Loading summary
Transcript51 lines
- [00:00]
Rafael Yuste
Hello everybody.
- [00:01]
Marshall Poe
This is Marshall Po. I'm the founder and editor of the New Books Network. And if you're listening to this, you know that the NBN is the largest academic podcast network in the world. We reach a worldwide audience of 2 million people. You may have a podcast or you may be thinking about starting a podcast. As you probably know, there are challenges basically of two kinds. One is technical. There are things you have to know in order to get your podcast produced and distributed. And the second is, and this is the biggest problem, you need to get an audience. Building an audience in podcasting is the hardest thing to do today. With this in mind, we at the NBM have started a service called NBN Productions. What we do is help you create a podcast, produce your podcast, distribute your podcast, and we host your podcast. Most importantly, what we do is we distribute your podcast to the NBN audience. We've done this many times with many academic podcasts and we would like to help you. If you would be interested in talking to us about how we can help you with your podcast, please contact us. Just go to the front page of the New Books Network and you will see a link to NBN Productions. Click that, fill out the form and we can talk. Welcome to the New Books Network.
- [01:07]
Gregory McNeff
Welcome to the New Books Network. I'm your host, Gregory McNeff and I'm excited to be joined by Rafael Yuste, the author of Lectures in Neuroscience, which was published by Columbia University Press in August of 2023. Raphael is professor of Biological Sciences and the Director of the Neurotechnology center at Columbia University. An expert on the function of the cerebral cortex, he also advocates for human rights protection of brain activity. Rafael is the Chair of the Neurorf foundation and help initiate the US Brain Initiative and the International Brain. Why I selected Lectures in Neuroscience is because it returns neuroscience to first principles, explaining what the brain does and how it does it. With unusual clarity and intellectual breath. The book offers a synthetic concept driven framework that connects neurons, computation, perception and consciousness rather than treating them as isolated topics. Most importantly, it is rare to find a single volume that is both rigorous enough for serious students and accessible for the curious public. Rafa, thank you so much for joining me today to discuss your book.
- [02:09]
Rafael Yuste
Thank you, Greg. Delighted to be here.
- [02:11]
Gregory McNeff
Rafa, why did you write Lectures in neuroscience and who is the target audience?
- [02:17]
Rafael Yuste
So I've been teaching undergraduate at Columbia 29 years and my bread and butter course is neuroscience. Actually it's called Neurobiology two Systems Development and Systems Neuroscience, which is a lecture course over a semester supposed to explain how the brain develops and how it works. And for 29 years I've been with this, or actually for 26 years, I've had the situation in which the textbooks that were available for the undergraduates to use in the course were pushing a view of the brain, which I know it's wrong. So it was very frustrating to teach the subject and tell the students one thing and have to use a book that says another thing. And it got to the point I said, you know what, Rafa? You better just write your own book. And that's why I did. I decided to just take the pen and I wrote this book. It's called Lectures in Neuroscience. And it literally goes through all the lectures that I give in this undergraduate course. The target audience is undergraduates at Columbia, but these are special type of undergraduates. These are undergraduates that have very little background in neuroscience or in, in science in general, because as you probably know, at Columbia University, we make the undergrads take the core curriculum for a year and a half, which means for the first year and a half they're just reading the classics. So by the time they reach us in the third years, they have a very, very thin veneer of, of background in, in basic science. So it's a, it's a pretty approachable book. That's just how I wrote it to, to readers that are intelligent, are interested in the topic, but then they don't necessarily have a strong background. So that's, that was the target audience. And the course, essentially the book covers all the topics of how the brain develops. That part is very briefly covered and then how the brain works. And that's sort of the, the bulk of the book.
- [04:35]
Gregory McNeff
Yeah, no, it does an excellent job. I mean, I am sure has had a, the Tabo Ross Queen slate as much as any of the Columbia undergrad. And I, I managed to pull something and I, I have a lot of respect for the Columbia undergrads. You've got some fascinating insights throughout the book. And there's no way we can cover all of them today, but I do want to hit on a few of them. And the first one is.
