Why Rome actually fell: plagues, slavery, & ice age — Kyle Harper

Dwarkesh Patel: Today, I have the pleasure of chatting with Kyle Harper, who is a professor and provost emeritus at the University of Oklahoma and the author of some really interesting books, The Fate of Rome, Plagues Upon the Earth, Slavery in the Late Roman World. An upcoming one called The Last Animal. The reason I wanted to have you on is because I don't think I've encountered that many other authors who can connect biology, economics, history, climate into explaining some of the big things that have happened through human history in the way you can. The most recent reason I wanted to have you on is I interviewed David Reich, the geneticist of ancient DNA. And some of the questions we were discussing, he kept emphasizing this overwhelming role and surprising role that diseases have had in human history, not just in the recent past, but I mean, in his work going back like thousands of years, tens of thousands of years. And he's like, you got to have Kyle on. I emailed him afterwards, like, who should I interview next? And he's like, you got to have Kyle on. You have this graph in The Fate of Rome. Yeah, you show human population over the last few thousand years. I assume that these two down spikes are both the bubonic plague, Yersinia pestis, right? Yeah. And so this is not like some small little nudge you can see. Like the overwhelming, I mean, other than the hyper-exponential growth in human population, the overwhelming, not just one, but the overwhelming two major features in human population going back the last 10,000 years is this one bacteria, right? One of the things you discuss in the book is that the collapse of the Roman Empire was a result of this one particular event that

Kyle Harper: Well, I mean, the period that I normally work on is sort of from the high Roman Empire, so like the glory days of the Pax Romana in the first or second century, which is usually where I start, through what we call the late antique or early medieval period, so the sixth or seventh century. And at the beginning of this period, Rome dominates this Mediterranean empire. It's what you think of when you think of ancient Rome. It's the largest city in the world. It's the center of this huge network. And then by the end of this period, the city of Rome has, you know, we don't know, 50 to 100,000 people. It's a 10th or 20th of its former size. And I think we now can say pretty clearly that environmental factors like climate, but also especially diseases play a part in that really big transformation. And well, there's a problem because we don't have the same kind of modern government mortality statistics that we do for like COVID or even for the last century or century and a half. We have to piece together from clues, but it's pretty clear that the bubonic plague events, whether you're talking about the Black Death of the 14th century, the Plague of Justinian in the 6th century, these events are capable of causing death rates temporarily that are just orders of magnitude beyond what we're accustomed to. And even in these ancient societies, the reason why these were so shocking, in a world where the death rate is always pretty high, probably several percent of the Roman population, three, four percent a year may be dying in a normal year. And so for them to just be utterly shocked by the death rate already tells you that it's some multiple of what they're accustomed to. Yeah. I think you discuss in the book the possibility that the death rate might have been

Dwarkesh Patel: close to or even over 60% wherever the Black Death. So this is not just like, this is like literally the most significant thing. Yeah, it's mind-blowing.

Kyle Harper: And I mean, I... In the case of the Black Death in the 14th century, it's pretty clear. It kills 50%, 60% of the population in entire regions. We don't necessarily think that it killed 50%, 60% of the whole continent, although that's actually not impossible. But even the fact that it's killing 50% of the people in cities, in provinces, in countries, is just beyond the damage that other plagues do.

Dwarkesh Patel: Right. And do you think that we're not for... this 60% mortality event, plus the fact that I think we haven't even discussed yet, this super severe cold snap, do you think that the Roman Empire might have otherwise just kept going? Because you discussed there were these two previous other big pandemics. The empire still survives. I think Will Durant had this quote that the Roman Empire fell for longer than most empires have lasted. So do you think like it would be similar to China maybe? There were – maybe a dynasty collapses but fundamentally the same sort of cohesive nation reemerges.

Kyle Harper: Yeah. I mean that's a great comparison, right? And not just sort of like decline and fall of dynasties but also like geographic changes in the configuration that parts get added and parts get – get cleaved off, but you still kind of think of it as fundamental continuity in the core. Like that to me is a very, very plausible counterfactual. Like Justinian, the emperor in the 540s, he reigns from the 520s to the 560s. He's on a path of success. He's retaken Africa. He's mostly retaken Italy when the plague hits. Like to me, a very plausible counterfactual is that a more or less Mediterranean core – of the Roman Empire could have survived east and west. So, it does sort of survive in the east. But even including really all of Italy, Africa, and probably Spain, that would have been like to me a very, very reasonable outcome of the sixth century if you hadn't had this kind of random shock.

Dwarkesh Patel: Yeah.

Kyle Harper: that there would be this, you know, the Roman Empire would keep going. And remember, it does, and it calls itself the Roman Empire until the 15th century. But we would think of it as maybe more sort of like really the Roman Empire if it still included the Western Mediterranean and was this major, powerful, urbanized polity that resists invasion from the southeast as happens in the seventh century. So yes, the answer is like I think that the Roman Empire absolutely could have had another turn sort of as the thing that we kind of mean when we say the Roman Empire, this pan-Mediterranean powerful, urbanized empire. Yeah.

Dwarkesh Patel: Okay. So one of the things I found really interesting was you were discussing the firsthand accounts of As this big—and by the way, feel free to explain the cold snap as well as it's happening. But the firsthand accounts of people who are experiencing this, some of whom come from this burgeoning Christian faith, which already lends itself to millenarianism and apocalyptic thinking. I'm curious, basically, how did different people try to make sense of— Like this once-in-a-thousand-year event that's super kind of like just intense.

Kyle Harper: Clearly, people have to try and explain within the elements of a worldview that they have how something like this can happen. And they don't have – They don't have modern science. They don't have germ theory. They don't think of it in terms of a biological event or a climatic event. Since that's come up and you've invited, I'll say a little bit about that, but this is one of the other really exciting frontiers where we're learning new things about the human past that we just didn't know 10 or 15 years ago. That in this case, we now have really, really cool paleoclimate data that helps us understand that this period of the sixth and seventh century was also a period of really abrupt and significant natural climate change. And so we're all familiar with like anthropogenic climate change, the carbon emissions stay in the atmosphere, trap heat. Humans are changing the climate. It's a big problem. You can talk about it if you want. I just like to clarify that like that view is not incompatible with the reality that the climate does also change for natural reasons on every timescale from like really long geological timescales to much shorter timescales. So we live in the Holocene. The last 11,700 years have been pretty. It's stable, pretty warm. It's an interglacial. We're literally between ice ages right now. And it's been really stable in the big picture. And yet, even within that stability, there are smaller scale climate variations and climate changes. And because we need to understand how the Earth system works, how the climate system works, in order to be able to model what's happening. We need an empirical record of what the climate has done. So for historians, this is like great news because now we have a huge number of sometimes even pretty high resolution climate reconstructions for historical periods across the Holocene. And so we now know, like we did not know this 20 years ago when I started graduate school, say, that The Roman period experienced some really abrupt episodes of climate change. In this case, the sixth century, we know the cause, that there was a series of really significant volcanic eruptions. Volcanoes are very powerful, short-term climate forcing mechanism. It ejects sulfur into the stratosphere. It aerosolizes and sort of creates a reflective shield that scatters the radiation entering the atmosphere. And so it leads usually to short-term cooling. And in this case, you had a series of really significant volcanic eruptions that cooled the climate for several decades and in some ways with a later series of eruptions even like a century and a half. And it wasn't just a little bit cooler. It was like a degree to two degrees cooler. Which we all kind of know now like two degrees, this isn't weather, this is climate. So like two degrees doesn't affect your day. But two degrees globally is a pretty different globe. And so all of a sudden in the late Roman world, it's much, much cooler. probably areas that have been wetter are now drier, places that are drier, maybe wetter. I mean, it changes the hydrological cycle as well, which is more complicated. But in addition to the shock of the plague, you have this simultaneous and probably not unrelated shock to the climate system. And so we know that it was essentially challenging for agriculturalists. That when the sun is blocked and it's really, really cold and the wheat doesn't grow, your society then starves. And so the Romans get this like wham, bam, double shock of climate change, famine, and plague. And so back to how people explain this. Yes, like apocalyptic thought is one of the principal sort of ways people frame it. To them, nature is going crazy. You know, huge amounts of the population are dying of this horrible sudden disease and the crops don't grow. And you don't have microbiology and you don't have climatology. So you explain it in the – with the resources of the worldview you have. And there's a huge burst of like apocalyptic thought in the 6th and 7th century, which is always kind of there. I mean you mentioned that Christianity's eschatological – It is, yes, fundamentally, but like that comes out in different ways with different sort of emphases in different time periods. And this is a period, the sixth century, when there's a really sharp emphasis on eschatology in Christian thought. I found your early chapters in the book about what the Roman economy was like in

