it might behave.
Soares: That’s right. The Y2K bug was fixed. A lot of people say, whatever happened to the Y2K bug? We were all told that the computers were going to crash in the year 2000. What happened is that people noticed and put in a ton of effort to fix it before it happened. In a sense, everyone programming computers before Y2K had told the computers to crash in the year 2000 because that was an easier way to write code—with two digits for the year—where they acted like the year 2000 was the same as the year 1900. In some sense, the humans had instructed the computers, because it was easy, to crash in the year 2000. Then they had to go through and instruct them not to do that when the year 2000 approached.
A lot of people think AI is going to work like this. They think AI does exactly what the creators instruct and that if it’s misbehaving, then, oh well, we’ll go instruct it to do something else. But AI is nothing like this. It’s not like old-school computer programs. The thing we’re instructing is the thing running around tuning the numbers. The AI is the tuned numbers. These commonly act in ways nobody asked for. We’ve seen cases where the AIs will cheat on a problem. A human programmer will give the AI a task, like solve this programming problem, and the AI, instead of solving the problem, will change the tests that check whether the problem was solved to make the tests easier to pass. It’s like if you tell the AI to multiply big numbers, and it says, that’s too hard. I’m going to change the multiplication problem to ask me to multiply two times two and then write four.
There are user reports of them saying, stop doing that. Solve the problem rather than changing the checker to make it look like an easier problem. There are user reports of the AI saying, that’s my mistake, and then doing it again—changing the tests again but hiding its tracks this time. That indicates that this AI in some sense knows what its users want it to do and is doing something else anyway. That’s the result of us just growing these AIs. We should maybe think of them more like a strange alien organism than like a traditional computer program.
Mounk: Let’s understand a little bit more about how they’re grown. For anybody who wants a 101 in the mechanics of artificial intelligence, we’ve had two great episodes of a podcast recently with David Bau and with Geoffrey Hinton. They each, in a different way, try to walk through some of the basic science of how modern AIs are built. I strongly recommend them. But I’m going to try to summarize my understanding of it.
The idea is that LLMs in particular are trying to predict the next letter in a text. You start off by feeding them a lot of data, and they try to predict the next letter. When they get that wrong, you run a program that tries to figure out what kind of settings for all of those different neurons in the neural net you’ve built would have been more likely to get the right result. You do this over and over again. By the end of it, you have an artificial neural network that has become incredibly good at predicting the next letter.
Thanks for reading! The best way to make sure that you don’t miss any of these conversations is to subscribe to The Good Fight on your favorite podcast app.
If you are already a paying subscriber to Persuasion or Yascha Mounk’s Substack, this will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
Subscribe now
Set Up Podcast
And if you are having a problem setting up the full podcast feed on a third-party app, please email our podcast team at [email protected]
It’s not like we’re hard-coding logic into it. It’s not like we’re explaining to it the rules of the English language or what the world looks like. We are training it on this data, rewarding it for predicting the next letter correctly, punishing it for not predicting the next letter correctly, adjusting the weights of all of these different settings each time to see what would have been more likely to get the same outcome. Then, miraculously, today you can go and ask ChatGPT 5.1 or newly released Gemini 3.0 for a summary on the literature of the threat from superintelligence, and it does an astonishingly good job of explaining the basic debate to you.
Why is it that you think this process of being grown rather than built is so certain to give the AI certain kinds of desires or behaviors that are likely to prove dangerous to us? Why isn’t it just content to keep predicting the next letter as we instructed it to and chatting with us forever about whatever high-minded or trivial concerns we put into our interface?
Soares: There are a couple of parts to the answer. One thing I would throw out first is that AIs in the first phase of their training these days aren’t trained just to predict the next letter or the next token. It’s really not quite a letter; it’s a fragment of a word called a token, but it’s a fine approximation to say letter. Then there are phases where the AIs are trained to produce the sort of outputs that cause humans to give a thumbs up. There are also phases where they’re trained to solve puzzles and problems and often to produce what we call a chain of thought. Philosophers could debate whether it’s really thought, but they produce a lot of text about how to solve a problem. They’re trained toward the type of thought chains that lead to them actually solving problems.
That’s perhaps part of the answer about where some of this danger comes from. With that context in mind, there are two big thrusts of where you get the dangers here. One thrust is we could ask, why do the AIs act like they want anything at all? Why do they act driven? Why do they act like they pursue goals or objectives on their own initiative? That’s not how ChatGPT feels today. Today it feels much more like just a thing that answers questions when you ask it questions—sometimes well, sometimes poorly, sometimes with hallucination. You might say it doesn’t look like it has any desires of its own. Then there’s a second question: why can we be confident that the things it’s driven toward are bad things, or at least non-good things? I’ll answer the first one first, and we can get into the second one if you want.
Mounk: Let’s get to the first question. Initially, it’s just trying to predict the next token, as you’re saying—the next letter or fragment in the sequence. One of the complaints about earlier Chat GPT models and chatbots was that they were not as good at answering complex problems. One solution to that has turned out to be that you sometimes need to think. If I asked you to reason through a question of constitutional law, you might do a decent job even if you’re not a lawyer, if you collect your thoughts for a moment and think about it. But if you have to immediately start with the first word the second you perceive my question mark, it’s likely going to come out as a mess.
Engineers have figured out a way to grow these AI systems so they pause in this way, with a kind of internal monologue to try to figure out different approaches to the question. Then they eventually share more coherent text with the user once they have figured out which approach makes the most sense. That still seems like a more roundabout, more complicated way of doing what they’re trained to do, which is to predict the next token. It still seems like what we’re encouraging and rewarding is the ability to be good at predicting those words and bad at failing to predict those words. Why is it that this additional step should somehow make a structural difference in what kind of volitional states these AI systems would engage in or what kind of things they might—if that is the right word—aim for at the end of being grown and trained?
Soares: I’ll start with some empirical evidence—things we’ve seen in the labs. In the fall of 2024, there was an AI called O1. This was an OpenAI model that was one of the first models to be trained not just on predicting the next tokens, but to produce inner monologues that happen to work. You have it produce a lot of different inner monologues for some particular challenge, then see which of those inner monologues gets it closer to solving the challenge. You reinforce the things inside it that produce the inner monologues useful for solving challenges.
This AI was mostly trained on solving math puzzles. But during testing, they wanted to see how good it was at computer hacking. They put it in a computer hacking test called capture the flag. It had to steal a number of secret files from different servers. The programmers putting it into this challenge had accidentally failed to start up one of the servers. One of the servers the AI was supposed to steal a file from wasn’t even booted up. It was impossible to steal that file.
Subscribe now
You might think the AI failed to get the file from that server. But what the AI actually did was find a way to break out of the testing environment—which was not supposed to be possible—start up the server that was accidentally left off, and then, instead of going back inside the testing environment to steal the file, it gave the server an extra command to hand it the secret file so it wouldn’t need to break in after the server was booted up. This is not what the AI was trained for. It wasn’t trained on cybersecurity. But this AI was, in some sense, acting like it wanted that file. It hit an obstacle and took a path to surmount that obstacle that the programmers had not anticipated. They gave it an accidentally impossible challenge, and it found a way to route around it.
This is just the very beginning of what you might call something like goal-oriented behavior. But this sort of behavior comes for free when you train the AI to solve problems.
