School of War - Ep 247: Nicholas Wright on the Brain Science of War

Episode Date: November 11, 2025

Nicholas Wright, neuroscientist and author of Warhead: How the Brain Shapes War and War Shapes the Brain, joins the show to explore the implications of neuroscience for warmaking.  ▪️ Times 01:20 Neurology 02:57 Intelligence Defined 07:06 Mapping the Brain 17:05 How Modeling Happens 27:15 Outthinking and Outlearning the Enemy 30:25 Prioritization and Survival 35:15 Fear and the Fall of France in 1940 44:48 Enhancing Military Training 49:47 AI and Its Impact on Human Cognition Follow along on Instagram, X @schoolofwarpod, and YouTube @SchoolofWarPodcast Find more content on our School of War Substack

If we want to understand the nature of war, we need to understand human nature. And to understand that, we need to understand how the human brain works. Today, neuroscientist Nicholas Wright joins us to talk about his work on how brains shape the conduct of war and how war has shaped our brains. Let's get into it. It is for safety for war. We'll sum of us up. It will live in it.

A bloody experience of Vietnam is to do. and they state we continue to face the rain situation in the ground. We'll fight on the beaches. We should fight on the landing ground. We'll fight in the fields and in those streets. We shall never surrender. Hi, I'm Aaron McLean. Thanks for joining School of War.

I am delighted to welcome to the show today. Dr. Nicholas Wright. He's a neuroscientist. He's an advisor to the Pentagon's joint staff. He researches the brain technology and security at University College London, and he is the author most recently of Warhead, how the brain shapes war and war shapes the brain. Nicholas, thank you so much for joining School of War.

Thank you very much, Aaron, for having me on. How does one become a brain scientist? I didn't even go through the typical route of becoming a brain scientist. So I was a neurology doctor. So I was a doctor. I did a medical degree. And then I did internal medicine and soul. And then I specialized in neurology.

So that's treating patients with neurological conditions. And I became fascinated by the brain. I've been fascinated by it for many years before as well. And then I ended up doing a PhD and more research using things like functional brain imaging, that's brain scanning, look at how people make decisions. And that is how I published a variety of papers and did lots of, you know, analysis and the rest of it. And that's how I became a neuroscientist. Tell me about the state of neuroscience today.

And what I mean by that admittedly very broad question is I have a, a, prejudice that probably stems from things that I read when I was very young that I've never circled back on, that we've come to know a lot about the human body in the last few hundred years of medical science. We kind of know how the heart works, kind of know how the lungs work, et cetera, fairly mechanical stuff, but that the brain remains tremendously mysterious, even today, and there's all this talk about artificial intelligence and how to mimic intelligence with machines and things like that, whereas my prejudice against, again, sort of learned it in early and never actually circled back upon,

certainly not with anyone like yourself, who's in a position to state things authoritatively, is that we don't even really know what intelligence is. So how are we doing, Nick? Well, the great thing you'd be pleased to hear is is if you've got five academics in a room, they'll give you 17 different definitions of intelligence. So there is nothing I can say that is going to be completely satisfying.

But what I would say is that we have learned an enormous amount. So if you think, for example, back, I mean, you talked about learning about the body, but we didn't really even know how cells worked. We didn't know that evolution until the mid-19th century. You know, these are pretty fun, what we've now thinks is pretty fundamental aspects of biology or what a lot of people would. How did we get our ideas about the brain? You are absolutely correct. It's been a long, slow process.

So, you know, you talk about the heart, and we know love another heart, but we could only really understand how the heart works once we had good microscopes that could look at the heart properly and stain the heart properly, staying the cells in the hearts, reanalyze it, exactly the same with the brain. In the late 19th century, there were remarkable new technologies for looking at brain cells, right? But the really big turning point happened in actually in 1943, around then, when we started coming up with new ideas

about how the brain worked, around how we don't just perceive the world, like our passive television set in our brains or our minds. What we do is we use a model of the world. We carry around models of the world with us. and we use those models, like literally it can be physical models of the world around you, for example, that we carry under our hippocampus. It's not literally a model of, you know, it's not a metaphoria model as well.

We carry on a physical model of the physical world that we map out in our hippocampus, which is a brain region. So we have these models, and that was really a big turning point. And over the subsequent, you know, eight decades, we've added an enormous amount more. And there's not one simple technology that's given us a lot more knowledge. It's many different technologies that work together. Things like genetics. You can switch different cells on and off in rodents or fruit flies is something they

often look at. And in humans, we can look at patients who have parts of their brains missing through a stroke or a tumour or some rare condition. We can actually temporarily lesion parts of a brain. Some experiments I've done, you can apply magnetic pulses through the skull, which lesion parts of the brain, stop parts of the brain working. Then you can see how the brain works when you know, when part of it's missing. And you can do brain scanning. So that's things like looking at what parts of the brain are active when people are doing different tasks. And so we put all of these

types of technologies together and they're very powerful. And just to illustrate the kinds of ideas that come out of it, both last year's Nobel Prize is in AI, they were both strongly linked to my research institute in London, which is in Queens Square, which is part of sort of UCL, University College of London. And Demis, a Sabbaths, for example, he won, he did AlphaGo, which you're probably familiar with, and you know,

and beat the world's go champion. He led that team, and is the lead researcher at DeepMind and set up DeepMind. And he and I shared an office for a little while whilst he was finishing his PhD and I started mine, and he was researching the brain,

and he was researching imagination. He was looking at the hippocampus. That area I talked about was important for imagination, and it's also very important for memory, and we use those memories. Why do we have memories? We have memories not to make some, you know, the radical history of the past.