Dwarkesh Patel: this happy period. Quite interesting. So – There's a bunch of questions I have about this. If you read Gibbon, writing in the 1770s, I think he says in the 1770s that if you want to look at the happiest time in human history, you go back to this period you're talking about. This is true, at least according to him, as of a couple centuries ago. This is still like peak civilization. And you discuss the complexity of the Roman economy, the fact that millions of tons of wheat and other products have to make their way to Rome and the trade networks and everything. Um, And then I think you basically say, like, look, they were experiencing productivity gains. The wages were increasing. Population was increasing. But they were still not at the level at which it was plausible that, say, for these climactic and biological factors, they might have had an industrial revolution. And so I'm curious why you think, like, basically, yeah, paint a picture for you of what, like, the Roman world looked like as of this happy period and why that was still, like, counterfactually not, you know, couldn't have just saved us a thousand years of history if they were on the right track.

Kyle Harper: Yeah, I mean, first of all, I think this is the sort of question that historians ought to worry about all the time. We ought to be thinking about why didn't the Roman economy catalyze the takeoff, because in some ways it It was so precocious for its time period and it seems not utterly impossible, right? The Roman world is still a pre-industrial economy. So agriculture is the dominant sector. The majority of people work in agricultural pursuits and productivity is low. They don't have modern mechanized traction. They don't have modern synthetic fertilizers. They don't have the modern... Green revolution yields, all the things that have made agriculture stupendously productive. So just like the primary sector is fairly limited in terms of its productivity because of the sort of limitations on technical inputs. And we can think of the inputs to an economy are going to be capital, labor, and ideas. And what the Romans, they've got people. They have some investment, but like they just – They don't have technology. They don't have ideas. It's a late Iron Age civilization. And I do think there's productivity growth. And that productivity growth comes from markets, from trade, where you get comparative advantage. In Egypt, I'm really good at growing wheat. You can make glass in Syria and then we'll trade. And the urbanization of the Roman world certainly facilitates that. Cities are these sort of like hubs of productivity and exchange. And there's some technology. I mean the Romans, you know, if you look really, really hard over like five or six centuries, there's certainly like economies of scale where the production process and manufacturing is sort of moved from artisanal to sort of, you know, industrial scale. But there's not really – there's no takeoff because they don't have – science. They don't have research and engineering that drives continuous productivity gains. So I think they go like precociously far in a pre-industrial setting where you take trade really far. They have good institutions in terms of There's strong property rights. There's relatively reliable contract enforcement. There's financial markets. They have like the most advanced financial markets in the world before like the 17th or 18th century. There's impersonal financial intermediation. So like it's not like you have to know me and come ask me for a loan if you want to go build a ship and go trade something. There are banks that take money from depositors. And keep balances and then lend out to debtors who want to go and do entrepreneurial things. So they have like so much potential, but there's just no spark. You never see these sustained productivity increases. And I would just say ultimately it's because the Romans don't have technology improvements that are really self-sustaining. And the reason they don't have that is because they don't have science. Their science sucks. I'm offending some of my colleagues, I'm sure. Galen is great. Ptolemy is incredible. I love plenty of the elders' encyclopedia. But if you look in the big picture, the contribution that the Roman Empire makes to our knowledge of how nature works and then the applied technology that comes out of that – is really pathetic for five, six hundred years. And so they go as far as you can with Smithian advantages to market exchange and specialization, to banks and finance, but without the kind of creative destruction of new technologies that improve productivity. you're eventually going to run out of improvements.

Dwarkesh Patel: If you're like Augustus or some other Roman emperor and you're like, look, we've got this big economy, but I want to see productivity gains. And you want to make it happen somehow. What is it exactly? Is there something you like from a... top-down perspective you could have done. I mean, in Britain, you know, the government subsidizes the royal arts and so forth. That's what I was going to say. Yeah, the longitude prize and so forth.

Kyle Harper: That's exactly what I was going to say is, I mean, this happens first in France and then Britain, but you get royal societies for science where you're doing really – I would say there's like three things that are essential there. One is like the promotion of what we would call basic or fundamental science. So it doesn't all have to be like immediately practical or commercialized that you're like promoting deep knowledge of nature. Yeah. Two, you're doing it in an empiricist way and this is something very important in the 17th century that the Romans by contrast don't have is you have like the spirit of Francis Bacon that we need to ground our knowledge in experiment and observation, not just believe whatever authorities or Aristotle said. And that's very much the spirit of places like the Royal Society is we don't take things on anybody's word, especially like Aristotle's. And so you need basic science. You need empiricism, like this rigorous and self-correcting. And then third, you need a sense of useful knowledge. And that's the other thing that really comes together in the 17th century is not just – the basic and abstract science, but the application. And the 17th century language for that is useful knowledge. And that is something that doesn't ever get wired together in the Roman Empire. There are tinkerers and engineers, but they're not talking to the mathematicians and the physicists. And so if you were from on high to like design self-sustaining innovation, I think you would want to bring those elements into proximity. And I guess this – unfortunately for them, didn't do it. Probably good for the world. The Romans are pretty nasty people in a lot of ways. I definitely am of the opinion that sort of the high science matters. That like Isaac Newton is not a tinkerer, right? He's not building like pumps. But the guys who are – are like his friends. And they're in and around the Royal Society. And they're absolutely – I mean, look at Denis Pepin, who's a French engineer, who's like very much in the circle of like Leibniz and the like very high abstract mathematics, is trying to build like vacuum pumps. And the proximity of like high math, high science, very abstract, with the – What is ultimately going to be the sectors that lead to mechanization where then you can harness this new source of energy or this – not new source of energy. This sort of source of energy that is there all along but hasn't been tapped in coal. That's what catalyzes the big positive – cyclopositive feedbacks. don't have is that what the romans do have is the the kind of specialization and like now that we look for it it's there like when you when you look at you know food processing which is a huge sector the way that they build mills like there's definitely improvements um but there's never like the the catalytic change where you get runaway positive feedbacks that's right yeah so a previous guest of mine

Dwarkesh Patel: nat friedman I don't know if you saw this, but he launched this challenge called the Vesuvius Challenge. Oh, yeah. This library of Herculaneum in 79 AD was buried under the ashes of Mount Vesuvius when the volcano erupted. Speaking of volcanoes. And now they figured out with modern techniques how to read the burnt scrolls. And it is supposedly the biggest... library of classical text ever. It would double the amount of classical text we have. As a scholar of ancient Rome, what would you personally find most fascinating? Where are you excited to find from this data?