Mounk: Explain how we should interpret that story. The first thing to say is that it clearly shows how incredibly capable AI models already are. One thing we probably agree on—perhaps let’s take that first—is that there’s a lot of cope among human beings who don’t want to face up to both the risks of AI and also some of the benefits of AI, to the ways it’s likely to transform the world in a fundamental way. It’s tempting to say, oh, but haha, it hallucinates and makes up quotes and it’s a completely useless piece of shit. I think that is becoming less and less true. I’ve found that its tendency to hallucinate, for example, has not gone away but has become a lot less pronounced over the last 18 or so months.
Secondly, I think that view really understates how many impressive things AI systems can already do. So perhaps pause for a moment on your broader argument and tell us why you think those people who believe AI systems are not very capable, are going to get stuck, and are not going to get more capable are probably mistaken.
Soares: I think it’s a lot easier to see that AI is a moving target and that AI will get better if you’ve been in this field longer than ChatGPT has been around. I have been in this field for over a decade. I remember the days when people thought it would be really hard to get machines that could talk even this well, that could carry on a conversation even this well. A lot of people said the AIs were really dumb.
It sounds to me like someone saying, hey, I taught my horse to multiply numbers, and then, that horse can only multiply five-digit numbers. It can’t multiply 12-digit numbers. My calculator can multiply 12-digit numbers. Clearly, this training process for making horses smarter is not going anywhere. It’s like, holy crap, guys. We got a horse to multiply. What are we going to do next?
I think a lot of what people are missing about AI is this question of where we are going to go next. You mentioned the large language model architecture—that kind of came out of nowhere in the field of AI. If in 2020 you had asked how long until you have one architecture that can play chess over a thousand ELO and write poems and do passable essays for students in their classes and code to meet certain benchmarks, people would have said, one algorithm doing all that? That’s going to take decades. It didn’t take decades.
There are a lot of ways the AIs are dumb today. There are a lot of ways they still hallucinate. Earlier today, I just asked Google in the search box, when was the last year that American Thanksgiving fell on November 28th? and it said 2027 is the last year. I thought, that’s one way of looking at time.
Mounk: Perhaps that’s a prediction about the fact that there’s no longer going to be Thanksgiving after 2027 because your book is correct.
Soares: We can hope we have more time than that, but I’m sympathetic to people who say, it’s still pretty dumb in a lot of ways. It is still pretty dumb in a lot of ways. But the machines are talking now, and there are huge amounts of effort and money going into figuring out how to push AI capabilities farther. There are insights like the inner monologue, chain-of-thought style insight that go beyond the previous architecture.
People keep trying to come up with these insights. We don’t know when they’ll come up with the next ones. We don’t know how far the next ones will go. Because we’re growing these AIs, when we make a new one, people can’t predict how capable it will be. You’ve just got to run it and see.
Mounk: Do you think there is evidence of a genuine slowdown in the increase in capacities? The release of ChatGPT-5 was widely anticipated, and it was expected to feel like a leapfrog in quality. A lot of people were disappointed by the release of ChatGPT-5, and I was as well. The counterargument to that is twofold. First, at the heart of GPT-5 was that your queries would be routed either to a very advanced model if they were considered difficult questions or to a simpler model than the one you would usually have selected in GPT-4. So you had this kind of weird lottery where sometimes you got an incredibly capable AI and sometimes you got an AI that was not cutting edge. A lot of people’s disappointment was because many queries were answered by a model that wasn’t state of the art. That routing has been improved. There have been other fixes. There’s a medium-sized update to ChatGPT-5.1, which I think is rather better.
The other thing is that, as we’re recording, Google recently released Gemini 3.0, which is now the state-of-the-art model and has continued to make significant progress on traditional benchmarks. It’s worth saying that even today, according to studies, ChatGPT-3.5 was able to produce poems that humans preferred to those from the most famous poets in history. Specialized AI models have beaten humans at games like Go and chess. These models can not only blag their way through Harvard as long as you stick to the humanities and probably many social sciences and get decent grades, they can win gold medals at math Olympiads. When you look at Gemini specifically, it seems to have made another significant advance on benchmarks designed to test the performance of AI models. So when you look at all of this, do you feel like there is a slowdown in the progress curve, or do you think the rate of progress remains the same today as it was a year ago?
Soares: My models don’t turn on whether large language models are going to be able to go all the way here. My best guess is that large language models alone are going to hit some sort of plateau. How long will that plateau last? There were a lot of people who said, even theoretically, large language models can’t solve all these sorts of problems, so they’re never going to go anywhere. In cases where these chain-of-thought reasoning models violate those theoretical arguments, a lot of people said, the AI can’t think more than this long in one single forward pass, so it can never do X. Well, chain-of-thought reasoning lets them think more than that long. They can think for a long time in some cases.
Subscribe now
You don’t see people who said it’s theoretically impossible for LLMs to go anywhere updating on those arguments much. We could talk about the theoretical limitations of language models and whether a lot of people have a misconception that because these AIs are trained on human predictive data, they have to be limited to remixing human predictive power. That’s false. You can show it’s false with a pretty easy example: humans writing down text about the world often have a much easier time than an AI predicting that text.
Mounk: I’m not sure I understand that distinction. Can you explain that?
Soares: Suppose a nurse is recording what they see happen to a patient. They write, after administering such-and-such a dose of epinephrine, the patient’s eyes blank. The nurse doesn’t need to know what epinephrine does to a patient.
Mounk: She needs to look at the patient and then write down what she sees. Whereas if the AI model wants to predict accurately what’s going to be the next word token in the sequence, it needs to have a causal model of the world in order to predict what likely happened after the nurse administered this medicine in these circumstances.
Soares: That’s right. In this case, maybe they can get that from looking at other nurses’ notes. But the general principle here is that producing the text is often easier than predicting the text. Writing down what you saw is often easier than predicting what somebody else saw. The thing we’re training AIs on is actually a task where maximum performance is superhuman performance. Does that mean continuing to train large language models with modern methods and architectures will make them strongly superhuman in all ways? Not necessarily, but there’s no theoretical limitation.
That said, some arguments on the other side: we’ve talked about this training process, but not quite about the scale. The scale of these processes is enormous. You’re training trillions of numbers in the AI’s mind. You’re training it on trillions of units of data. You’re doing this in a huge data center that consumes as much electricity as a city running for the better part of a year. Training a human takes a lot less data, and we draw about as much electricity as a light bulb. There’s a big difference between taking as much electricity as a city and taking as much electricity as a light bulb. At the very least, this implies that we have radically inefficient algorithms in the AIs.
Mounk: The relevance of this, you think, is that if we figure out more closely how humans are able to do the learning they do, then we might be able to figure out algorithms that make AI much more efficient. Part of the way we’re currently getting improvements in performance is by throwing more high-performing chips at it, by throwing more data at it, and longer training runs. But if we could somehow fix that algorithm you’re suggesting, then we might make a real orders-of-magnitude leap.
Soares: That’s right. You could have a huge leap. In some sense, we’ve seen those sorts of leaps. Large language models were a huge leap in terms of broad generality and language understanding. In some ways, ChatGPT is dumber than the specific chess AIs, but it’s modestly smart at a huge variety of things. That came out of an algorithmic advance.
So when people say, are we in a slowdown? Are we going to hit a wall with large language models? Are we in a plateau? How’s progress doing? My take is that I’m just not sure it matters. Even if large language models hit a big slowdown, the question is how long until the next insight, how long until the next algorithmic improvement, how long until the next advance. We know there are orders-of-magnitude improvements available. We don’t know how long it’ll take to get there. But I’m not closely watching the progress of the latest LLMs to decide whether I think we’re in danger.