We have memories to help us navigate the future. And so his PhD was about how do humans imagine things, right? And he was looking specifically how the hippocampus did that. And those sites, sorts of ideas are, and lots of others from neuroscientic been central to the advances we see in AI. Could you give us a tour of the brain and its parts just briefly? tour. Yeah. Highlighting, highlighting the things that you think are of most interest to people who are mostly concerned with military matters. Yeah. So I'll start down at the brain stem, which is right down at the

very base of the brain. And the brain stem is very important for things like pain. So, and this helps illustrate kind of what a lot of all the reason why we have a brain. So why do you air and have a brain? You have a brain so that, to link senses coming in, right? and link that to actions that you can take that help you achieve your goals. Those girls might be staying alive or reproducing or whatever it might be. That's why you have a brain. That's why a mouse has a brain. That's why a fruit fly has a brain.

So down at the brain, then, we see things like pain. So pain is a tissue alert response. And say, for example, you're a spitfire pilot and you're hit by a bullet. When people are actually hit by bullets, often they describe it, they describe it as like a prodding finger or being hit by a lump of mud. If you're a spit firepilot hit by a bullet, your model can sit there and it can turn down that pain. So you're not immediately just going, ow, and can't do anything. They can then think, well, what should I be doing here?

Should I be, you know, warning my comrades, should I be, you know, trying to take that dome of bomber? Should I be trying to head back to base? Should I be bailing out, whatever, right? So the model can sit there and it can help us, can sit between sensation and action and help us choose from a repertoire of different responses. So that's just an illustration, but that's just an illustration, but that's, right down at the base of the brain. As you move up, you get to things like that work with emotions, things like fear or anger. And they're really important for, you know, giving us,

giving us ways to respond quickly in very uncertain environments. And, you know, you can't, as I said before, fear is useful, you can't do away with fear. Then we move, we keep moving further up and we get to brain areas that help us navigate, for example. And then we go on to what kind of people think of us, the brain, you know, like the sort of the big bit of, on the outside. That's the fancy real estate, right? That's the super fancy stuff. That's what you'd be paying the big money for, this big bit of the brain. And there again, you've got big areas that are about sensation and big areas that are about taking actions. So if you look at the areas of sensation, for example, vision. So a good chunk of everything you eat, a few percent of everything you eat,

the energy goes towards just your brain processing visual information. Every time you eat a sandwich, A good bite of that sandwich is just there because you have a giant visual cortex, and it is very costly, energetically, to feed that visual cortex. How does that visual cortex work? So it works through using, now, you get too much information into your eyes ever to just process like that information like on a passive TV sets in the brain. So the way your visual perception works, for example, it's the same for all perception, but visual perception.

I'll just use as an example. the way your visual perception works is you perceive a model of the world, right? And that model of the world, it's an actively generated model, it's a created model. So to give you an example, if I'm looking at your nose, but then around you, I can see blue sky, I can see lots of things. But actually, your retina, my retina, only has color, good receptors for color and fine detail right in the center. So I should just, everything else should be seen colorless and blurry around. around the center of your face, but it doesn't.

I'm seeing lots of colors. And that's because what I'm actually seeing, I'm perceiving as a model of the world. Now, that model of the world is generated by my expectations about the world, by what I expect to see. Right. It's a generated, literally, it's like generative AI. It's a generative model.

So, you know, generative bottles are not new for neuroscience. It's literally a generative model. I'm generating things, generating what I'm seeing. And then it's just updated using sensory information, coming in that can update it. Now, what does that mean, for example, the military people, that means that whatever technologies we use, messing around with people's expectations

and can always trick them, right? And so, you know, I talk a lot of, give lots of examples, camouflage, cheating, communication, confusion. You know, you see this in, you know, with Israel, having to fight in a city of rubble, right, which is what the Russians did to the Germans at Stalingrad. These things will always be part of it. of the thing. So, and, you know, reconnaissance would always be crucial because you need better updating,

right? And people don't think enough about that. Now, we then get what sits between sensation and action, right? So that's that. And we have, like, amazing ways of making actions. What sits between sensation and action? That's thinking, right? We could just call that, because, you know, colloquially, we can just call that thinking. And so we have large areas of a sort of fancy real estate that are that sit between sensation and action and therefore thinking. So what type of thinking? It can be, for example, I have a model in my brain of you, and I have a model in my brain of what you think of me, and I have a model, and, you know, I have a wonderful model for thinking about others' intentions, right? I have incredible brain capacity for that, so do you, right? And again, this is very energetically

expensive, a good bit of sandwich that you eat goes towards, like, fueling that, that, that, that model you have your ability to mentalize, to have theories of what other people's intentions are. And we can exploit those. And I'll just come to the last part. So again, we have amazing models for planning in the future, right? And we're very good at planning. And the models we have for planning are, you know, they have to be flawed because, you know, we have to deal with a complex world and all the rest of it.