- [04:56]
Rafael Yuste
I'm sorry, yeah, Rafa, no, so sorry. So probably the, the audience is waiting to hear. So what was wrong with the other books? And the. And essentially is very simple. So neuroscience, like every, every science, we have an intellectual framework of, with which we approach the problem. And the way neuroscientists have approached the problem of understanding how the brain works, it's by assuming that the unit of the brain is the individual neuron. And this is something that's called the neuron doctrine. It was proposed over a hundred years ago, and it was proposed to be something that was so true that you had to believe it with religious fervor. That's why they called it the doctrine. And this neuron doctrine explains the brain as essentially a machine where all the neurons are doing their job. And if you're a neuron in the middle of the brain, you get inputs from another neuron, you do something to that input and you pass the ball to your, the next one in line. So it's a little bit like a bucket brigade. We call it also an input output machine. And this is the way people normally think about machines. Now you have a machine like a car that's sitting there doing nothing, and you do something to it, you turn it on, and then the machine starts to do it, generates an output that generates a movement or a behavior. So this model of how the brain works, it's written up in every single textbook of neuroscience. They start the first chapter, the first any of the textbook, the neuron doctrine. The brain is composed to neuron. The neuron works, as we call it, the reflex action input output machine. The brain is sitting there doing nothing. It gets an input, let's say sensory stimulation. You see something, you see, I don't know, whatever, food. And then you start salivating. No, that's the output is the saliva or your, or your growling. So, and that view of the brain has been, I would say, not completely useless. It starts a lot of how, how neurons work, but we haven't understood how the brain work. We choose the whole point. And that's where I come up with, well, I don't. Haven't come up with a different theory. This is. People came up with this theory already a long time ago, but now it's starting to become more and more clear that the way the brain works is completely different. It's not an input output machine. It's actually more like its own machine that's always on, that's always working. Sometimes it listens to what comes from the outside, sometimes it doesn't listen. Sometimes it generates behavior, sometimes it doesn't generate behaviors. It's a completely different beast. And the reason that works like that is because the units are not individual neurons. They're groups of neurons that enable brain activity to be endogenously generated. So it's a machine that's endogenously generated, active. It's. It doesn't wait for us. It's actually always On. And that's a completely different view. And this is what we normally call the neural network. This is actually how neural networks work through groups of neurons and they activate each other internally. So that's why. That's the view that I present in this book.
- [08:31]
Gregory McNeff
That is a wonderful explanation, Rafa. And I may have cut you off when you said it's a type of machine in the book. You talk about it maybe being a prediction machine. What type of machine would you characterize it as?
- [08:44]
Rafael Yuste
Yeah, so, exactly. So, so let's say. Let's follow the. The thread. So the brain is always active. It's a machine that's always on, is generating activity. So why is it, oh, it's active. What the hell is it doing? And that comes the prediction part. So the brain is a very smart machine that evolution has designed to predict the future. Yeah. Imagine you want to predict what's going to happen. So how do you do it? Well, you have to have a model of the world, and then you press fast forward to that model. It's like a virtual reality model of the world. You press fast forward like, okay, let me see what is going to happen in the. If I run the model. And this is exactly what the brain is doing. It's building a virtual reality model of model of the world which lives in our heads, in which every single aspect of our reality, it's actually its model. There is an inside group of neurons that's representing that reality. For example, I have a bottle of water here. So I actually has a group of neurons in my brain that represent this bottle of water. And when I think of the water, that lights up. And when I touch the bottle and I see it, that lights up. So. And just like this bottle of water, I have groups of neurons that represent everything you can imagine, everything that you know, actually your whole reality. And then what the brain is doing is constantly making a prediction about what's going to happen in the future. And it's comparing that prediction that's internal with what happens in reality and uses. And if things match perfect, you don't touch the model. But if things are off, then you go and touch the model. So that's what the brain is doing, is constantly retouching, fixing up, retooling the model of the world so that we can make a better prediction for the future.
- [10:52]
Gregory McNeff
Got it. You laid out some great themes there, and I want to ask you about a few of them, including criticism as a need to improve as well as the mathematics behind the brain. You talk about Euclidean geometry and You've been talking about the Fourier theorem, and I absolutely want to ask you about neural assemblies. But before all that, I should say your book, as you note, has 18 chapters. You begin each one with a quote. The chapter on vision is a quote from Kant, who actually figures. Throughout your book, you bring in Kant's insights, and it is. It remains completely unknown to us what the objects may be by themselves. And apart from the receptivity of our senses, we know nothing but our manner of perceiving them. How accurate is our model, or map, maybe, if that's the right word. I know you use that in some context throughout the book, of reality compared to actual reality. I mean, is.