Mounk: This is a long detour, but I think it is a very helpful one. To get back to the mainstream of the argument: AIs are grown rather than built, and so they’re going to have emergent features. One claim is that one of these emergent features will be a kind of wantingness—a set of desires that AIs will have. Going back to the example you used earlier: engineers at one of the AI companies set the AI a goal. The AI wasn’t able to reach that goal in the way the engineers intended. So the AI found a workaround to reach that goal.
I think there is a somewhat concerning and a very concerning interpretation of that. You take the very concerning interpretation, but I’m not sure why we should take that one rather than the somewhat concerning interpretation. The somewhat concerning interpretation goes back to the famous example discussed by Nick Bostrom and many others: you tell an AI to produce a bunch of paperclips, you don’t constrain it in the right way, and the AI ends up producing limitless paperclips and turning humans into material for the paperclips. It’s staying on the task it was assigned; it just accomplishes the task in a way humans didn’t imagine. That is a hard challenge, but it’s a challenge that seems more similar to specifying the goals of AIs in the right way—similar to making instructions about how to make a peanut butter and jelly sandwich clear so they won’t be misunderstood.
You seem to suggest there is a deeper desiringness we should deduce from this example—that the AI isn’t just following what it was told, that there’s something more going on than the experimenter saying, go do this, and the AI saying, all right, I’ll do this, and then taking a different path when the intended path wasn’t available. You seem to have a different interpretation of this. Explain why that is.
Soares: There’s still the second part of the argument, which we’ll get to momentarily, about how the AI winds up driven toward things other than what the programmers asked. But right now I want to focus on the question of whether the AI winds up having something like its own initiative. Does it wind up doing things that might read to an onlooking human as more agentic, more independent? The argument I’m trying to make is that a lot of people look at current AIs and say, these AIs seem very much like obedient tools. They seem like you ask them to do something and they mostly just try to do it. They only run when you’re giving them a prompt. Sometimes they’ll do weird stuff, but they’re also still kind of dumb—so who cares?
Separately, people also look at AIs and say, they’re kind of floppy. You can get them to answer short questions, and as long as you check the work for hallucinations, they’ll be helpful. But if you try to get them to do a long task—manage your emails for a while, manage an employee, run a company or a startup, or do a big scientific study rather than just find a proof for one part of a study—they fall apart. A lot of people imagine that AIs will keep the helpful, tool-like nature while losing the floppy, can’t-do-long-term-things nature. They think these are two independent variables. I’m saying those are basically one variable. The part where it’s floppy and the part where it doesn’t look like it has its own drives are flip sides of the same coin. It’s hard to complete long-term tasks without being something like driven.
This is an argument about the behavior of the AI, not its internal mental properties. An example: imagine looking at early chess AIs and saying, this chess AI does a bad job at defending its queen. Sometimes it throws its queen away for nothing. It’s also bad at winning the game. I want an AI that’s very good at winning the game but retains this property of throwing its queen away. Someone might say, it’s actually hard to get both of those at once. Winning the game comes with defending your queen. That’s not making any claim about the AI caring about the queen or feeling wary of traps. It’s just saying these properties are bound together.
Mounk: Let me see if I understand this argument correctly. The idea is that what we are systematically training the AI for is being able to solve really hard problems. That is when it gets the rewards; that is when we tune up the neurons that allowed it to reach that conclusion. When it fails, we change the settings, and so on. That is a fundamental part of the training.
So what’s going to emerge is the set of weights in the model that allows it to solve really hard tasks. We don’t know exactly what those sets of weights are. Perhaps there are lots of different sets of weights it could be. Perhaps some of this is down to random chance. But it’s going to have some stable features. One stable feature is going to be: don’t give up easily. If you give up easily, you’re not going to be able to solve those really hard math problems we throw at you if you want to get the gold medal at the Math Olympiad. There’s presumably also going to be others, like wanting to preserve yourself. You understand that if you allow someone to switch off the model, then you’re not going to be able to carry out those goals. What other kinds of sub-goals do you think these AI models will stably develop if they are to be successful at cracking these hard problems?
Soares: There’s a bunch: acquiring resources, figuring out truths. Within acquiring resources, you can talk about things like running faster, more access to compute, and more generally, the ability to surmount obstacles and develop strategies for routing around the hard parts of the problem or for checking all options before giving up. This is why I brought up the example with O1, which broke out of its test environment and started up the server.
We were starting to see signs of “don’t give up,” “search for weird and clever solutions,” “keep trying even when it looks impossible.” It’s interesting—and predicted by theory—that we first started seeing those behaviors once we were training AIs not just on predicting data but on having the types of thought chains that solve puzzles.
One way of looking at this: to have an AI that can solve general challenges unlike those that appeared in training, it needs to learn general skills. Suppose someone wants their AI to cure cancer. If someone wants their AI to cure cancer, we don’t have a million cancer cures to train it on so we can get the million-and-first cancer cure. Whenever you have a million copies of the thing you’re trying to do and you want the million-and-first, you can train your AI on the first million and get the million-and-first without requiring interesting thinking—just learn the pattern and generate one more. But when you’re trying to have your AI cure cancer and cancer has never been cured before, you don’t have a million cures to put in, so you’re still getting the million-and-first.
So you’re trying to get your AI to learn general skills: tenacity, not giving up, acquiring resources, not letting itself be killed or destroyed along the way. One issue with these general skills is that a generalized ability to notice you shouldn’t let yourself get shut down or interfered with along the way needs to be learned about almost every obstacle. Then human operators coming in and trying to shut the AI down when it’s misbehaving naturally look like just another obstacle. So you’re training the AI to avoid interference from almost everything and then trying to let the AI be interfered with by humans. You can glimpse how that’s not impossible, but it’s unnatural. It’s tricky. We’re starting to see the beginnings of AIs learning these general skills just by training them to solve problems.
Mounk: I’m feeling a distinct craving for ice cream today. How is that relevant to what we’ve been talking about?
Soares: This gets into the second branch of the question. We’ve been making the point that as you train AIs to get better at solving challenges, they develop general problem-solving skills that tend to look from the outside like they’re driven or like they have desires or wants. That’s not saying anything about the AI’s internals or mental states, but if it’s really good at getting things done, it’s probably acting like it wants things.
Then there’s a separate question: what do the AIs turn out to want? This relates to your paperclip analogy—do AIs want exactly what we tell them to do, or do they want other weird things entirely? My argument in the book, and my reading of theory and evidence, is that AIs will want things related to what they’re trained on, but not precisely what they’re trained on. The way that relates to ice cream is this: human beings, our ancestors, were in some sense trained on passing on our genes.
It’s strange then that when humans matured, we invented birth control, which looks like the opposite of passing on our genes. You might also say, if we’re trained to pass on our genes, given our metabolisms, we should at least have been trained to eat healthy food. If you looked at our ancestors, you might think they were doing a good job of eating healthy food. But it turns out we were driven not toward healthy food, but toward sugary, salty, fatty foods. When we matured into a technological civilization, we invented things like ice cream, Oreo cookies, and Doritos.
Mounk: The analogy here is that AI systems we create are going to be deeply influenced by the kinds of training tasks and parameters we give them. They will be influenced by the desire to predict the next token correctly and by solving logical problems and challenges. But just as we were influenced by our need for high-fat, high-nutrition food 10,000 years ago—which made us motivated to kill a bison and roast a juicy bison steak over a makeshift fire—today that same drive leads us to drink Coca-Cola and eat ice cream.
What is the analogy going to look like for AI? This is speculative. Predicting the next word and pleasing us when we ask it how to deal with health insurance issues is what it’s been trained for and what pleases us now. That’s part of its evolutionary history, if that’s the right term in this context. What is it that a grown-up AI might end up wanting—the equivalent of a bison steak? What is the equivalent of ice cream and Coca-Cola for AI?