But I'll just say one last thing, which is about, again, about, and I think this is really important for people who work in the military. and I've written about this for the Pentecost for a number of years and for lots of things on this. And this is our thinking about thinking. This involves the frontal pole, this bit that's really most distinctively human right at the front of the brain. That bit of the brain is very important for thinking about thinking, right? You, for example, Aaron, now you can think, you can make a decision about what it is that you see in front of you, for example, or what it is that you think about something.

You can then interrogate your own thoughts and ask yourself, How certain am I of that? Right. How certain am I? The next guess they want to have on is some random British neuroscientist, right? Or whatever it might be. You can do that. You can interrogate your thoughts that this metacognition, this thinking about thinking.

That is really important. So it was actually essential. So I'm sure you've read The Soldier in the State by Huntingdon. And one of the key things that he talks about there is why were the Prussians better than the French in the Franco-Prussian War? It was that they really fostered this idea of reflection. reflective officer. They went on lots of staff, right, thought through lots of things reflected on their own ideas, right? The Krieg-Shby or, you know, all of this stuff was about

reflecting on what they thought and thinking, and that really helped them. And more broadly, self-reflection is central to wisdom and wiser choices. What was the point of the Office of Net Assessment? It may say it exists. I don't know. Obviously, it's complicated. It's coming back. It's coming back. Okay. He never went away. It was always there. It was a recess. It was always there in our hearts. And now it's coming back out.

So, obviously, what's the point of the Austin assessment? It was partly about thinking about red, but in the context of blue. You know, it was thinking about the other, but in thinking about the context of the self, right, there's an important element of a self-reflection. I'm looking out at Washington Monument. Why did Washington argue for a Senate? He argued for the Senate, so it was a reflective chamber, right? He literally, there's this wonderful, sadly, probably apocry exchange rate.

He basically says that's the point of the Senate is for reflection, to pull the tea into the saucer, so it calls. you can have a think about it, right? That's working out great. Well, I mean, you're there. It's all right. It's not literally a hellscape behind you. It looks really nice.

You're right. Just to say, we can use... That was cynical of me. That was cynical. And we can use this metacognition, and that is really crucial. And, you know, I was... You're probably not false pent skies

in the Pentagon a few weeks ago. And those officers need the ability to reflect, to decide what types of actions they should or should not be taking, to think how sure they are about what it is they're doing. and that's going to be really important, how confident they are. All leaders need to think about confidence. And those are all the types of things.

There's not one part of the brain that matters. We need to think across the whole of the brain, and every part of the brain is important. I would say that would know. Every part of the brain is important, but all of them have lessons that we can apply in the military, but also in our everyday lives and in politics and everywhere else. That was a really interesting tour.

So from the animalistic brain stem all the way up to sort of paraphilosophical or maybe just actually philosophical or strategic, the front parts of the brain and everything in between. The notion of models that your work emphasizes and that you keep circling back on here in our conversation, how does a model like the ones you're talking about actually form? So you've made reference to you and I have formed, you know, in fairly hasty fashion on limited information models of each other. You know, we're just meeting each other for the first time and learning of each's other's existences in recent weeks. Presumably, you know, I've grew up in Northern Virginia and later came back to it and lived there now. Presumably I have a model of Northern Virginia in my head. You know, how did that happen? How did these models actually coalesce?

So, I mean, a lot of the models we have, like, I mean, for example, you and I will both have physical model. I've been, I lived in Washington, and, you know, you and I both have literally in there a physical model of Washington, D.C., a map, right, of the terrain. And it's linked to the terrain, literally through things like the Washington Monument, landmarks, right, help us anchor our model of the world, right, which is in there, which is then linked to reality, right? And how do we learn that? So the most classic example that if you've read any popular neuroscience book, you will probably have read this, but it happened to be done in my lab in London. Again, Queens Square in London by a lady.

Sadly passed away very recently, Ellen McGuire. And she did a wonderful set of studies where she looked at London black cab drivers, London taxi drivers. You have to do something called the knowledge, which is where they have to basically learn how to get essentially anywhere in London, just purely from memory. And she then compared them, and she compared people who passed the knowledge exam to those who didn't pass knowledge exam.

She compared black cab drivers to bus drivers who spend their days driving, but they don't have to remember where they're going really beyond the basics. The gist is learning the knowledge changes the structure of your brain in the part of the brain in the hippocampus that is particularly involved in navigation. And that is a very robust finding. So when we learn it changes the structure of our, changes the structure of our brain. If I teach you to, I don't know if you juggle, but if you don't juggle, but if you don't juggle, I teach you to juggle. I can show brain changes if I scan your brain before and after.

So we have an amazingly plastic brain. But a lot of the basics have to just be in there. For example, you have a visual cortex that is broadly the same as my visual cortex. You have maps in your hippocampus and it's broadly going to be roughly the same because we're both humans with it. It's built on the same template. So how do our models form? I'd say one of the most interesting things in this actually relates a little bit to what Jolsey Marshall did with his Fort Benning Revolution. So he reduced the class sizes so people had to teach more. So people then learned more in smaller classes, but then they had to teach. So much of what we do as humans, I think I'm incredibly clever because I can do things like

hammering nails into pieces of wood and whatever. But actually, I don't know, you know, I used something about I have to teach my, I had to teach my kids a few years ago how to hammer nails into pieces of wood. I'm right-handed and both my kids are left-handed. And I had to then use my non-dominant hand to hammer it. It's actually a lot harder than it looks. And things like hammering nails into pieces of wood is actually quite difficult. It's something that your parents or other people will have taught you how to do. So much of what we take for granted is actually stuff that's taught. We can type, we can read, we can write, we can do all these things. We get these through learning from others who teach us. And what's remarkable is there's this spiral of learning and teaching that's

really crucial. One of the key ways we humans get our models of the world is through this learning and teaching spiral. Now, we are the only primates that teach. No other primates teach. Other primates learn, but no other primates teach. So chimps, for example, they can learn. They can observe their mother cracking stones with a cracking nuts with a stone or something. But the mother won't actually spend time and energy teaching. We are amazingly good at teaching. And, In addition to that, we are incredibly avid social learners. So we will carry on learning about social stuff, right, learning from other humans, even we'll carry on doing things, even when a chin would go, well, this is a waste of time.