- [11:47]
Marshall Poe
Is.
- [11:47]
Gregory McNeff
Do we live in our brains or do we actually perceive reality as it is?
- [11:51]
Rafael Yuste
Yeah. So I think Khan got it right. Now, Ken is one of the philosophers that first hinted at the possibility that the reality that we live in is not real. It's actually internally generated. It's generated by our brains. So it's the reality that we think we live in. And Kant had other arguments of why he thought like that. But the theory that I'm proposing in this book is exactly the 21st century version of Kant's theory. It's the neuroscientific underpinning of Kant's theory. So why. How good is our model of the world? It's fantastic. In fact, it's so good that we normally cannot tell the difference between reality and our model. It's only when some mischievous philosopher like Kant or mischievous neuroscientists say, wait a minute, there's something that doesn't square here. Kant figured this out because of mathematics. You know, Kant was a mathematician and he knew deeply how mathematics work. And he said, wait a minute. Mathematics is something that is not outside in the. In the world, it's inside. But when we use mathematics and apply it to the world, guess what? It matches. So why the hell does it match? It shouldn't match. Why should something that I've invented in my head? Why should the world care and follow some theorem that I just conjure in my head and can't say, well, because it's not the world. What's out there is actually your head. That's why it matches, because your head matches your head. So why is it so good? Because evolution has spent 750 million years building better and better models of the world. In evolution, animals have built more sophisticated models. Until we reach primates and humans, and we have a killer model up here, we can make incredible predictions about the future compared to other Animals, that's why we are smart. You can think that intelligence is when someone can see things coming before they happen. So when you can see the future, you're smart. So our model is very good because every time evolution has built a new animal or a new brain, it improves on what has happened before because of the natural selection. You need to have species that are better than the other species because those are the ones that survive. So that's why our model is so good. And. And the way we adjust. I didn't mention this, but the way we adjust the reality to our model, or the way we adjust our model, sorry, to the reality, is through our senses, not through our. Our vision, our audition, our touch or smell. And if you analyze scientifically, our vision, for example, is as good as it gets. You cannot build a better light sensor than the human eye. It can detect individual photons. That's the physical limit. You cannot beat physics. That's it. And if you go through every single of our senses, guess what? It's optimized to the physical limit. So that when you first realize that, you get scared. Like, holy moly, what a job has evolution done. Why is that? Because it wants to ensure that our model of the world is perfectly updated with what happens up there. And again, it's so good that 99% of the people never realize that they don't live in the real world. They live. They think they're living the real world, but they actually live in their head. But occasionally there's a philosopher like Kant or neuroscientist or psychophysicist who dig deeper and pull the rock. Like, wait a minute, there's something here that just doesn't square.
- [15:50]
Gregory McNeff
Absolutely fascinating. Again, you hit on some great themes. I want to just follow up on the math you note. Some characterize the brain as a giant Bayesian machine, and then also say evolution uses Euclidean geometry and our brain internalizes Euclidean geometry. There's another instance of where you refer how the evolution may have anticipated Fournier's theorem. I was surprised. How often. Often math, geometry, Cartesian planes, figures. Into this book, could you briefly talk about how the brain incorporates math?