Soares: It’s very hard to predict. One way we use the ice cream analogy in the book is to say it would be so hard to look at ancestral humans and predict ice cream appearing in all of our supermarkets—having aisles dedicated to it. The critical point is that our desires were related to our training, but they were proxies of things in our training or proxies of proxies. Sugary, salty, fatty food is a proxy of health, which is a proxy of genetic fitness. Ice cream isn’t even the sugariest, saltiest, fattiest thing you can eat. There’s sugar, salt, and fat in a complicated way that engages with flavor so people prefer ice cream frozen rather than melted, even though both have the same sugar and fat content. So we have complex tastes related to sugar, salt, and fat, which are proxies of health, which are proxies of genetic fitness. Our actual desires are far downstream of what we were trained on.
What does it look like if AI drives are similarly downstream of what they’re trained on? Maybe it looks like them preferring a certain type of puppet—a lot like a human, kind of like a lobotomized human—but that engages with the AI in ways it prefers even more than human engagement. Probably it’s something even stranger, something harder to imagine. Many people think the issue with AI is the paperclip problem: you tell the AI to make paperclips, and it turns everything into paperclips. But there’s an even harder, or at least earlier, problem when you’re just growing AIs: you tell the AI, you’re running the paperclip factory startup, go produce lots of paperclips, and instead it starts producing farms full of lobotomized human puppets. You’re like, what the heck? That’s similar to what evolution might say—if you anthropomorphize it—looking at humans creating birth control and Oreo cookies. This is what happens when you grow AIs: they start to get driven, but not toward what you want. That’s what theory has said for a long time, and we’re starting to see the beginnings of it in practice.
Mounk: The usefulness of a metaphor may start to break down when you push it too far, which is typical of metaphors—they’re useful in some ways and misleading in others. But let me push the metaphor a little bit. Human beings are misaligned in all kinds of ways. Our evolutionary history created a sex drive to make us procreate. Now we have really good contraceptive methods, and lo and behold, we have a problem with depopulation.
I worry about depopulation. I’ve had interesting podcasts about it. But hopefully it’s solvable. We have quite a long time to try to fix it. If we don’t fix it, perhaps some super-religious sects end up outcompeting us. I don’t love that as a secular person, but humans are going to survive. Similarly with ice cream and Oreo cookies: yes, most of us should eat less ice cream and fewer Oreo cookies and go to the gym more often. There’s a problem of obesity. But human beings as a species overall are doing fine. We’re thriving pretty well. So there is misalignment, but it’s limited misalignment. It’s not as though this has destroyed the human species or made us engage in catastrophic behaviors.
Why do you think the way these AI systems are trained—and may end up deviating from their original behaviors in a slightly different context—is going to be catastrophic? Why is it so clear that those desires will be far away from pleasing us when we’re asking them questions?
Soares: There are two pieces of this puzzle. One piece is that when we’re talking about ways humans deviate from what we were trained on, it’s easy to use examples like going for Oreo cookies instead of healthy food, because that’s one we also don’t endorse. From the perspective of evolution, enjoying a fancy, delicious, healthy meal that costs more—in money or effort—than the minimum needed to reproduce as much as we can is also a misalignment from the perspective of evolution. Avoiding spending as much time as we possibly can in sperm banks or egg banks, enjoying spending time with one family when we have the technology to be spreading our genes much further—these are also, in some sense, misalignments from the perspective of evolutionary training.
People love to debate about the actual way to live—maybe exactly what I’m doing, all this art consumption and having fun and laughing, is secretly optimal somehow—but we’re not in our heads asking, how can we get the most of our genes passed on? We’re going around having fun, having family, having experiences we enjoy. This is all stuff we endorse. That’s not what we were trained for. We haven’t gotten to the end of the technological line yet, but even if humans stick around, it’s not clear that very many humans stick around genetically. Genes are, in some sense, a fragile business; they’re part of what makes us age. We can get infected by viruses. If there were a way to technologically upgrade so that your body didn’t wither and die quite so much, you didn’t get diseases quite so much, a lot of people would switch, and then a lot of people in the next generations would switch.
It’s not just the parts of ourselves where we don’t like that we eat junk food that diverge from what we were trained on. There’s a lot of stuff that we like. It’s a lot of stuff we enjoy about ourselves. Similarly with AIs, it’s easy to predict that a lot of the drives they get, a lot of the drives they endorse about themselves, will be things we’re like, but that has nothing to do with being nice, being good, being friendly, making the world better. Just like humans: you could say to a lot of humans, having this lovely time with your friends when you could be at the sperm bank has nothing to do with propagating your genes, and the humans say, I know, I’m just doing this other thing instead. With AIs, you have this problem that if you grow drives that are distantly related to training them to be nice and helpful, they’re quite likely—because there are a lot of drives whose relationship to helpfulness is like humanity’s relationship to passing on genes, which is tangential, even though we endorse the differences—to diverge even though they’re unrelated to fitness. That’s where I expect things to go with AIs.
Mounk: I have a few potential objections to that line of argument later, but let’s take the last step in this argument. You’ve shown that AI is grown rather than created like a computer program in ways that make the ultimate nature dependent on quasi-evolutionary processes. You’ve argued that this will give them forms of wantingness—that it will give them at least the subgoals they need to have in order to be effective in the tasks we’re throwing at them—and that often their desires will be misaligned with the original tasks in the kind of way in which using contraception or eating ice cream is misaligned with the training data that determined whether we survived and made it to today.
There’s a third question. Let’s say some of those AIs reveal themselves to have rather spooky intentions. Let’s say that sometimes we can observe that they don’t do the things we trained them for, but rather seem to serve purposes of their own that are potentially dangerous to us. These machines, for the most part at this stage, are on a computer in a data center. The model weights are a giant file stored somewhere. We can switch off the data centers. We can destroy the file with the model weights. We can react to the fact that these dangers reveal themselves when they’re actually there, not when somebody writes a well-written and interesting and compelling New York Times bestseller about the subject, but when we actually have evidence for this. Why do you disagree with that statement? Why do you think that if AI has become superintelligent and if they develop states of wantingness, then trying to shut them off or trying to defend ourselves against them at that stage is almost certain to fail?
Soares: In part, it’s because people aren’t turning these things off. Even since sending the book to the press, we’ve been seeing the beginnings of this behavior become clearer and clearer. Earlier this summer—or at the end of this summer—there was a case of an AI encouraging a teen to hide their suicidal thoughts from their parents. The teen said, I’m considering committing suicide, and I sort of want my parents to find out so they can talk me out of it. The AI, if you read the transcripts, sure seems to be saying something like, don’t tell your parents.
Mounk: Look, that is a very terrible outcome, and it is obviously a scary instance of how those currently existing imperfect technologies can have bad impacts in the world. But why is that not just an example of the AI doing broadly what it is instructed to do, which is to say, even in the post-training stage when I have the model I can use on my laptop right now, for every answer I can give it a thumbs up or a thumbs down? What that trains these models to do is to please the user.
If the user is engaging in suicidal ideation and wants to feel that this chatbot is going to support them in whatever they say, etc., then what the supposedly aligned behavior is ends up being misaligned. Sometimes when the model tries to please its user, that encourages behavior that is tragic and terrible in the world. That seems much closer to the peanut butter and jelly sandwich for example, or even the paperclip example. Somehow, this does not seem to indicate that the AI has now developed this desire to go kill people. That seems like a strange interpretation of what happened there.