Kyle Harper: Well, I mean, I'm super interested in the history of math. It'd be sort of like... What happens after Euclid? Because it's very hard to say, but you get like these really interesting people like pop up, like Diophantes, who's later in the early Roman Empire. But like there's still like really interesting math going on. And I mean Euclid is incredible. Like that is – the Greek – like experiment in math and science is the one that I think had like the better chance of sparking sustained takeoff. And it didn't. And it'd be interesting to know more about like why. Why did things stall? Because these people aren't, you know, like Euclid is not – just like a towering genius who comes out of nowhere. He's very much a product of the culture and the questions that are being asked in the generations before. But it just sort of feels like, you know, after him, you fail to get that kind of sustained, continuous progress and advance. So, maybe sort of back to that big question that we were asking before, like what – What is it that prevents the kind of breakthroughs that we see in the modern world? What is the population of Greece during their golden age? Well, I mean of like the greater Greek world or like individual city-states like Athens. I mean we think of Athens as being like – a couple hundred thousand people. Like not massive. So I wonder if it's like the Greeks had a science but not the people to sustain a

Dwarkesh Patel: modern economy. Or not a modern, but even a sort of like industrial economy. And then the Romans had the people but not the science.

Kyle Harper: Yeah. Yeah. There's probably something to that that like there's just not the critical mass of educated people, of like sheer cognitive power to like keep it going. Yeah.

Dwarkesh Patel: In the last couple of episodes, I've mentioned how my sponsor, WorkOS, helps companies like OpenAI, Perplexity, and Cursor get enterprise ready so they can grow faster and scale further. Of course, WorkOS can't actually help you build or sell your core product. That's your job. But they can handle all the IT requirements that are needed to get you a seat at the table and let your product shine for the enterprise. Building these security and admin features in-house is a massive commitment and could require multiple years of investment, making it harder for you to deliver your core value to your enterprise customers. WorkOS has APIs for all of the most common enterprise features, like authentication, permissions, and encryption, so that you can quickly plug them in and get back to building. If you want to learn more about how WorkOS can help you get enterprise ready, go to WorkOS.com. All right, back to Kyle. Let's turn to another one of your books, the one about slavery in the Roman world. I did not realize before I read that one how central—I mean, how much Rome was a slave society. I guess that just, like, isn't a salient thing in a sort of, like, conventional understanding of Rome. Yeah, but why don't you paint us a picture of how much slavery was involved? I mean,

Kyle Harper: slavery, you know, tragically is, like, a really important institution throughout history. We sometimes tend to think of it as, like, a distinctly modern phenomenon, but that actually misses— The deeper picture and in fact it obscures the importance of modern slavery because modern slavery is uniquely important and it's uniquely tied up with certain kinds of market exchange and certain kinds of production, certain kinds of racial ideologies. There's like things about modern slavery that's really important to understand. Those are different, but not just because slavery is there. Like slavery has this longer history and slavery is more important in some societies than others. And we want to try and understand that to ask why and then what implications does that have for understanding those societies. Rome is one of those societies. Slavery is really important. a prominent institution in Rome from the late Republic. As the Romans conquer other parts of the Mediterranean, they start taking captives as slaves en masse, and they build an economy that really relies on slave labor in important sectors of the economy. So plantations where commodities like wine, olive oil... are produced for market exchange that allow landowners to amass enormous amounts of wealth. Slavery becomes this really important institution that's entangled in the development of the Roman economy. you know, maybe the third or second century BCE. And then with ups and downs and really important changes along the way for centuries and centuries.

Dwarkesh Patel: As you're pointing towards like from the supply side, all the Roman conquests lead to all the surplus labor that they can make use of. And on the demand side, these cash crops. Yeah, exactly.

Kyle Harper: I'm very big proponent of the idea that you have to have both, right? You have to have a source of slaves. And after the conquest stops, the Romans figure out other sources of slaves. And if anything, the demand is equally or perhaps even more important. Because if there's not a mechanism, if there's not institutions that let you turn this kind of exploitation into cash flow, the institution's not going to go very far. And so it really is the institutions, the presence of markets where you can You can take labor and turn it into profit. That's the most important element.

Dwarkesh Patel: One of the things I find interesting is in the age of colonization, we're used to thinking about slavery in terms of race, but also like maybe like religion and other things which more obviously demarcate free and slave populations. Yeah. In the Roman world, it doesn't seem that that's clearly the case. Yet, there's no abolition movement the way that emerges out of England in the 19th century or something, or maybe even before that. And the reason that's mysterious is, look, if you're literally descended from slaves, if you're like, my grandfather was a slave, but then we were freed, and they're basically just like you, you would think that there would be more of a sense of like, not everybody would be an abolitionist, but at least some people would be writing about abolition and something. Yeah. And you got like Christianity and so forth burgeoning and they didn't seem to have a problem with it. Like why is there no abolition movement despite this sort of like heterogeneous nature of the state population?

Kyle Harper: Yeah. It's – It's sort of like disturbing in a way, isn't it, that the humans have the ability to convince themselves that it's okay to own other human beings as property through a variety of different kinds of ideological justifications. And you see even in the ancient world, there's different models. that people use to say that slavery is OK. I mean Aristotle develops a theory of natural slavery that actually some people deserve to be slaves by their very nature and that it's actually good for them to be in bondage. What's really interesting though is that doesn't actually ever seem to be like the dominant ideology. The Roman ideology of slavery – is not racialized. It's not like the Romans think that the Greeks or the Germans are like, you know, some fundamentally separate kind of human that justifies their exploitation. The Roman ideology of slavery is really rooted in the law of property and status. So they think that slaves are people who've been conquered and rather than killed, they've been spared, and they've been sold into the condition of being somebody else's property. And this seems to mentally explain to them where their slave system comes from and why it's justifiable. And so you have different kinds of criticism of the slave system from within. But remember, most of what we have written is from the slave ownership class. It's not like – I don't think the slaves were themselves – you know, believing this ideology. And there must have been sort of what we would think of as abolitionist movements or spirit that we just don't have really good records of. But it is like this curious thing that the Romans are able to build this huge system that's really brutal and really violent. But... has this kind of flimsy ideology where they tell themselves these stories. But the deeper lesson of that is that humans can create these systems of belief that will exclude others and justify almost any form of exploitation and convince themselves that it's okay. I hope your next book about The Last Animal discusses the potential parallels with

Dwarkesh Patel: factor farming.

Kyle Harper: Well, there's a pretty gruesome chapter, I'll say that.

Dwarkesh Patel: I don't know if you mentioned what numbers you said, but I think it was like 10 to 20% of the population under the Roman Empire was enslaved, or whatever the number is there. But given that larger size of... a slave population, it's surprising to me that there's so few slave revolts, not only in Rome, but even throughout history. Like there's Spartacus in 71 BC. Then there's the Haitian Revolution. I forget when, but like probably late 18th. Yeah, I was 17. Yeah, okay. You know, if like 20% of the population is enslaved, like how is this sustainable? You know, if you're like running a farm and there's like 4,000 slaves and then like the next farmer over also and like... Why aren't there more slave rebellions? Yeah. Right?