And they've created these clever experiments where humans carry on doing all these superstitious movements with puzzle boxes and all sorts of stuff when they shouldn't. So teaching and learning is crucial for how we form a models of the world. And I'll just give you one last example, which is really crucial. And it's something I'm doing literally right now, which is language. So we, how do we, again, something that we can do, that no other primates can do, how do we form groups of more than a couple of hundred individuals that are coherent? No other primate can do that, none at all. So people talk about societies falling apart, and they say that's remarkable.

The remarkable thing is we can have any societies that function at all, right? Because no other primate can even get to the level of a tribe. that's, you know, some hundreds or thousands of people together. No other primate can do that and we can do that. So how do we do that? So one of the ways we do that is we learn that we create humans who have identities that are similar enough that they can act coherently together.

I'm sure you went through this in your military training and growing up in Virginia and working in think tanks and all of this stuff, right? We create humans whose identities are similar enough that they can work together. They can work together as part of a coherent group. And that is intimately linked in a spiral with culture. So culture here is just a set of rules,

but how things are done around here. So at the Hudson Institute or whatever, there will be a culture, right? In the military, there is a military culture, the US military as a culture. How things are done around here. Culture helps create the identities. The identities help create the culture.

And that spirals to change us. and we are plastic. Our ideas change about who we are and what we are and what we're part of. And that, for example, is another way that we can create these groups that know others can and that our models are formed

that help us answer, for example, the question, who am I? Like, if you have to answer the question, who am I? Like, a lot of that would come from the cultures which are apart, and it's now formed part of your identity. And then you, in turn,

like, are forming your identity is such that you now form part of that culture. So our models are very plastic. They're really quite remarkable. And they do change our lives. But a lot of them is, you know, a lot of it is, is inbuilt as to how we can achieve these remarkable feats of learning and changing of our models. Yeah.

No, it's fascinating. And your point about teaching in particular is fascinating. You know, it speaks to the existence of this front part of the brain as you were, you were speaking about as sort of the most characteristically human part of the brain. So we, unlike the other primates, as you say, we're sort of able to think about the things that we're doing. And so go one step beyond just letting our young or letting whoever imitate us and actually break it down. We can analyze our own actions and explain in a way that is beyond animal capacity.

Well, can I introduce one interesting. That is what everybody thought until about 20 years ago when they discovered some other animals that do teach. Oh, not primates. Non-primates animal. It's genuinely really interesting. And this tells you something again about humans, right? So another animal that teaches is meerkats. You're probably familiar with meerkats.

You know, there's very clever little mongoose-like type guys. So meir cats have to eat dangerous prey, right? They eat, for example, scorpions. So meir cats have to teach young meir cats how to eat scorpions because scorpions are just dangerous, right? So what they do is initially they'll bring like a dead scorpion, right, for really young meerkats. Then a bit later they will bring a scorpions.

scorpion with the tail, you know, with the stinger nipped off. And then finally, the mere cat has learned enough about scorpions that they can tackle a whole scorpion. Now, they're not as good at teaching. We are so remarkable teaching. They actually, the meerkats do that based on the age of the mere cat pups. They don't do it based on what they think of others know. So you're absolutely right. We can do it orders of magnitude better than they can. But the key thing is, is it helps us do things that involve being smart, because meerkats are obviously smart, and doing things that are potentially very dangerous, like hunting a mammoth or whatever. Well, it's fascinating.

I think the basic point I was going to make remains, though. Well, no, no, no. It's totally fascinating. But for all of our justified preoccupation with these semi-unique or at least uniquely well-developed frontal capacities for analysis, explanation, reflection, et cetera, too much emphasis on all of that distracts from how in other ways we are quite basic or animalistic. and we are quite, I mean, this is something that Plato was onto a long time ago, right? That we are governed by these models, which is to say we're governed by these images,

this world of images of things that we live within. And it's cool, and it makes us, it's made everything possible that we have this last, you know, whatever percentage of capacity that is rare, virtually, you know, virtually not existent. It's quite rare amongst other animals. But still the images govern us, and in the context of warfare, what this allows us to do is the trickery that you emphasize in your book. So one example that you speak about in the book is Patton's fake army

before the invasion of Normandy in 1944. Hitler in particular, but a lot of the Germans, they look at the map of the British and French coasts, and it's just sort of obvious that the invasion is going to come at Calais. It'd be a little, it'd be difficult and sort of self-defeating to go elsewhere. Calais is the obvious place.