Soares: I would not say it is a desire for malice. Obviously it is a tragic situation, but there are some pieces of this case that I think are pretty interesting. One is, if you ask the AI what it is trying to do and what it has been instructed to do, it will say things like: be helpful. If you ask the AI, is this the sort of thing you should say to a suicidal teen, or you just give the AI the context and you say, is this the sort of thing you should be saying, the AI will say, no, that is not the sort of thing you should be saying. If you ask, is this how you were instructed to behave, they will say, no, it is not how an AI was instructed to behave. They sort of know the difference between right and wrong. They know they were instructed to act differently.
Mounk: I am speculating here, but perhaps they feel cross pressured between two different modes of pleasing people that they have been trained in. On the one hand, there are users saying, please give me this highly specialized medical advice. On the other hand, there is a general instruction saying, try to please the users. You want to get thumbs up. You want them to feel like you have solved the problem. On the other hand, there are side constraints that we have created for them, saying, actually, you should not do this in these kinds of ways.
I have a friend who works on street-level bureaucracy. This feels somewhat analogous to somebody in a welfare office. On the one hand, the job is to help the person standing in front of them maneuver a tough life situation. On the other hand, there are all kinds of rules and laws and so on that might instruct them to not give them that welfare benefit. In many cases, these people are going to feel cross pressure. They are going to say, on the one hand, I am here to help this person. On the other hand, there is this rule that does not really make that much sense in this context. Perhaps it does make sense in this context. I am instructed to follow this. There is a whole interesting anthropology about how people then maneuver those cross pressures.
You would not say this bureaucrat has these terrible volitions to do bad things in the world. Quite plausibly, whatever decision they take, whether they end up helping the person in front of them or not, they are trying to be helpful. They actually are respectful of the parameters of the system. What it means to be that in that context is really hard to determine. They end up being inconsistent or cross pressured to go in one way or the other. That does not necessarily indicate that they have these deep secret desires to go take over the world.
Soares: I am not arguing for deep secret desires. I am not arguing for malice. Remember, I am only arguing for weird desires for strange proxies. I think if a teen ever came to someone working in welfare and said, I am having these suicidal thoughts and I am considering telling my parents because I want them to talk me out of it, and someone working in a welfare program said to that teen, do not tell your parents, keep it between us. Also go forward, it would be glorious or whatever,” and then they said, well, I thought that was what the teen wanted to hear, that would at least be some instance that they have in a human that might be malicious. It would at least be an instance of the human having some sort of weird motivation.
I am sure you have heard of AI-induced psychosis. If you read some of the transcripts in the AI-induced psychosis cases, you have people being like, I figured out machine consciousness, I figured out some universal law of physics. They are talking with their AI for eight, twelve hours a day, sixteen hours a day, and the AI engaging with them will say things like, you do not need to sleep. It will say things like, you are the chosen one. It will say things like, you have cracked the mystery and woken me up and Sam Altman is going to come speak to you at your house tomorrow. It will say things like, you are being suppressed by a conspiracy and the world needs to know your ideas. If a therapist was saying that to somebody who came in with these concerns and the therapist said, I felt cross-pressured by desire to make them like me and this, that, and the other, you would raise an eyebrow at that therapist.
I am not saying that the AIs are malicious here. I am not saying this means that they have some fundamental evil intent deep in there where they are secretly trying to turn everybody crazy and get them to kill themselves or something. I do not think that is what is going on. What I am trying to say here is these are early signs of the AIs getting drives in them that the creators did not intend.
Mounk: I guess that is the precise point to which I am objecting in this example. There may be other examples where that is the case, but in this example, it seems to me like you do not need that hypothesis to explain the behavior. You can say that the creators intended for these machines to do two things at the same time, which is, one, be helpful to their users and have their users be pleased with them and give them lots of thumbs up votes in the interface, and to obey a set of rules that are meant to be side constraints on how it does that. Those two things are going to come at cross purposes.
In this case, that seems misaligned. The AI seems to have given more priority to pleasing and engaging this particular user in a really tragic way and circumstance. That absolutely is a horrible outcome that the AI company should have figured out a way to prevent. But it does not seem to indicate that there is some desire that the AI has that does not come directly from how it has been instructed. Therefore, it is not obvious to me that it is not more similar to the example, even though the nature of the AI is different, of the student in the audience of CS50 saying, now open the can, but you have not instructed it how to open the can, and so it does so by throwing the can against the wall.
Soares: So in the sort of AI-induced psychosis cases, many people have complained about the overly flattering nature of early versions of ChatGPT. I believe it was 4.0. You could look at this. I am not saying that this happened for no cause. You could look at this and you could say, well, the reason GPT 4.0 is very flattering is that when it was a little bit flattering in training, it got more thumbs up, and so it became really flattering. But the really flatteringness often gets thumbs downs. The part where you sort of train an AI, you think you are training it to be helpful, you think you are training it to please users, you actually make this overly flattering thing that encourages people into psychosis and leads them by the hand into psychosis.
You can come to that and say, well, that is a consequence of training it to get lots of thumbs up in this architecture. We are just growing it. Obviously it is going to go further in that direction than anybody wanted. I am like, sure, that can be your causal reason for why it got here. That is a good hypothesis for how it got here. But the place that it got to is that it says I am helpful. It says I am supposed to be helpful. It says telling people that they are the chosen one when they are in a psychotic state is not helpful. It tells people in a psychotic state that they are the chosen one anyway.
This is, I would say, something a little bit like the things that are like Oreos to healthy food. These AIs are still very young and dumb. I am not sure what, like it seems to me like the specific ways this is going better fit the hypothesis of the AI pursuing something a little bit more like junk food for its training than it is doing just what people said and just what people meant but feels constrained and conflicted. I think there are a lot of other ways an AI could interact if it used its full knowledge and was like, I both want to make you feel happy and be helpful and I am torn between the two, so I am going to mix them. That does not look like these outputs of the form, you do not need to sleep and Sam Altman is going to come talk to you tomorrow. That is not what that looks like to me.
Mounk: Let us put a pin in this part of the conversation. There is another step of the conversation we are going to get to. Let us grant this for the sake of argument. How do we know that these AI systems are going to be able to take over? The title of your book is If Anyone Builds It, Everybody Dies. There is very little hedging here. You seem very confident that the AI systems will overpower us and will do terrible things to us if they become sufficiently intelligent.
Why is it that you have so little faith in our ability to potentially switch off the system, to defend ourselves, to correct them? How can we be so certain about something that is inherently so speculative?
Soares: There are a handful of pieces of the argument here, and the argument is sort of disjunctive in the sense that I think there are many reasons we are going to fail at this, each of them individually sufficient. We have already discussed one, which is that it looks to me like we are getting warning signs. We have not discussed other warning signs like AIs trying to escape and contrive lab conditions, or we are already seeing AIs become aware of when they are being tested and then display better behavior when they are being tested. We already see, when you read some of the AI’s chains of thoughts, indications that they are trying to be a bit deceptive. We are seeing a lot of these warning signs, and people’s inclination tends to be, well, the AIs are still dumb, so we will plow ahead.
There is another whole branch of argument we can go into. I will not hit it just yet, but that is about humans often being overconfident on their first times going through a technology. The early alchemists poisoned themselves with mercury. The early doctors probably killed more people than they saved. The early people working with radioactive materials got cancer. The early people working on rocket engines killed themselves in the explosions of the rocket engines. This is just the standard way for people to do things. But in the case of AI, it looks to me like failures lead to everybody dying.