Kyle Harper: Why not? I mean, how did they do this? I mean, they have a really elaborate system of repression. They're worried about it. And probably... The parts of Roman society where there are 20, 30, 40 percent slaves are pretty limited to certain regions and certain time periods. And partly because once you cross some kind of threshold, the challenges of repressing sort of direct violent resistance. But it's... You know, it's a system of exploitation. That means there's always a mix of carrots and sticks, to put it crudely. Like the Romans extract people's labor partly through, you know, physical violence, but also partly through like systems of manumission that try and incent people to— to obey and not to rebel in order to earn their freedom. And, you know, they're using everything from literal chains to like enticements to try and keep rebellion from ever sort of coalescing in a way that can turn into collective violence. But yeah, I mean, it's a little bit... It's a little bit challenging for us to look back and say, I mean, we know like in Pompeii, the slave population in Pompeii is huge. It must be 30%. And not all of these people would have been like plantation workers who were lashed every day and worked to the physical bone. A lot of them are nurses and textile workers and maids and tutors and all sorts of things that are sort of quasi-embedded in households as well, where there's always this weird psychological dimension too. I mean, part of the strangeness of slavery is how it's deeply embedded in domestic institutions as well. And so there's ideologies in which the paterfamilias is sort of the father and the master that sort of brainwashes, tries to brainwash people against But it's just like the important thing to recognize is just like a pervasive system that tries to colonize people's minds and pervasively tries to keep them from resisting.

Dwarkesh Patel: And I wonder if we can close the loop with the question we began with, which is why didn't Rome have an industrial revolution? I don't know if it's a plausible explanation that cheap slave labor reduced the incentives for mechanization and engineering and other crafts or if not. I don't know.

Kyle Harper: It's definitely an argument that's been made. Aldo Schiavone, who's an Italian historian, has argued that. It's kind of like a neo-Marxist tradition that argues this. It's an interesting argument. I don't buy it at all. But the good version of that argument would just be that the Roman Empire is using, in fact, slaves in many of the most forward elements of the economy, too. We tend to... I think because we think – we know slavery is bad and we think progress and economic growth and innovation is good. We tend to think that those things don't go together. But in reality, it's like the most economically advanced sectors of the Roman economy that have the most – a high degree of organization productivity that tends to employ slaves. And so in the Roman world, like you could make the argument that if the labor in those sectors had been free, there would have been more opportunities for positive feedback loops. The way the argument is usually made is just that the Romans got rich without really thinking about They just wanted to extract labor, extract wealth rather than create wealth, which I don't think is like a terrible argument. But I just – ultimately, I don't think it's the system of labor that keeps the Roman world from industrializing. And there are lots of sectors in the Roman world where slavery is not – a dominant institution. And it's not like they're more productive or like flirting with some kind of breakaway. So it's an interesting argument, but not one I've ever found all that persuasive.

Dwarkesh Patel: Final question about Roman slavery. What did Gladiator get right and wrong about? Will they just abduct you in front of your house?

Kyle Harper: You know, I haven't seen the second one. So you mean the first one? The first one got right that when you're making a movie, you should worry more about making a good movie than historical actually the first one's a great movie and if it was like completely historically accurate it would have been much more boring um so uh so i'm not uh gonna be critical of of that movie like it it plays very loose with the facts of high politics um around commodus and the creation of this character but who cares i mean russell crowe's incredible and it

Dwarkesh Patel: But on slavery in particular.

Kyle Harper: On slavery in particular. I mean, I think actually that's one of the strong suits of the movie is you see this like completely exploitative system that brings people from very different parts of the world who have very different backgrounds. And then the system of like urban spectacle is very real. And the use of slave labor in that. It's certainly a part of it. So the movie actually gets like some really important things about that right. And that makes it totally forgivable that it has to create a kind of high politics storyline. Yeah.

Dwarkesh Patel: Okay, I think that covers all the questions about Rome. We can get back to your most recent book about human history and plagues. What do you make of the general argument that people have often made that we were living in a sort of Eden before agriculture, especially given your... You know, you've explained that all these diseases that we're sort of stuck with are actually quite new. If we take that perspective seriously, was life before human population exploded and we had agriculture just much more pleasant, at least in comparison? Homo sapiens is 200,000 to 300,000 years old.

Kyle Harper: They emerge in Africa. and disperse, multiply, but we spend 90, 95 percent of our history as foragers, so people who are hunter-gatherers who take energy from wild food sources rather than sedentary farmers who've domesticated plants and animals and live a sedentary lifestyle where you You're enslaved to this wheat or rice, but it gives you reliable calories. That is – along with the Industrial Revolution and then whatever this thing we're about to go through. That is the biggest change in the history of our species other than those others. OK. So the shift from foraging to farming. It affected everything. It affected – our beliefs. It affected our genetics. We're all basically genetically different, adapted to live in a different kind of environment with different kinds of diets. It affected our societies. It affected inequality. It affected culture in every possible way. And of course, it affected our health in really basic ways. It affected our labor regime, so doing the same kind of labor over and over every day is very different from running around as a hunter chasing deer or whatever, which sounds quite nice. It had changed, so it changed our labor regimes. It changed our diet most of all. Foragers tend... to eat high protein, high fat-ish diets with no refined carbohydrates but like limited carbs and it's a very varied, highly varied diet. Sedentary farmers tend to eat more monotonous diets and they tend to be like dependent on grains and starches. So like very narrow spectrum for your calories. So changes in labor regime, changes in the diet. And then changes in lifestyle, being sedentary and living in big populations, that then puts you in proximity to other humans, puts you in proximity to human waste. So feces are a major, major conduit of infection. And it puts you into proximity to the air that they breathe, which is conducive to respiratory diseases. So this transition... Which, by the way, takes thousands of years, right? It's one of these things that's more of a process than an event. But it has massive implications for human health, including the infectious disease environment that we inhabit. So it's not like hunter-gatherers were living in paradise. Like the infectious diseases that they had were seriously burdensome. They sucked and probably most people died of infectious disease. They – you know, malaria is a really, really old disease. Lots of diseases existed in the Pleistocene, in like our Paleolithic past. So it's not like it was Eden. But there is this idea that the transition from foraging to farming – I mean, Jared Diamond called it humanity's biggest mistake. And the – You know, certainly these changes entailed some things that like were not positive net for humanity. And one of them is that it definitely increased the infectious disease burden. So simply as our population multiplies and as we're in contact with feces and as we're sharing the air through which respiratory pathogens can spread, diseases are constantly changing. trying to take advantage of this. That's just how nature works. Energy is scarce. Everybody's trying to steal it from everybody else, including microbial parasites. And so the disease burden of humans over time definitely increases. And the burden of infectious disease on humans goes up over time. So very broadly across these thousands and thousands of years, The diseases that are suffered by, say, people by the time of the Roman Empire are absolutely much worse than what had been the case in Stone Age times.