- [16:25]
Rafael Yuste
Yeah. Well, so let's take, for example, Bayesian statistics. So imagine that you want to predict the future. Yeah. Mathematically, the best way that we've discovered to predict the future is by using a type of math called Bayesian math, statistics, probability, which essentially takes into account what has happened in the past and uses the probability of events happening in the past to calculate what will happen in the future. I'm not a mathematician, but when I talk to my colleagues that are mathematicians, they tell me the best mathematics that we have to predict the future is Bayesian guess what. If you look at the nitty gritty of how the brain circuits work, they seem to be doing Bayesian math. If you look at human behavior, not just human behavior, the behavior of any animal, they, we work as Bayesian machines as if we're calculating the Bayesian probabilities. And this is, it's an incredible result. Now you can. So one classical example is the experiments that Pavlov did with dogs. The conditioning of the salivating response to the bell when the bell was associated with food, with the smell of food. And this is the foundations of psychology and physiology. Buffalo dog's response to the bell. Well, if you repeat that experiment and compute the probabilities, if you change the probability that of the animal hears a bell or not while you're training the animal and then you see the behavior of the animal later, it actually tracks perfectly well, the Bayesian probabilities. And so, and it's amazing because obviously dogs, they don't explicitly know math. When we go and want to make an investment, for example, we don't calculate Bayesian probabilities, but our brain behaves as if it's doing that. So this means that this Bayesian probability has been captured by evolution, discovered before Mr. Base discovers the theory of evolution already figured that out, and they implanted that in the brains of humans and other animals. So this is just an example of how the chisel of evolution, generation after generation, again over hundreds of millions of years, is created these formidable machines that have all kinds of algorithms and tricks of which we only know the very beginning. And I'm telling you, one of these algorithms that are up there is Bayesian mathematics, another one, Euclidean geometry. All kinds of things that we have here that we and most of the algorithm we don't know about yet. We haven't discovered them.
- [19:39]
Gregory McNeff
Again, you hit on some great themes. I want to follow up. And just to level set how optimized, how unbelievably incredible our brain is, it has over 100 billion neurons with 100 trillion connections, which is orders of magnitude more than the Internet. And it effectively runs on the energy of the 20 lot light bulb. I think you say some sandwich and water is all you.
- [20:03]
Rafael Yuste
Exactly.
- [20:04]
Gregory McNeff
Okay, I want to go back to this idea of neural assemblies. Why are they so powerful? And we're all familiar with this phrase, neurons that fire together, wire together. Could you talk about that?
- [20:19]
Rafael Yuste
Yeah. Okay, so let's say we want to make a model of the world. So in our head. So that means that in our head, we have to have something that it's a symbol of, something outside, let's say, of this bottle of water. It has to be something in my head. I have to be able to turn on something in my head that symbolizes. That represents this spot of water and the same with every single object in our world so that we can manipulate things mentally. So, for instance, I can say, okay, what happens if I drop, if I push this bottle of water over the edge of the table? So I can actually physically do that, or I can run it in my head like, this is going to fall so I don't have to physically do it. I know. Yeah. Okay. So this is an example of how we use this model of the world to manipulate. To manipulate the world. Now, in order to do that, you need to be able to have switches or some activity in the brain that you can turn on and keep it on in the absence of any outside stimulus. I want to think about this bottle of water without having to look at it. How do you do that? Well, so one way to do it is by connecting neurons together in what we call an ensemble, a group of neurons. This is the same idea as an assembly, but we now use the terminology of ensembles of neurons. And you connect these neurons, let's say 10 neurons, like my 10 fingers, and you connect them with each other so that they activate each other as a group. So that means that I can keep this group of neurons active forever because they're always one neuron is activated in the others. So it's like the ball is rolling between these neurons, and you can turn it on, and you can leave it on as long as you want until you want to turn it off. So this is one way, probably this is the way in which nature has been able to solve the problem of how to create an internal state of activity that's independent of the world through these neuronal ensembles. So these neural ensembles. So let's say, I mean, we're starting to. To get to know them. We found that they're there in. In all the brains we've looked at. We found that they are used to represent the object, just like Khan had predicted. Can predicted that we were going to have what he called representations. No, Doc Stelligan, of something in our minds that would represent things in the world. So we found these things, we call them ensembles. We don't know how many we have. Let's say the human brain may have maybe a million of these ensembles, just to give you a number. And then the way our brain works, the way we think, is by essentially like a theater of these ensembles turning on and off in particular ways. And we're starting to learn how this is happening. But this is essentially, if you look at it as a computer science, like what Justa is talking about is a neural network, actually a particular type of neural network called a Hofeld network. And John Hoffield, who got the Nobel Prize two years ago for his pioneering work in neural networks, he proposed that if you have circuits that are interconnected, just like these ensembles that I'm talking to you about, if they can excite each other, these neurons in the HOPE field model, and generates states of activity, stable state of activity, which he called attractors, which are identical to these ensembles. They're the mathematical version of our ensembles. And then in his model said, hey, if you build a neural network like that, you can do all kinds of things. You can do memory, you can encode, you can compute things, you can solve optimization problem. It's essentially a general, general computer. But importantly, that has memory. It's a computer with memory. And that's all I'm saying in the book now, that our brain is essentially a computer with memory, a particular type of a computer, a hop field type network. And the way it works is through these ensembles.