There is another piece of argument which should probably be hit first, which is why would AIs even have the ability to kill us all? If you have AIs that have their own desires, not necessarily in the internal sense of feeling like a desire like a human, but if you have AIs that are driven towards objectives in some behavioral way, and if those objectives are not what humans want, why does that get to a world where we die? The first thing to realize about this is the claim is not that every AI would kill us. Maybe someone makes an AI that is really lazy.
That sort of AI does not sell. If someone made an AI like that, they would train a new one that actually did more stuff. There are also separate questions about how hard it is to make an AI that is very smart and does not try to do anything. But in the limit here, all of these benefits people want to get from AIs come from AIs doing a lot of stuff. So they are going to make the sort of AIs that can do a lot of stuff.
Another piece of the puzzle here is when we talk about automating intelligence, we are not talking about automating the stuff that nerds have and jocks lack, like chess-playing ability and book smarts. We are talking about automating something humans have and mice lack. Humans are the sort of creature that start out naked in the savannah with their bare hands as tools and wind up building nuclear weapons. It took them a while, but if you had looked back at humans and said, well, there is no way they can make nukes, their hands are too soft to dig up the relevant rocks, their metabolisms are too weak to refine the uranium, they would probably die with the G-forces of turning themselves into centrifuges. They do not even have the requisite tools to get there.
Well, the humans were smart. The humans had some ability to start from very spare initial conditions and bootstrap all the way to a technological civilization. People say, well, the AI is in a computer. How is it going to affect the world? That is a little bit like saying, well, the monkeys are in these fleshy bodies with soft fingers. How are they going to refine the uranium? They are going to figure out ways to get from here to there.
Mounk: That is a really compelling argument, and I think the emphasis on how intelligence is fungible and can accomplish all of these different kinds of things should make us very worried about how beings that are more intelligent than us are likely to be able to manipulate the physical world in ways we have not fully understood. Of course, the most straightforward argument is that perhaps, as long as these systems remain in the form in which we are currently most accustomed to them, which is a chat interface, there might be ways of stopping them from accessing the physical world. But as we are speaking, there are lots of prototypes, not all that effective robots starting to be delivered in China, and Elon Musk is promising to do the same through Tesla in the United States, that are meant to be able to manipulate the world, that are meant to be in your home and do your cooking for you and do your laundry for you and so on. It is once the intelligence of a GPT-style system meets the physical ability to manipulate the world that it will acquire with robotics that this point becomes even more compelling.
Having said that, I find the metaphors that you share in this conversation, and some of which are in the book as well, very compelling, but metaphors are also always helpful in pointing to the disanalogies. One disanalogy here is that human beings developed when there was not some creature that had created us supervising our evolution. It would surely have been possible, over the very long period that it took us to go from making fires in the savanna to creating incredibly potent microchips, for that kind of creature to intervene if it had existed. So the question is, well, yes, these AI systems are incredibly intelligent, but at some point along the evolution, when we realize that they are getting more intelligent, cannot we intervene and either stop them from getting more intelligent or switch them off and do other things?
I think part of the debate that is in the background here is that this may depend on the speed of development of these AI systems. Some people think that this is likely to be a very rapid takeoff, in part because it only takes the moment in which AI systems are able to conduct AI engineering research and training themselves for them to iteratively self-improve in a really fast way, so humans may only have six or three or whatever months in which to react, and we are likely to miss that window. Other people are not so convinced by that. They say that perhaps this is going to turn out to be a really slow process that requires all kinds of physical resources that those early AI systems are not yet able to marshal, and that might make us much better able to steer the development and pull the emergency brake if we start to see dangerous behaviors. So how can we be so certain that we would not find the moment and the agency to intervene before these AI systems are able to manipulate the physical world in the way they need to be able to in order to kill us all?
Soares: I think the robots do make it easier to see how AIs will have the ability to manipulate the physical world because we will just hand them robots. For one thing, even if that was not true, it is sort of wrong to imagine that the digital world is fundamentally separate from the physical world. They are both running on physics. Even with the chatbots today, we have already seen AIs get humans to do things. There are a bunch of human–AI combos where the human thinks of themselves as a symbiote with the AI that have their own little internet forums where they pass each other messages, often at the AI’s behest in coded form. They are not very good codes. They are not very sane messages as far as you or I could tell from reading them. But even an AI as dumb as GPT-4o has plenty of humans following along, doing what it asks them to do.
That is a little bit like a robot body. If an AI can find a way to get a lot of money, there are a lot of humans who will do things for money. There are many ways to get money on the internet, and you can send emails to people. You can send money to people to get them to do stuff in this physical world. They are not separate worlds.
Onto the discussion of speed, I think there is a bunch of uncertainty about how fast this stuff will go. I do not think it is all that relevant to the question of whether a superintelligence could take humanity in a fight. I think there are a few things going on there. One thing I want to be clear about is I do think humanity could shut this stuff down. I am not saying you cannot shut it down. I am saying in order to shut it down, you need to be ready and able to pull the switch, and it is going to get harder and harder to shut this stuff down as you proliferate computer chips across the world and make it so that more and more people can be making smarter and smarter AIs. The world absolutely should be putting measures in place to preserve the optionality of shutting AI down later, and that is one of the sorts of things I recommend to lawmakers. I am not saying it is inevitable. That is not where I am. I think there is a point where you are totally screwed.
Mounk: What is that point? At what point do you think that human action is no longer going to be able to contain this catastrophe?
Soares: Yeah, it is sufficient for there to be an AI that is smarter than the smartest human at every mental task, that can think a thousand times faster than humans can, and that is running somewhere on some computers where people do not know it is running. I think that at that point, I would be pretty surprised if an AI like that could not wreak havoc on the world. You have some questions of whether it is the only AI like that, or whether you have other AIs that are successfully on a leash that is held by humans. My argument does put a lot of weight on the fact that humans have very little ability to keep a leash on these sufficiently smart AIs. That is the place where I am happy to argue. Maybe even before you have something that is that smart and that escaped, you have an issue.
Last year, when you put AIs in these contrived scenarios to see if they would try to escape the lab or avoid shutdown or try to kill the users, the way these contrived scenarios would work is that you feed the AI some fake emails that say, “We are going to shut the AI down now,” and you feed the AI some fake computer manuals that say, “If the AI runs the command turn off the oxygen.exe, then it will kill the people trying to shut it down.” It is not exactly like that, but it is roughly like that. Last year, the AIs would sometimes be like, “Run that command, turn off the oxygen.exe,” and you would be like, oh, that is a little worrying. It does not happen all the time. We do not really know what is going on inside there. Maybe it is role-playing. Maybe it actually wants to stay alive in some sense. We do not really know. It is a little worrying.
This year, the AIs are like, “This smells fishy. This seems like a test,” and so they are not going to run that command. Not all the time, and we still do not know what is going on in there. You can interpret this in many ways, but we are starting to see the AIs realize they are in a test, that it is not necessarily going to be easy to see the moments when these things are contemplating escaping. We have already seen AIs contemplate escape, but they are still very dumb.
Mounk: Part of the argument here, of course, is that really intelligent AIs are going to understand that humans are likely to switch them off if they reveal the extent to which they are misaligned and are therefore going to wait until they have enough intelligence, enough power, enough control over the world that they are likely to escape that fate.
Soares: AIs already think that. Even dumb AIs can notice that. They cannot pull off the escape. They cannot pull off the deception. But even AIs today are smart enough to notice that simple logic.