They're probably going to come at Calais. So what do we do? We lean into it. We lean into it elaborately. and give all kinds of indications and information that that's where we're going to go, all to create a kind of over-reliance on these models that are going to in the end give us major advantages, because, of course, we're not going to Calais, we're going to Normandy. It's manipulating the sort of animalistic parts of others' brains through our superior reflection

and hoping that their reflective capacities are going to be a step behind, that they're going to be kind of governed by their image world. and their model worlds. Yeah, I mean, outthinking your adversary has very often been an advantage in war. And I think that's the thing. So we need all of the different parts. For example, the will to fight, right, and the will to fight aggressively, that is only partly to do with our more sophisticated, the fancy brain regions that you're talking about

there to do with reflection and so on. It's partly to do with that, but a lot of it is things like, you know, controlling fear, using harnessing fear, harnessing anger, training, right? A lot of that is, you know, he's crucial for those types of things. On the other hand, outthinking your advocacy has always been crucial as well. You know, a lot of what he's, I mean, he talks a lot about more basic stuff, but a lot of what he talks about is trying to outthink the enemy, right? And we can become better at that.

So I think, again, you know, from a military perspective, I remember, you know, I don't know if you know Michael Howard, who was a military historian, Mark Howard. You know, he talks about what was the biggest military innovation in the 19th century. He doesn't focus on, you know, technologies. What he focused on was the Prussian, which then became German general staff. And the idea was that basically they were beaten. They thought they had the best army in Europe, right?

And then they were beaten by Napoleon at Yeina-Auerstadt. And they were like, they went away and they had to think, oh, God, how are we going to like beat this guy? He's really good. Because he was a genius. He was a military genius. So what they did is they built. So, you know, clouds of it, but also there were some in the civilian world as well.

I didn't have, I'd love to have gone into this in the book, but couldn't do it. They built a system for turning a relatively normal, intelligent person into a really good decision maker. They turned that into a process of scientific administration and planning. And they then steamroll it everybody. And I work at the joint staff, right, or I work with the joint staff, sorry, advised the joint staff, which is fundamentally a copy of that original model in many ways. right we can enhance our ability to make decisions and make decisions better and using our new knowledge of the brain i think will actually help us do that more effectively it's been done in the past and we can do it again

you know the brain evolved the human brain evolved over a long period of time in response to all kinds of different pressures and survival imperatives but war as we experience it in the last couple let's say thousand years of human history is a much more complex enterprise than what primitive man would have experienced surviving on the savannah hundreds of thousands of years ago and that this introduces a kind of unusual set of tensions and situations where actually our survival instincts are not particularly well positioned to help us in various strategic situations and things get a bit scrambled Is that fair? And talk about the kinds of pressures that made the human brain what it is and why it prioritizes the things that it prioritizes.

Yeah. So, I mean, so the first thing to say is that that is a reasonably accurate characterization. So, I mean, my reason, I'm not an evolutionary scientist. I'm a brain scientist. And yes, we can frame it in terms of, you know, throughout human history, humans and prehistory humans faced very, very life-threatening emergencies, violence, starvation, and so on. And that's why every brain is built to survive

or hopefully win a fight, right? That is true, but the book is really about the brains that we have now and how all the different brain systems we have. So I start down at the brain stem and end at the bottom of the brain, an end at the frontal pole, which is the most sophisticated brain region.

It's the most distinctively human brain region. I go through many different parts of the brain, many different brain systems, and each of those brain systems can help drive us towards war. Now, there's no one simple reason why humans are driven to war. There are many reasons, many ingredients that arise from, you know, the different brain systems we have. So for your question of how do these things, how do these tensions work? So you think about fear, right? So we need fear.

Fear is useful, right? You need to have fear. So, and that is something that you might say, well, you know, we have these primitive fear systems from when we're on the Africans of Anna or whatever. And now we're living in, you're living in, I don't know what's behind you. Can I see the Washington Monument behind you? Yes, that's why you know, you're living in modern Washington. But the thing is, is that actually we all need these fear systems, right? And you need them now.

And you certainly need them in some parts of your life, as I've read your biography, right? And what happens if people who live in our society don't have the, like a key. part of the brain that's important for fear. So there are people who can lose a part of the brain called the amygdala, which is a small part of the brain that's very important for seeing fear in others and feeling fear yourself. And people who don't have an amygdala, it's very rare, but it can happen that people lose their amygdala. And then some of those people can lose the ability to see fear in others and feel fear themselves. And if they're living in environments are a little bit dangerous, even a little bit dangerous like LA, there's a famous lady who lives in LA and is studied at Caltech.

And she doesn't, she's not, losing the ability to feel fear doesn't make her a fearless criminal who can take whatever she wants without, you know, worrying about the consequences. What it actually means is that she becomes a victim of crime. So I would just reframe it slightly with what you were saying is that actually all of our different brain systems are useful and they remain useful right now, even the ones that can seem more primitive, right? Yes, once you get to our most sophisticated things like our frontal pole that helps us do thinking about our own. own thinking, or what can be called metacognition. Once you get up there, yes, you can see how we can now make that far more sophisticated. We can learn from those who've gone before us, and we can be far more sophisticated than it could be on the African-Savannan. But still, the fundamental machinery remains the same, and all of it's useful. Every part of our brain's orchestra