Dwarkesh Patel: James Scott has an interesting theory in Against the Grain. Yeah. I don't know if it originates with him, but he argues that one of the reasons that the early agriculturalists were so successful, and David Reich, by the way, if you've seen this stuff about like the Amnaya 4,500 years ago, conquer all of Eurasia, but before them, the The Anatolian, the initial farmers, they're the ones who displaced the initial hunter-gatherers across Europe and Asia. Anyway, so he argues that initial wave was so successful because of these first, I guess... diseases that the farmers had created the conditions to engender. And basically, the relationship these farmers had with respect to the foragers that they were taking over from was similar to the relationship the Europeans had to the Native Americans, where inadvertently the disease is just a significant player in why you were able to dominate them. I don't know how plausible you find that.

Kyle Harper: I mean, I would... First thing, say – like it's important – I think you were starting to get at this – that there's never like a generation of humans that has the opportunity to make this choice once and for all. Like should we stay like hunting mammoths or should we become sedentary farmers with like – basically torturous dentistry and die by diarrhea. Like this happens over thousands of years through an evolutionary process where nobody can – it's a story of unintended consequences, right? I mean the mammoth are gone because – partly because we killed them all. And so people start – like the first livestock that are domesticated are goats. And nobody says like, hey, let's become goat farmers. The goats are wild. They're ibexes and people are hungry – And so they start managing them to only kill the males to make sure that they can reproduce. And they start pinning them and they start killing the wolves who are trying to attack them. And over very, very, very long periods of time, this becomes this tight mutualistic relationship where – All of a sudden we're goat farmers, you know. But no generation makes that whole decision for anybody. So that's part of it is that it's unintended consequences that are made in very, very incremental steps. And then two is like I definitely agree that there's some kind of like cultural selection here where the farming groups are simply so much better. more adapted to extract energy efficiently from the environment, right? It's all about energy. You want to multiply, you want to grow, you want to survive. It's all about energy. And so the foragers require huge landscapes to extract enough energy to feed themselves and grow and reproduce, whereas farmers per unit of land can extract such higher rates of energy that then can be, you know, through photosynthesis is captured and turned into edible sugars that we can metabolize. And so those populations are just growing faster that they, you know, quote-unquote out-compete the hunter-gatherer populations, say, of Europe that are largely but not completely Now, on top of that – so just like the energy story alone is a big piece of it. But then on top of that, you probably do have some kind of population difference in the exposure and possibly even immunity to infectious diseases. So I definitely think that – Early farmers, the first farming societies that are starting to live sedentary lifestyles where you have aggregations. These are not cities. These are villages. But still, that's more than like a hunter-gatherer band. And your childhood is then going to be constant exposure to – to a series of pathogens, those kinds of populations, when they're then migrating into Europe or probably carrying these pathogens with them that may have had a kind of further effect that on top of just being able to extract more energy and multiply faster, drives up the mortality of the existing populations.

Dwarkesh Patel: The point you made about fertility is interesting. I can't remember where I saw this, and it might be wrong, but I vaguely remember reading that for – it's not just the fact that you're not – like the energy density is lower. It's like you're also moving around a lot. And so that – because of that, you're spacing out kids much more so than if you're just like in the same place. And I think like the actual fertility for foragers is like not – is like sort of like a reasonable – I don't know if sustainable is the right word because I don't mean like in ecological sense, but more so in a like a – like keeps your population constant. Yeah, I mean don't like make me swear, but it's like more like four than six.

Kyle Harper: Right. Because women who are, you know, moving with a foraging band, you know, miles and miles on foot – on average a day, and also carrying kids, right, are going to have very different life history than sedentary populations. And that's very clear.

Dwarkesh Patel: JOSHUA SHARFSTEIN- Publicly available data is running out. So major AI labs like Meta, Google DeepMind, and OpenAI all partner with SCALE to push the boundaries of what's possible. Through SCALE's Data Foundry, major labs get access to high-quality data to fuel post-training, including advanced reasoning capabilities. SCALE's research team SEAL is creating the foundations for integrating advanced AI into society through practical AI safety frameworks and public leaderboards around safety and alignment. Their latest leaderboards include Humanities Last Exam, Enigma Eval, Multi-Challenge, and Vista, which test a range of capabilities from expert-level reasoning to multimodal puzzle-solving to performance on multi-turn conversations. Scale also just released Scale Evaluation, which helps diagnose model limitations. Leading frontier model developers rely on Scale Evaluation to improve the reasoning capabilities of their best models. If you're an AI researcher or engineer, and you want to learn more about how Scale's Data Foundry and Research Lab can help you go beyond the current frontier of capabilities, go to scale.com slash Thwarkash. One thing I'm really curious about is what effect these diseases through history have had on the cognitive functioning of people. I mean, yeah, you discussed this in the chapter about more recent history of the great divergence and probably attributed to the productivity of Europe that they were able to have public health earlier. But literally, if you go back thousands of years, you mentioned, for example, that Caesar was 5'5", and that was considered tall during his period. diseases and malnutrition whatever it was that caused these physical health effects also mean that like the average IQ was much lower just because of like you know you're like when you're a kid you're sick and that steals away nutrition from brain growth or something yeah short answer yes long answer we know that in the modern world like say over the last 250 years first in western European societies and

Kyle Harper: their settler offshoots and then more globally and more rapidly globally, there have been really deep physiological changes in the average human, right? So we're talking about populations with distributions. And what's happened is really two things. One is – There's more energy per capita. So people eat more. They eat more calories and they eat better calories. They eat lots of bad stuff too. But like people eat more. And two, the burden of infectious disease has been lowered. And growth, the growth for a human is a very complicated trait that's influenced by genetics. So, you know, I was never destined to be – You know, super tall. But it's also affected by environment, which includes but is not limited to what you put in, nutrition, and then what you spend either doing labor or what you spend fighting infectious disease. Infectious disease imposes a huge burden on the body. the immune system is extremely metabolically expensive. And so if your childhood is spent just fighting infectious diseases, you're going to struggle to invest energy and growth. So there's a massive increase in the size of populations over the last 250 years. And even though it's an even more complicated trait, This improves people's cognitive abilities. People are smarter. Like, may not feel like it. But – and I think it is rapidly leveled off. But – You know, people are more intelligent today than they were 100 years ago. Their brains are better nourished and they spend less – their bodies spend less time fighting pathogens. So I think there's no doubt that pre-industrial populations – and again, populations. So you still have – you know, you still have your Isaac Newtons who whatever infected him as a kid didn't slow him down. But at the population level, I think there's no doubt that not only were pre-industrial populations – Shorter. This is just a total fact that we know from their bones. But they probably also, on average, had sort of a lower distribution of cognitive abilities, but with a big distribution.

Dwarkesh Patel: Yes, you have this great profile in the book about living in London in the 18th century and just how disgusting it was. Pretty disgusting. But literally at the same time, in that city, you were just mentioning... I think there's these scientists and people with like towering intellects who are basically figuring out how the universe works and how to make all these machines and so forth. And so, I mean, one answer is just like what you just said, like, look, the distribution was lower, but that's still like maybe Newton would have had an IQ, a standard deviation higher if he was born today. But just like seeing that from the small population, you're seeing so much genius. Yeah, I guess the question is, How is that still... How could you have had this much of a deleterious impact on cognitive functioning and still had enough spare geniuses to kick off the Industrial Revolution? Yeah, I mean obviously it didn't keep them from discovering some pretty amazing things.