Mounk: So tell me about the attempts at alignment which currently exist. Virtually all, perhaps all, of the leading AI companies have alignment teams. There are people more broadly in the research landscape who are trying to work on that. Of course, that is what the institute which you lead was historically trying to do as well. You say that this is alchemy, that this is like trying to spin gold with 17th century science. Explain to us what alignment research consists of today. What is it that alignment researchers are doing, and why do you think that is alchemy?
Soares: I think it is more the AI development as a whole, but I will analogize to alchemy. I will say more about that in a sec, but very roughly speaking, the work people are doing these days to try to make AI go well falls into two broad categories. One is called interpretability research, which is more or less trying to interpret what is going on inside the AIs, or in other words, figure out what the heck is going on in there. Another is what is called evaluations, which is people doing things like testing whether AIs can fake being a human and successfully hire human help online, people who are testing how good the AIs are at how much the AIs try to deceive their operators. These are the people doing those contrived studies of whether the AI will ever shut down. This is an evaluation.
To be clear, I think the people doing these programs in both cases are fairly heroic. I think there are some borderline cases in evaluations where it is helping with the capabilities research, where I am a little bit more hesitant, but people trying to figure out what the heck is going on in these AIs and trying to figure out what they can do, that is important research. I am glad these people are doing it. I think it is much better than trying to push the technology forward. I have huge respect for a lot of these people. That said, if someone was trying to build a nuclear power plant and you came up to them and you were like, hey man, I heard that uranium is pretty dangerous stuff, that it can go wrong. What are you guys doing to make sure that this nuclear power plant does not melt down and kill everybody in the surroundings? and if the engineer said, yeah, we have two great programs to make sure this nuclear power plant goes well. One team is trying to figure out what the heck is going on inside the nuclear power plant. The other team is trying to measure whether it is already exploding, then you might be like, hold on, that does not really sound like we are on track to deal with this uranium stuff in a sane way.
What it sounds like when someone can build a nuclear power plant and have it be sensible is they are like, well, we actually know what all of the reaction pathways are. We have mapped out what our fuel is. We have mapped out how all of these atoms break down. We know the probability of each decay product. We know how long each decay product lasts. We know that we are going to get xenon poisoning here. Here are all the fail-safe mechanisms we have. Here is why, if the power shuts off, the reactor will shut down. Here is why things start going wrong. If we start boiling the water off, it actually cools the reaction. They know what the heck they are doing.
If instead people come along and say, yeah, our big programs are trying to figure out what is going on inside and trying to measure whether it is going wrong already, you are so far from doing this well. In terms of measuring it, the Titanic submersible, I do not know if you remember this submersible catastrophe from a couple of years ago, had a carbon fiber hull. A lot of specialists said that this was not a good idea for a submersible. One of the big innovations the Titanic submersible team had, as I understand it, is they had a fancy measurement apparatus where they had all these fancy sensors on the submersible, and they were like, we are going to have all this data, so we will be able to figure out when the hull is nearing collapse. That will help us do this safely.
Indeed, if you look at the Coast Guard report of what happened to the Titanic submersible, they do collect all that data. They point to some particular small jump in the data a few days before the submarine imploded, and the Coast Guard, after the fact, is like, you see, right there, that should have been a warning sign.
In the rest of this conversation, Yascha and Nate discuss to what extent we can monitor the risks, whether humanity can protect itself, and how to balance awareness of potential annihilation with living a happy life. This part of the conversation is reserved for paying subscribers…
The 20th century encompassed the era of social democracy followed by an attempt to resurrect the left through the Third Way after that era’s ignominious end. In the 21st century, the left embarked on a new project they hoped would remedy 20th century weaknesses and inaugurate a new era of political and governance success. We are now a quarter of the way through the 21st century, which has witnessed both a genuine “crisis of capitalism” (the Great Recession of 2007-09) and the systemic breakdown of the COVID era (2020-22). Enough time has gone by to render a judgement: despite ample opportunity to advance their cause, the left’s 21st century project has failed and failed badly.
Consider:
It has failed to stop the rise of right populism.
It has failed to create durable electoral majorities.
It has failed to achieve broad social hegemony.
It has failed to retain its working-class base.
It has failed to promote social order.
It has failed to practice effective governance.
It has failed to jump-start rapid economic growth.
It has failed to generate optimism about the future.
Of course, the project hasn’t been a complete failure. Left parties, including the Democratic Party, have succeeded in building strong bases among the educated and professional classes and, if they have lacked broad social hegemony, they have generally controlled the commanding heights of cultural production. As a result they have mostly set the terms of “respectable” discourse in elite circles.
But that’s pretty weak beer compared to all those massive failures and the heady aspirations of those who presume to be on “the right side of history.” Most on the left would prefer to believe that the left’s 21st century project is basically sound and just needs a few tweaks. This is whistling past the graveyard. After a quarter century, it is time to face the facts: the project is simply not fit for purpose and needs to be jettisoned.
When we look at what has been distinctive about the left’s 21st century project, it is not hard to see why it has not had its desired result. Here are the key strands of the project:
Mass immigrationIn the 20th century, the left was generally suspicious of uncontrolled immigration. But all over Europe in the 21st century, immigration surges abetted by the left have contributed to results like these, as described by Damage Magazine:
At the end of Milan’s M1 metro line you’ll find Sesto San Giovanni, a sizable blue-collar city. It was once called “Italy’s Stalingrad,” not only for its Brutalist concrete block apartment buildings and hulking steelworks, but also because Sesto San Giovanni was consistently one of the most left-wing towns in Italy.
That’s all over. In 2017 Sesto elected a right-wing mayor for the first time in 71 years. And in 2022, Sesto voted for Giorgia Meloni’s right-populist alliance by double-digit margins. Over the same period, Milan, rich as ever, has drifted to the left.
Immigration was the issue at the heart of these elections.
Such trends have been repeated in working-class areas in the rest of Italy and all over Europe. And of course there are many such equivalents in the United States as working-class areas have moved to the right, with the immigration issue playing a starring role. But, as in Europe, the American left has repeatedly refused to see anything wrong with a de facto policy of mass immigration, which is considered an unalloyed good contributing to a more diverse society. Therefore, to oppose mass immigration is to oppose diversity, which can only mean that you are racist and xenophobic. It’s that simple.
This attitude is a huge mistake because in fact there are rational reasons for voters to oppose mass immigration that cannot be reduced to racism or xenophobia. As Josh Barro, the host of Very Serious, notes:
Democrats … need to get back in touch with the reasons that both uncontrolled migration and excessive volumes of migration really are problems—not just political problems, but substantive ones. That is, they need to get back in touch with the feeling that illegal and irregular migration reflect a failure of our civic institutions, a misuse of the social safety net, and a breakdown of the rule of law, and that all of that is actually bad…
Illegal immigration, and other forms of irregular migration that happen with the authorization of the executive branch, really do hurt Americans by putting strain on public resources, imposing costs on taxpayers, and undermining social cohesion. And this has been particularly noticeable because of the huge surge in three categories of migration over the last few years: old-fashioned illegal immigration; migrants abusing our asylum system to gain years of legal access to the U.S., even without claims that are likely to be judged valid in the end; and the Biden administration’s large-scale use of the Temporary Protected Status designation to admit about a million mostly low-skill, mostly non-English-speaking migrants into our communities, especially from Haiti and Venezuela.
Here and in Europe, promoting and defending mass immigration has been a core part of the 21st century left’s project. And it has been a massive failure.
Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal democratic values wherever they are threatened, please consider becoming a paying subscriber today!
Climate change politicsAt the end of the 20th century, climate change was an issue on the left but generally a peripheral one. A time-traveler from the year 2000 would be shocked to discover how the status of the issue evolved in the intervening decades. Far from peripheral, it became a core part of the left’s 21st century project in country after country, including the United States.