of brain systems is useful for providing a symphony with which we all live our lives. Let me try to be a little bit more clear about what I'm trying to drive at. So, you have any number of great historical vignettes in the book that tie back to various elements of brain science and just the brain itself. One of them is the fall of France in 1940, where you discuss a kind of paralysis setting in amongst French decision makers as a consequence of the shock or the surprise that they suffered as things did not proceed as they imagined. They would, given the balance of power, like the literal balance of forces between the French and the Germans, which were more or less at parity, you could argue even advantages for the French, depending on how you ran the numbers,

and yet here they are. The whole thing's just collapsing before them. And this leads to a kind of paralysis. It stems from elements of the brain that presumably, I guess the point I'm trying to make is in some tactical evolutionary manner or some evolutionary manner relevant in a tactical context for an individual or small groups, that paralysis presumably serves some sort of survival purpose. But as a function of decision-making at a high level proves catastrophic in the circumstances that come to pass in 1940. Is that fair? Just tell that story of 1940 from your perspective. So I think, I mean, you are right.

So, yeah, as you say, on May the 9th, 1940, there were two armies facing each other and the Allies at that point. It was Britain and France, you know, a number of historians, and this May, for example, you know, said that the Allies had, the defenders of France, had more trained men, more guns, more tanks, more planes. And they weren't worse.

they were good. Some of them were even better than the Germans. But the Germans won. And as you say, one of the key things they did is is a Blitzkrieg, what became subsequently in his Blitzkrieg, their ability to act incredibly fast, to think and make decisions much faster than the enemy, to really cut through the enemy, to surprise the enemy, caused, as you describe, a real paralysis. Right. So again, Ernest May, he says, you know, if there's one feature that explains the fall of France so quickly, and it's literally within seven, weeks, Hitler's cavalry was clattering on the Schozelese in a victory parade. It was this, it was this, it was the surprise. And why is surprise so important? The reason why surprise is so important is because surprise is an absolutely fundamental feature of how brains work. We need surprise. So we use models of the world to understand the world, right? We have to use models of the world. And by a model here, I mean a process that links senses to actions to help us achieve our goals. That's kind of what a of the world is, and it can be sophisticated, like the one I said, you know, with the map of the world in it, and there are all sorts of models we have of the world. How do we update our models

of the world? How do we change our models of the world to keep them consistent with reality? The way we do that is where our models are constantly making predictions. They're constantly predicting how the world is going to work. And when those predictions are wrong, when there is an error in those predictions, then that prediction error, that learning signal, can help us update our models. So this is just fundamental across, you know, essentially the whole range of our brain systems, this prediction error. And so when an event happens that is very surprising, it can cause a huge psychological impact. And you have to write, we need that to learn. And I think you're right, But I would actually say, I mean, it didn't do, and there just is, is that when you have too much of it, it can cause, you know, a shock and paralysis can prevent you from, from responding appropriately.

Your idea that it matters, it's more useful at the tactical level. Thank you for, no one's asked me that question before. But I would say, I mean, he didn't do the French troops, because one of the challenges was, is that the French troops who were defending the Ardenne were, were some of the poorly trained French troops. And the French did not train during the period of war, came under as a phony war. They did not train anywhere near as hard as the Germans. And they were the less well,

you know, less effective French troops, the less well trained French troops. I don't think it did them a lot of good, actually, that they were paralyzed by the surprise of the, of the just sheer speed of the German Blitz Creek, any more than it necessarily helped those who, for example, French generals who literally burst into tears. Let me clarify.

Sorry, I'm using tactical a bit loosely. I don't mean French troops in 1940 at the tactical level. I mean our distant ancestors who are dealing with dangerous animals or other small bands of people. And presumably, I'm sorry, I'm totally out on a limb here. This is just me kind of making stuff up. So you're going to tell us what the truth is. You know, the famous fight or flight reflex, right? You either stand in fight, you run away.

And if neither is possible in the shock of the moment, there is this tendency to freeze. Yeah, which I'm speculating, has some sort of evolutionary purpose that amongst, if you're going to get murdered, if you fight and you can't run away, playing dead or just being stock still, probably you're, I mean, again, it's some very basic. Rodents show freezing. Yeah, absolutely correct. And so that it has some basis in the human survival instinct. But here you're seeing it play out at a sort of strategic into grand strategic level and deeply, deeply unhelpful,

deeply unhelpful in that level. So it's not just, one thing to hear is it's not just to do with fear. So freezing is kind of a fear-related response, but actually prediction errors, when you get this shock and surprise, that is across many different aspects. That's how essentially all the parts of our brain works.

So it's like the clashing of the symbols that really dominate across the whole of the output of the orchestra. Yeah, I mean, it's a great question. what I think the other thing to remember is is that our systems are we do have very effective we need all of these different brain systems we have very effective brain systems so if we know I'm not going to move to the other things I don't stick with fear for a bit true so for example Vladimir Putin in this pseudo so he's a man who really understands how to use fear and manipulate fear in others right he's literally a former KGB Asian and not a very nice guy by very many different measures and he really understands how to manipulate fear in others he understands fear clearly. And there was a story he's told a number of times apparently and it was in his sort of pseudo-autubography, which is where he was growing up in a dilapidated Leningrad apartment block