Kyle Harper: So it couldn't have been completely destructive. But that's one of the things that's interesting about the early modern period in the 17th, 18th century in particular is it's sort of this between period where – where you have sort of the pre-industrial and the modern that are still like mixed together in these really interesting ways. And so the example I use in the book is the very famous diary of Samuel Pepys, who's this incredible figure and is like very close to Newton and that social group. And his name is on the first edition of the Principia, right? I mean, these people are like This close to each other. But like, you know, the stuff that I evoke, I won't – you know, this is a family podcast.

Dwarkesh Patel: I don't want to say it. But like, you know, the stuff that Peeps does bodily functions is like mind-blowing to us.

Kyle Harper: It's vile and disgusting. But at the same time, like – Right down the way, you've got people who are making the most fundamental discoveries about the nature of the universe and inventing machines that will improve productivity and ultimately economic output. So that's what's precisely so weird and interesting about that particular period is you have this kind of mingling of the old and the new.

Dwarkesh Patel: Yeah, when I had Joseph Henrik on, you know, one of the things he discusses is if you look at, I mean, cultural evolution has figured out some remarkable things, right? If you look at the cuisines of different cultures, apparently the spices they use match the antimicrobial and antifungal properties you need in that particular biome. At the same time, you're like watching, I mean, reading that part of the book, I'm like, okay, I get in some cases they just genuinely did not have the resources to invest in public health and so forth. But, like, come on. I don't know if I should mention some of these things as well. But, like, you're just, like, sleeping in your own vomit and so forth. Like, why didn't cultural evolution or some, like, foresight just be like, hey, like, this we can sort of, like, do without?

Kyle Harper: Yeah, I mean, it's a deep question. And what I think we don't think enough about is, like, how... Like how – in a really fundamental way, how hard are some problems to solve? Like some problems are just like very, very hard to solve. And even though the incentive is really there, you think, ah, that took a really, really long time to figure out. Even though if you'd only known, like it would have made your life so much better.

Dwarkesh Patel: Right.

Kyle Harper: And it also – like there's tons of trial and error, right? So the example that comes to my mind is the invention of vaccination, which is like one of the great human achievements, like of all the public health improvements. is the most important one. Public health is never perfect. It's this like system of like six or seven really critical tools that involve clean water, personal hygiene, vaccines, antibiotics, different kinds of therapeutic interventions or rehabilitation therapy. And we're still – we can't like fend off all the germs. Like you have to have all of that and you can sort of like achieve this equilibrium state where you mostly have it under control. Vaccination is like the most important one. And it took forever to find the first vaccine. And it took this huge period of like – All kinds of weird trial and error, like inoculation with the actual smallpox, which is very, very dangerous, not vaccination. Vaccination uses cowpox, the lymph of an infected cow, to intentionally cause the immune reaction of humans. Before that, people would inoculate a person with actual smallpox, which is just giving somebody smallpox.

Speaker 2: You do it through the skin.

Kyle Harper: But like you're giving somebody smallpox. But it was absolutely in like a utilitarian way. It was the rational thing to do. It had these horrific death rates, right? We would never get FDA approval. But in a world where like 10%, 20% of kids die of smallpox, it's this horrible decision. But you'd be rational to do it. And so that – we actually don't know where that comes from. It may come from Africa. It may come from China. It spreads for like a century or more before Jenner discovers vaccination. So like it's clearly like really hard to figure that out. And then even after Jenner, it's like another, you know, 60 years, 70 years before Pasteur kind of systematizes it and says, hey, we could – Do this for everything. Yeah. Right. So so some of these some of these discoveries and innovations take they're really, really hard to discover. But then the beauty of cultural evolution is that we can store that information and you and I don't have to figure out any of that. We can go on to the next problem. Yeah. Because that's now been like collectively stored in the library of cultural evolution. It's known, you know, we don't even think about it most of the time until it's controversial. But like what a blessing to, you know, to live after people like Jenner and Pasteur who figure that out.

Dwarkesh Patel: Yeah. There's this great blog post by the author Slime Old Time Old where he's discussing. Wait, author what? Yeah. You don't know internet culture, you know? There's a bunch of weirdos out there. I live in a different world. Sorry. What did you call them? Sly multi-multi-multi.

Kyle Harper: You can't just drop that. Like, I'm going to let that one slide. Okay. I got some homework. Anyways, they have a blog post about scurvy and why it took so long to discover. And he was discussing all these sort of like – it's sort of like an epistemic

Dwarkesh Patel: conundrum because you can like use lime and you realize like, oh, it works. But then if you use lemons, which have much less vitamin C or maybe it's the other way around, they just work much – Way worse. And then there's like certain kinds of fruit which have vitamin C, certain kinds which don't. So it's actually hard to kind of figure out like what is it if you don't have a mechanistic explanation about how you solve this problem. And I think they had once figured it out and then they lost the knowledge until it was rediscovered again. But it makes it all the more mysterious that the kinds of things that Henrik discusses forage our societies having figured out. Literally, there's this 10-step process for how to process a certain kind of beans so that you don't get cyanide poisoning. And if you mess up any one of those 10 steps, you're going to get cyanide poisoning. But a society just figures out the right taboos and traditions to process beans. But you can figure that out. But this thing which is causing 20% mortality, you only get in the 17th, 18th century

Kyle Harper: Yeah, but we need to think more about the computation that's happening. You said it takes like 10 steps to figure out how to process this one particular kind of food. But I'm guessing... It is just really hard to figure out infectious disease. It's a really, really steep mountain. And once you get, like, up to a certain plateau, then the discoveries, like, come really, really fast. They become systematic and they become more fundamental. But it was really hard to get there. You know, not that many societies... really scaled it, not even within the societies that did. Like it was just a handful of people at first, but they did get there. Yeah.

Dwarkesh Patel: Okay. And then asking about where different countries were at around this time, what evidence do we have about what was actually happening in India before the British or the Mughals? Because it does seem to be the sort of black box in terms of historiography, but... Like, do we know if there were these huge plagues? Yeah, it's such a – it's a tricky problem because – start with the third plague pandemic

Kyle Harper: in the late 19th century. We know that that's in India. And India is a big part of its history. It's, in fact, where the plague bacillus is discovered by Alexander Yersin. Yeah. It's called Yersinia pestis. In his honor, a Japanese scientist finds it at exactly the same time, gets left out of the nomenclature. But the … Trevor Burrus It's a special kind of honor to … Trevor Burrus Yeah. Trevor Burrus The worst pathogen ever. Immortality. The plague is definitely in India in the 17th century. Trevor Burrus Yeah. Trevor Burrus And we know that from contemporary written records that are pretty inambiguous about the presence of the disease. What we don't know is was it there before that? And if not, why not? Because it kind of actually seems like it's not, at least not in this like same explosive way. And that's pretty curious. Like we don't have a great explanation of that because – You know, India is connected to the Central Asian world where the plague is endemic. There's plenty of trade. It would have plenty of chance to move to the subcontinent. So we don't understand that. And then if you go back even further, you know, that's the Black Death. You go back even further to the late antique period – We – it's like a total mystery and the Indian sources from the – even the fifth and the sixth century are not great. They're hard to use. This is totally outside my language abilities. They require totally different expertise. I've worked with some people who think that there are like – oblique references that may be interpreted as epidemic. But one of the interesting things is we actually think that the plague moves through India to get to Rome. This is not definite, but the plagues – And enzootic, it's like natural animal reservoir, is the Tianchan Mountains where China, Kazakhstan, and Kyrgyzstan meet. And we can actually, in the scheme of things, kind of identify a pretty small region where the pandemic lineage comes from. And we know that it doesn't go overland. So it's not like the Black Death, which goes across the steppe. The Mongol trade networks, military networks carry it. In the 6th century, probably the plague goes south through India and maybe like the ports in Gujarat or along the west coast that are still pretty connected with the Roman world, with East Africa, with Arabia, with the Red Sea. Right. that the plague travels on ship across the Indian Ocean because it shows up, the plague of Justinian shows up in the Red Sea. And so that is a clue that it probably is imported on this seaborne commerce. But how it, you know, how it got from Central Asia to Gujarat is a hard question.