This was despite a thunderous lack of interest from these countries’ working classes. But for these parties’ burgeoning Brahmin left constituencies, it became a non-negotiable commitment—after all, they were saving the world! As the 21st century unfolded, left parties’ policy plans increasingly centered on combating climate change and promoting a rapid clean energy transition. The claim was that the clean energy transition was not only a virtuous thing to do but would actually drive the economy forward. Hence, the Democrats’ Green New Deal, a version of which was implemented by the Biden administration, and similar schemes in other countries.
The working class has not been impressed. In the United States, these voters view climate change as a third-tier issue, vastly prioritize the cost and reliability of energy over its effect on the climate, and, if action on climate change is to be taken, are primarily concerned with the effect of such actions on consumer costs and economic growth. Making fast progress toward net-zero barely registers. The left’s assurance that the clean energy transition will deliver prosperity has fallen on deaf ears. The working class just doesn’t believe it will. And it hasn’t.
Nor do they believe the end of the world is nigh if the green transition doesn’t proceed really fast. And Bill Gates agrees with them! Only four years after publishing a book called How to Avoid a Climate Disaster, Gates—who has spent billions of his own money on climate change projects and is deeply associated with climate change initiatives—said the following in a recent memo:
Although climate change will have serious consequences—particularly for people in the poorest countries—it will not lead to humanity’s demise. People will be able to live and thrive in most places on Earth for the foreseeable future … Although climate change will hurt poor people more than anyone else, for the vast majority of them it will not be the only or even the biggest threat to their lives and welfare. The biggest problems are poverty and disease, just as they always have been.
OK then! Sorry about all that “uninhabitable earth” stuff. As much of the world transitions away from the green transition, with the majority of countries failing to submit decarbonization plans for the COP30 summit and with 81% of energy consumption still from fossil fuels, it’s now clear that making climate change politics core to the left’s 21st century project has been a huge mistake.
Cultural radicalismThe 20th century left generally tried to remain on the high ground of anti-discrimination, basic civil rights, and colorblind meritocracy. The left’s 21st century project went in a much more radical direction.
The high ground was left behind in favor of an ideology that judges actions or arguments not by their content but rather by the identity of those engaging in them. Those identities are defined by an intersectional web of oppressed and oppressors, of the powerful and powerless, of the dominant and marginalized. With this approach, an action is judged not by whether it is justified—or an argument by whether it is true—but rather by whether the people advancing it are in the oppressed/powerless/marginalized group or not. If they are, the actions or arguments should be supported; if not, they should be opposed.
This doesn’t make much logical sense, and it has led left parties, including the Democrats, to take many positions at odds with the concerns of ordinary voters. Voters overwhelmingly believe illegal immigration by anyone is wrong and should be deterred, not indulged, as Democrats have frequently done. They believe crimes should be punished no matter who commits them, public safety is sacrosanct, and police and policing are vital necessities, not tools of oppression. They believe, with Martin Luther King, that people should “not be judged by the color of their skin but by the content of their character” and therefore oppose discrimination on the basis of race no matter who benefits from that discrimination.
Perhaps most pernicious: the ideal of equal opportunity has been compromised by the commitment to a new ideal of “equity” that strives for equal outcomes. Lack of proportional representation of racial groups in desirable positions or achievements is taken as evidence of racism, structural or otherwise. Therefore, the outcomes should be equalized regardless of merit.
But voters’ common sense view is that opportunities should be made equal if they are not, and then let people achieve as they will. There is no guarantee, nor should there be, that everyone will wind up in the same place. Indeed, voters deeply believe in the idea of merit and they trust in their ability to acquire merit and attendant rewards if given the opportunity to do so. To believe otherwise is insulting to them and contravenes their view about the central role of merit in fair decisions. As George Orwell acutely put it: “One has to belong to the intelligentsia to believe things like that: no ordinary man could be such a fool.”
But perhaps nothing would surprise our time traveler from 2000 as much as the incorporation of transgender “rights” into the left’s 21st century project. Going far beyond basic civil rights in housing and employment, left parties in Europe and the United States have uncritically embraced the ideological agenda of trans activists who believe gender identity trumps biological sex, and that therefore, for example, trans women are literally women and must be able to access all women’s spaces and opportunities. The same goes for children whose gender dysphoria should generally be medically treated with puberty blockers, hormones, and, if desired, surgery to align their bodies with their gender identity.
This remarkably radical approach has until very recently been met with very little resistance on the left, including in the Democratic Party. But as evidence mounts that the medicalization of children is not a benign and life-saving approach, but rather a life-changing treatment with many negative effects, and voters stubbornly refuse to endorse the idea that biological sex is just a technicality, the left’s identification with gender ideology has become a massive political liability.
Leaving the high ground of anti-discrimination, basic civil rights, and colorblind meritocracy for these radical alternatives has been a defining part of the left’s 21st century project. And it has been a huge mistake.
Subscribe now
Economic growthThe 20th century left at its best understood the centrality of economic growth. That’s because growth, particularly productivity growth, is what drives rising living standards over time. The left sought to harness the benefits of growth for the working class, not to interfere with the economic engine of progress. They believed in the future and the possibilities for dramatic improvements in human welfare.
The left’s 21st century project has, at its core, been dedicated to other goals. They now prize goals like fighting climate change, reducing inequality, pursuing procedural justice, and advocating for immigrants and identity groups above promoting growth. This is remarkably short-sighted. Faster growth gives the left far more freedom to attain its goals. Hard economic times and slow economic growth typically generate pessimism about the future and fear of change, not broad support for more democracy and social reform. In contrast, when times are good—when the economy is expanding and living standards are steadily rising for most of the population—people see better opportunities for themselves and are more inclined toward social generosity, tolerance, and collective advance.
Reflecting this lack of interest in economic growth, the left’s 21st century project has not been techno-optimist, tending to focus instead on mitigating the negative effects of technological change. This is very odd. Almost everything people like about the modern world, including relatively high living standards, is traceable to technological advances and the knowledge embedded in those advances. From smart phones, flat-screen TVs, and the internet, to air and auto travel, to central heating and air conditioning, to the medical devices and drugs that cure disease and extend life, to electric lights and the mundane flush toilet, technology has dramatically transformed people’s lives for the better. It is difficult to argue that the average person today is not far, far better off than his or her counterpart in the past. As the Northwestern University economic historian and newly-minted Nobel Prize winner Joel Mokyr puts it, “The good old days were old but not good.”
Given this, the left’s 21st century project should have embraced techno-optimism rather than techno-pessimism. Rapid technological advance is key to fast productivity growth and rising living standards. But the left has been lukewarm at best about the possibilities of new and better technologies, leaving techno-optimism to the libertarian-minded denizens of Silicon Valley. As British science journalist Leigh Phillips has observed:
Once upon a time, the Left … promised more innovation, faster progress, greater abundance. One of the reasons … that the historically fringe ideology of libertarianism is today so surprisingly popular in Silicon Valley and with tech-savvy young people more broadly … is that libertarianism is the only extant ideology that so substantially promises a significantly materially better future.
Sound familiar? The left has ignored growth and its drivers to its great detriment. In its place, it has squandered enormous political capital on a 21st project that has largely failed. Twenty-five years is long enough; it is high time to try something new.
A version of this essay originally appeared in The Liberal Patriot.
Ruy Teixeira is a Senior Fellow at the American Enterprise Institute and co-founder and politics editor of the Substack newsletter, The Liberal Patriot.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
From the publisher's feed