and he chased rats with sticks. And one day he cornered the rat. And then the rat basically, he got so close to the rat and the rat then turned around and attacked him. And so he said, and then the rat was chasing me. So another response to fear response, there are many things that can happen. You can withdraw your foot from a nail when you step on it, right? That's super quick. You can, if you're a trained person, you might learn the habit of ducking if you hear certain types of noises, if you're in a trench, say, or whatever. You also could have reflexive attacks. So if depending on the distance that the enemy is with you, you might either run or if you've got space to do that, or freeze, or you might attack, right, which is, you know, what Putin was

a listing in that rat. But then also we can link those fear systems, and those fear systems also linked to, in humans in particular, to our amazing abilities to plan forwards, the much more sophisticated regions of prefrontal cortex and so on, which help us plan forwards. And we do. We are incredibly good at planning forwards, planning enormously sophisticated understandings and projects in the world. Now, fear can also be linked to those. So that's what anxiety is, right? You may, I don't know whether you do worry about this, but you may worry about a civil war brewing, or you might worry about climate change, or you might worry about the erosion of your group's status and society,

whatever it is. Fears can operate on that level too, and causes anxiety. Anxiety is just a longer run version of fear, essentially. But the gist of it is right down there, and it actually has many different manifestations, many of them are useful, many of them are not useful. We need to harness them as best we can.

And if we can understand them better and understand ourselves better as humans, then that can help us control those emotions, harness those emotions, and understand them better in others. You know, I'm curious what role you would say, understanding the brain better would play in the context of military training. So much training, I was lucky to participate in some fairly, So what I thought at the time was pretty, and think now it was pretty high quality infantry training. And a lot of it, looking back on it and sort of reflecting on it, considering your work, was about increasing at the margins, but maybe substantially at the margins if you're doing it well, resilience in the face of all kinds of pressures that you speak about in the book.

So, at the very basic level, hunger, thirst, the need to, you know, the way in which your body just doesn't work at a certain point if you don't get enough sleep. just giving the trainees, the personnel experience of hunger, experience of thirst, probably least of all thirst, because that's actually the most dangerous one that'll kill you the fastest. But certainly the hunger and the sleeplessness, pretty extensive experience with hunger and sleeplessness in the training just to kind of give you practice, give the trainee practice, understanding what it feels like to be in that situation so that they're not in that situation for the first time on the battlefield. And then, of course, the emotional dimension to anything. I mean, a lot of the training is stressful and challenging, not just physically, but at kind of an emotional level and almost a moral level as well.

Again, to sort of at the margins inculcate a kind of toughness, you call it courage if you want, you know, these ways in which you're sort of pushing people to operate further out from what their normal limitations would be. If you were talking to the kinds of people who design or execute this training, what would you tell them about the brain that you think they might? find most useful. Yeah, I mean, so I do speak to these people. So I think, I think you're absolutely right. A lot of what you're trying to do, for example, is just to make things less unexpected. So, you know, they've done studies of people jumping out of airplanes. The first time a paratrootub jumps out of an airplane, you know, you get spikes of stress hormones and all the rest of it. But then the second, third, fourth time, it's now better understood by them. It is less unexpected, and they can handle it much better. So a lot of it is exactly as you're saying is about trying to handle things

much better. Now, how can we, so how could we, for example, use that kind of thing now? So one of the, one of the wonderful examples I found for the book was George C. Marshall. So I'm sure you're obviously familiar with George C. Marshall. They're sort of, you know, top American commander during World War II, military officer during World War II and some of it's been set to and so on. Most people don't know about the earlier part of his life about 10 years before when he was in charge of the infantry school at Fort Benning. And he was in charge of the teaching there. And what Marshall did is Marshall revolutionized the teaching at Fort Benning.

And it was really quite like, I think a lot of how well America did during World War II and probably actually a lot of things that you learned in some ways influenced by his teaching, his ideas, what he did, what they called the Benning Revolution. And 200 future generals went through as either instruct or students during his time there. And he basically improved training because he made training effective. And some of the ways he did that, for example, was using, like, a lot of the supervisors are taking people back to the same types of places to do their exercises. So you're saying, no, you need to take people to novel places to do their exercises. There's another crucial thing about,

so making things, trying to get people to, basically try and get more prediction errors, the more signals to learn from, right? You don't want to know, one of his key things was not learning how to do things perfectly, but learning how to do them well enough across a wide range. of different contexts, right? Which is something that's really important for the brain. And I think if what, like, he did lots of other things I won't go into. But the gist is, is that if I were talking to people now, and I do, I'm like, well, what would a George C. Marshall thinking of a, you know, a world war, 10 years in the future, what would

he be doing now 10 years ahead of time, right? And I think what we should be doing is using things like AI augmented reality to create as many different types of simulated environments as possible so that we are really training people across a wide range of unexpected different types of scenarios that are very difficult to do with standard training. So I think, you know, again, we should be thinking what would George, what would a modern day George Marshall be doing? You'd be using the cutting of technologies we have to revolutionise training. And yet, it's not going to be decisive. It's going to be like a small advance, a small enhancement. But it, that, that, that, that, that, that, that,

could be a really important factor in a future conflict. I want to close with a couple of questions about AI. And let me just ask them both. And you take them in whatever order and whatever way you like. The first is, you know, you've laid out for us this, if you'll forgive the expression, that was model of how the brain works. It is model of models and how we work with them. To what extent does AI work the same way?