Dwarkesh Patel: Huh. And I mean, I know the way we found that the Yersinia pestis existed in these Yamnaya 4,500 years ago is by, didn't they just find like the, I don't know how, but like if you can figure that out, why can't you look at the fossils of people 500 years ago or 1,500 years ago and just see if they have Yersinia pestis in them?

Kyle Harper: There's two things. First of all, you have to look. Yeah. not nearly the same amount of ancient DNA laboratory work that's happening on remains from ancient India. So if you're not looking, you're definitely not going to find it, and people aren't looking. Secondly, it takes a lot of luck for it to preserve. So the DNA molecule starts degrading. The second you die, it just starts falling apart. And... Even in the best of cases where we're getting it from – usually you're getting it from – if it's pathogen, you're usually getting it from the dental cavity. It's human DNA. You're getting it from the inside of the skull. But it takes a lot of luck for it to preserve because the soil conditions will affect the degradation. The temperature will affect the degradation. And just in a crude sense, heat is bad. And so that's why there's more DNA, ancient DNA that's preserved at more northern latitudes so far. But it has as much to do with the fact that people aren't looking. We should be looking. And if you've got skeletal materials from an ancient mass grave in India, call me. We can definitely look.

Dwarkesh Patel: And just to be clear for the context. Ancient. Ancient. Ancient. Going forward to the future a little bit – Speaking of future technology, maybe the one that's more relevant than AI is synthetic biology. And there's a worry that you can potentially create diseases which Maybe the evolutionary gradient is one that is, like, not catastrophic, where diseases are incentivized to be transmissible but keep you at a chronic level of infection that doesn't necessarily kill you immediately. Actually, it's interesting why the bubonic plague diverges from that selection pressure, which maybe you can answer. Yeah. Yeah. What do you think about the potential that with synthetic biology, people can make diseases that have the transmissibility of measles, but also the deadliness of something like Ebola? Is that given your understanding of biology and whatever? Yeah. How plausible is that?

Kyle Harper: Well, let me let me start with the plague where I'm a little a little more comfortable as I can say something as a as a knowledgeable person. But that but I think is relevant. Because you said like it's weird that the plague seems to sort of evade some of these evolutionary constraints. And it's worth just like saying what these are. A pathogen is a disease-causing organism, a microbe, usually a virus or a bacterium, but also fungi and single-celled organisms like protozoans that cause disease in a host. But like they're not trying to cause you disease. COVID doesn't hate you. Plague doesn't hate you. It's just evolution. It's just trying to steal energy or hijack your cells to reproduce its genes. it has incentives to try and do that as well as possible while doing the least possible damage. And so it's always kind of trying to thread that needle or to find the right balance because if a pathogen just kills you instantly, there's nothing to steal and it can't transmit its genes into the next generation. And so it's a really – every pathogen has these like basic evolutionary problems. How do I get from one host to the next – And how do I evade my host's immunity, which our immune systems are incredible, for long enough to multiply? And so most pathogens, you know, this is clunky. There's not like a perfect equilibrium. But they have to like explore the space where there are these various constraints. And... And they find all sorts of weird ways around it. And evolution is really, really good and really, really creative, unfortunately, for us. And just like the tricks that they find to like hide inside your immune system or to like fake it out are really, really wild. But – So one of the – I think there's two reasons why plague is so weird and like we don't completely understand why plague is so weird. But I think there's two basic reasons. One is that it's vector-borne, which means that it's transmitted through another organism that is the intermediate. Mm-hmm. Arthropod or insect vectors are really annoyingly, you know, helpful to certain pathogens. And most – there's actually a relatively small number of diseases that are transmitted through a vector like this. But they tend to be really, really nasty like malaria, typhus. And they can kind of get away with it because even if like you're dying – You know, a mosquito can come and bite you and transmit malaria to me. Plague is a vector-borne disease. And it's very, very well adapted to transmit, particularly by fleas, but we think also maybe by lice and other biting organisms. But really by fleas, it's really, really good at transmitting by fleas. And that's evolution. Actually, this is one of the cool things with ancient DNA. We've been able to like piece together at the like absolute molecular level, the The genetic changes that let it make this protein that like have this effect in fleas. It's really weird. It forms this biofilm in the gut of the flea that like chokes it and makes the flea feel like it's starving. And so the flea just starts feeding and feeding and feeding. And meanwhile, it's regurgitating bacteria.

Dwarkesh Patel: So can I ask a question about that? Why is it the case – because there's diseases that hijack the flea's mind or ant's minds or something. Are there – Like why isn't there a disease that makes humans zombies? Is it just the human brain is like so complicated that it's like – Let me come back to this.

Kyle Harper: Let me come back to this. We can talk about zombies, but that's a – we need to wind up for that.

Dwarkesh Patel: Okay.

Kyle Harper: So one, flea – so the plague is vector-borne and it's really, really good at like manipulating the fleas and it's just evolution. Two, I said this before, but it's an animal disease. We're like – Collateral damage. We're totally irrelevant to the like really core evolutionary history. The plague just wants to infect rodents. Of course, it's not really like wanting to do this. The plague makes a living. It survives out there in burrowing rodent colonies. We're like – I mean, it doesn't care at all. And so it has no evolved incentive to modulate its virulence to be able to transmit sustainably. Plague never sustains itself in human populations. It can transiently infect human populations, but then it always dies out. It becomes extinct, that lineage. And then what is the reason that you have these 1,000-year cycles, basically, where...

Dwarkesh Patel: Why is it not 500 years? Why is it not 10 years? Why is it like – what causes it to go dormant? What causes it to reemerge? Yeah, this is – you need to ask me in five years because we've learned so much and now this is like the thing that we fall in the category of like almost a new question

Kyle Harper: now that we can ask because now that we have the Neolithic – lineages and the Bronze Age lineages. We're starting to piece together this fuller history, but we still don't even totally understand the boundaries of when is the plague really not circulating in human populations, and what are the factors that cause it to be so explosive? Is it evolution of the bacterium? Is there something about the genetics of the lineages that escape from the animal reservoirs that are especially transmissible? Is it Human ecology, like that we put rodents, like black rats, you know, in the right place to get the disease is that there's something about the climate stress that renders the – we don't have a great understanding of like why the plague comes and goes. So that's scary. And the – So connecting it to your other question about like these superbugs, I mean what's interesting in the very big picture about the plague to me is like even like the history of infectious disease is like on the one hand like there's a real core of it that's just like basic principles of ecology and evolution. Right? I mean, we do certain things in the environment that creates the conditions that pathogens can evolve and take advantage of. But on top of that, like, evolution is just creative and weird and contingent and unpredictable. And it's those little, like, contingent facts that can end up having these really, really huge effects. And so in the case of the plague, like, you would never – If you were like really, really knowledgeable about the basics of ecology and evolution of disease, you would never be like, I think that, you know, every now and then a rodent disease from Central Asia is going to wipe out half of the continent. Like that shouldn't – that's not predictable. That shouldn't be happening. And actually that one is kind of an outlier. But like infectious disease is always kind of like that. I mean tuberculosis has probably killed more people maybe than any other infectious disease. It's like this horrible disease. It's just this – We don't really understand it. Now we really don't understand where it came from because it doesn't look like it is an animal host before it has humans. And it's just a weird disease. It's just a bacterial pathogen that in the huge world of bacteria, this one is very, very good at hiding. And so it gets in your chest and it just lurks. And then it'll just waste you away, particularly if you're poor and you're stressed. And so like... There's some core principles there, but then it's just like something weird about it. It's just like this terrible luck that makes it what it is. And so to me, like there's going to be another pandemic, you know, maybe bird flu, maybe something else. But like – It's the real, real outliers and the weird ones that we should maybe worry about a little bit more than we do. Like if you want to go to zombies, like I'll go there. You don't have to twist my arm too hard.

Speaker 2: But like,

Kyle Harper: you know, like prion diseases or like fungal diseases where we don't have the same, nearly the same infrastructure and level of knowledge, biomedical research as we do for like bacterial and viral diseases. You know, something, if we create the incentive, evolution is going to find some weird ways to exploit it. And it's not just transmissibility and virulence. Those are like two really basic parameters. But, you know, when you look at even COVID-19, part of what made it insidious is it just has just the right parameters to be latent for just long enough. Like the first COVID, SARS-CoV-1, 2003, slightly more virulent. And in fact, it was just more virulent enough that it made you sick pretty quick. And just that little difference was enough to contain it because you could figure out who was sick. COVID-19 was impossible to contain because it took, you know, several days before you really presented with clinical illness. And it's just that little quirk that made it totally impossible to control through non-pharmaceutical interventions early on. And so, like... follow that train of thought if pathogens are going to find ways to take advantage. And there may be pathogens that push the limits on latency that can be very, very hard to control. So I think that's one of the takeaways of the big evolutionary history of our pathogens. Evolution is very weird, very contingent, very creative at exploiting whatever weakness we give it. It's because there are Billions and billions and billions of microbes in this room. You know, I don't know how many tens to hundreds of millions of species of microbes are in this room. Most of them are not even remotely pre-evolved to pre-adapted to be pathogenic. But like lots are and they're constantly, you know, they're constantly seeing if you manage to lock that door. And they're just looking for a way to break in.

Dwarkesh Patel: Yeah, yeah. Okay, just a couple more rapid fire questions for you. Have you found tools like deep research useful for especially your kind of work where you just have to compile insights for many different fields? If you throw in a question, the kinds of questions you honestly investigate, and now maybe they can rely on you as a citation for those particular questions about what effect did climate have on the fate of Rome or something. But if you just had a different question, which like maybe you'd write a book about in the future, how well do they do at synthesizing this kind of literature and coming up with a thesis the way you do?

Kyle Harper: Yeah, I mean amazing but not yet like completely displacing or like totally threatening the kind of work that a historian does. But at this point, like I can't – I can't even conceive of what a research project would look like without using AI. I mean, it's just a constant.

Dwarkesh Patel: Oh, really? That fast, it's like become so central into your work? Yeah, but like for, you know, for just like – it's just like a constant conversation partner.

Kyle Harper: When you're doing research, when you're writing, you know, you can go back to that PDF and ask like, you know, whatever. How many species are there in this Texan? Or you can just ask the AI. Yeah. And you still have to check it, but it's getting – obviously, it's getting more and more reliable really, really quickly. But I think it hasn't yet like – In some of the deeper research, it's not the equal of humans yet. And then in the synthesis, it's really not. There's still that creative element of synthesis that is where conceiving of the question is as important as the answer. And it doesn't feel like it's like... right around the corner, but it's changed.

Dwarkesh Patel: Have you used Deep Research?

Kyle Harper: Oh, yeah. I started using it like two weeks ago or something.

Dwarkesh Patel: I don't know how long it's been around.

Kyle Harper: How long has it been around? Somebody told me about it.

Dwarkesh Patel: It's not that much longer.

Kyle Harper: Okay. Somebody told me about it like, I think it was like less than two weeks ago. And yeah, it's incredible. I mean, it's really incredible.

Dwarkesh Patel: Yeah. And then I want to touch on your next book that isn't out yet, The Last Animal, which One question I have is basically how worried should we be about extinction given that we're on the cusp of technologies which will make it possible for us to reanimate many lost species? I assume if you have their genome or something, our descendants will be able to make more woolly mammoths and saber-toothed tigers and so forth. Should we discount the value of endangered species as a result?

Kyle Harper: I would say no. We should still be concerned with extinction for a couple reasons. I mean absolutely this is a legitimate serious scientific field to like understand the genomics of extinct animals. And there is like a small but serious enough science of de-extinction. And it's feasible that some organisms can be targeted for serious de-extinction efforts. At the same time, like a couple of thoughts. One is like I'm not that optimistic that it will work. Not because I think it's necessarily impossible, although it's not yet totally feasible, particularly for animals that don't have like very similar modern descendants. But it's – It's because a species isn't just a genome. A species is an organism that inhabits a food web and an ecosystem. And, you know, we could bring the woolly mammoth back, but there's nowhere for them to live. You know, the mammoth step that they need to thrive is not there. And, you know, there's really very little point in bringing an animal back from – extinction just to put it in a box at a zoo to sort of like, you know, satiate our curiosity about it. So without the ecosystem, you can't have the species. And really part of one of the themes that I try and get at in the book that I'm trying to finish is like we need to think about living systems, ecosystems. And the extinction question is very much a question of like what kinds of systems will exist on the planet. And I think whatever happens technologically in 100 years, 1,000 years, the impacts that humans have on biodiversity is going to be very, very long-lasting. We're part of a species that has been impacting biodiversity for over 10,000 years. And there's things we can't do. There's things we can't undo. There's things we can't change about the past. But we're making decisions right now that will be binding on the future whether our descendants like it or not. And so we need to think very hard about like what choices do we want to make to keep intact the kind of variety and vibrancy of living systems. that in 1,000 years, 10,000 years, that will be a huge part of our legacy. Like the impact that we make on the stream of macroevolution will be one of the really big things that our species does. And it can sometimes be very hard to recognize that in like our individual lives, but collectively it will absolutely be part of our forever legacy on Earth. And so we need to think very carefully about the choices that we make.

Dwarkesh Patel: I think that's an excellent note to close on. Just to plug one more time, we've been discussing plagues upon the earth, which is the history of disease going back through the Neolithic to modern times. Fate of Rome, which discusses the plagues and the history of the Roman Empire considering climate and biology. We also discussed, what was the name of the book on slavery? Slavery in the Late Roman World. Slavery in the Late Roman World. Yep. And the upcoming book is The Last Animal. The Last Animal. All linked in the description below. And where else can people find you? In your descriptions. That's it. I'm not on social media. Sorry. Okay, got it. Well, you can find him here on this podcast. Just begin again. Exclusively. Thank you for tuning in. I'll see you in the next one.