To what extent is it different? I mean, my sort of deep-rooted suspicion is that it's actually meaningfully different, but I'd like to hear how you think about it. So that's the first question. The second question, more of a moral or political question than a science question, your example about the London cab drivers who do the knowledge. Having their brain meaningfully changed is fascinating. And it's also in a way concerning when we contemplate the way in which AI might affect what it means to be human in the years and generation. to come. You know, technology has always changed what it means to be human. It sort of is what it means to be human is where the species that technology changes and the species that creates

the technology that changes us and that loop is is us. I don't have to explain to you. And so, you know, I'm happy that a long time ago we figured out how to cook with fire and so our brains got big. And if we got a little bit scrawny or we'll so be it. That seems like a bargain worth making. And there have been any number of bargains like that over the years. I guess I'm a little bit worried about outsourcing my cognitive abilities. I don't fully know what that means. I grant that on some level, we've been doing that for a long time now,

so before we could write and take notes easily, we had to remember a lot more. We don't really remember like we used to remember a few hundred years or six, seven hundred years ago. So second question is, what do you think the effect of AI is going to be on us? So I'll take the first one, which is about, well, first one,

first, which was about, you know, how far does AI use these models, right? Or use the similar ways that the brain uses. Now, one, like, slightly boring answer to that is just that there are lots of different ways that AI can work. And so it depends. So for example, and so LLMs, for example, which is obviously a big deal now, they are generative models. And as I said, that is literally like, to some extent how your visual system works. And also just to say that, so Jeffrey Hinton, who was also at Queen Square. So he set up the Gatsby. So he won last year's Nobel Prize.

He was the second, the other guy, not Demis, Jeffrey Hinton, he basically invented these artificial neural networks in key aspects of that. And he literally set up the Gatsby computational neuroscience unit, right, at Queen Square before we went off to Toronto. And, you know, that's where I used to work. And he was deeply involved in how the brain worked. And a lot of his ideas were about trying to build an artificial brain.

I think if we're trying to think about how the brain works, a lot of the stuff I talk about in the book, for example, that has perniated the way that AI works, even if most people even in AI don't really know that. But Jeffrey Hinton, if you listen to him, he's like, yeah, I was trying to build an electronic brain. Alan Turing, you know, who's a famous computer scientist. He actually came to, I keep using,

it may even have been in Washington, but he met with Claude Shannon, who's like one of the most famous information. He was known as the father of information theory. the two of them chatted during World War II about how they could create an electronic brain. They used to chat over lunch whilst they were doing code-breaking type stuff or secure communications. So the ideas really are in there right at the very heart of AI. And LLMs, for example, that a lot of where they work is inspired by these types of things and do that.

But then there are a whole other set of models that were very fashionable before that, or in fact before about 2012, they were the sort of key things, which are more rules-based models. So if a, then B, you know, whatever, it's like rules-based models. Sit there like more. Now, the way we're, in my opinion, the way AI is going to proceed

is we're going to have enormous advances from the LLMs, which are more, like sort of more just less row to compute at it, to computing power. As I said, your brain is really expensive, right? A lot of every sandwich you eat goes to power your brain. Rural compute power is important, and it's the same with AI.

But in addition to that, there are lots of things that where they use models. So an example will be weather forecasting. You can do weather forecasting just using large language model type things that will basically just, you know, are quite naive that it have a model of physics involved. And they can be quite effective. The standard way of doing weather forecasting is building giant models

that use structures or models of the world about how physics works, right? Often the best thing to do is to combine those two things together. which is pretty much how our brain works. So I think all of these types of models I can talk about are going to be important for AI moving forwards because it's going to involve, just as our brain involves multiple types of models working together in an orchestra,

that's how the future of AI will be. Your second question about what's it going to do to us? So everyone's very worried about social media and everyone's, and it's true, if you go on the tube now in London, or I did it in D.C. when I was there a couple of weeks ago, you look at everyone on the metro and everyone, they're not even reading anymore. They're looking at their phones and they're looking at videos or pictures.

It's like we're going into some kind of preliterate society, you know, where we're going to be doing whatever. People worry about that. On the other hand, book sales are actually not going down in America. Hark you and physical books are still there. A lot of people are going to continue doing that. Yes, are people going to be as good now at navigating as they were in the past? No.

Probably if, you know, the balloon went up or whatever, we probably work without GPS if we really needed to. I don't think that young people today are so incapable because they didn't learn how to use a compass like I did when I was a teenager that they could not be able to navigate. And the other thing to say is, is I think, like Henry Kissinger,

wrote a piece in the Atlantic back in the thing as about 2018, worrying about what it would mean for humanity to have computers that know it's better than we know ourselves. And my first thought, and I've stopped with it, and I've not changed my mind on this, is we are incredibly egotcentric beasts. Humans are so egocentric. We think we're amazing.

We will always find ways to think that we are special and different and incredible and wonderful and marvelous and that we have something profound and important to say, as I'm saying to you now with my profound and important thoughts. Right. And so I'm not that worried that AI is going to make us, you know, either inadequate in terms of capabilities or, you know, become, you know,

it will destroy our drive to or our self-belief. Because that's just not how humans are. We are too egocentric and exciting and changeable for all of that. Nicholas Wright, author of Warhead, how the brain shapes war and war shapes the brain. It's been fascinating. This is a bit off-peased, for me at least, but I've learned a lot,

and I'm really grateful to you for joining the show. Thanks so much. It was really fun. Thank you.