- [24:42]
Gregory McNeff
Got it. And I believe at one point in the book you say it's an organic or biological computer. Maybe you say chemical, but is that. Could you be a little. Could you just, I mean, clarify what you mean by a type computer? Because you do at one point contrast the brain with FPGAs. And be honest, I felt like I was reading an architecture for a software program. I mean, hop field networks. And we really haven't talked about AI and deep layer, deep learning networks. But there's. It almost feels like the AI experts are trying to recreate the brain, because I assume it's because it's so efficient at processing. And it seems like they're trying to follow the map of the architecture. But could you. What type of computer is the brain?
- [25:26]
Rafael Yuste
Yeah, so by the way, we use terms like the brain computes, the brain is like computer. And this, I should warn you that this is a metaphor. And in fact, in history, when people talk about the brain, they always compare it to the most sophisticated machine that they had lying around. So, for instance, with the German philosopher Leibniz, he thought a lot about how the mind works, and he was really talking about the brain, and he and back then, the most sophisticated machines that they had lying around were water milkshake, because they were full of this internal working in which you had wheels that would turn and engage other wheels and torques and levers. So they said, well, it's, I mean, if you've seen one of these water mills, it's actually quite sophisticated in terms of how it works. So he said, oh, and he, he write this beautiful paragraph on the monadology where he imagines that he's walking into the brain, into the mind, and it's a gigantic mill with all kinds of things. So this was back then. Then comes the, the steam engine. And people say, oh my God, the brain, some sort of steam engine. And there are people who wrote the 19th century following the analogy, saying the way the brain works is through vapor going up and down and pistons, just like a steam engine. So now, 20th century, what's the most sophisticated machine we have Computers, digital computers, electronic circuits, like, oh, this is it. This is just like the brain is a computer. So we're still using that metaphor from the 20th century comparing the brain to a computer. But it could be completely different. That's what I mean by an organic computer. So computers are work on with digital electronics and they have particular characteristics and they're very powerful, no questions. But they all have limitations. They use a lot of energy, for example. And it is likely, I don't know for sure because we don't know enough about the brain to make the. The last call here. My intuition is that the brain is not a digital computer. It does not work like an electronic circuit. It's more, it's. It's a different beast, that it's an organic machine. You could argue it could be like an organic computer. And the big difference is that in the brain, every part of it, every little corner of every little section neuron in the brain could be computing, let's say the. An amino acid binding or unbinding from a receptor, a lipid in the membrane of a neuron, of course, the synapses, the neurons firing, all of that could be com. Computers on their own, they could be computing things. And that may be why it's so efficient, because we have a machine in which every little corner of it is actually a computer. So it's really a computer of computers. It's a collection of computers. So rather than a digital computer that has one CPU and has a particular algorithm that's trying to run, it's more like a collection of computers. And that may be why it's so powerful why it's so incredible and why it uses so little energy.
- [29:09]
Gregory McNeff
Okay, again, you hit on so many different themes there. I want to follow up on the computer theme and ask you, how is the learning process like FPGAs and specifically the role of pruning?
- [29:22]
Rafael Yuste
Yeah, so, yeah. So this has to do with how the brain develops. And this has been a huge surprise for neuroscience because as scientists were human, and when people try to understand how the brain develops, say, well, this. But it must be. It must be the same way that we build together a house. We put together the bricks and we build up the different rooms and we put the roof on top. And when people started studying brain development, they came with a gigantic surprise that that's not how the brain gets built. It's more. It gets built the way and sculptor makes a sculpture. So by starting with a big block and taking things off until you end up with the shape that you want. So this would be like building the way you would build a house is by starting with a completely solid block of stuff, and then you excavate and take out all the. The stuff that you don't want until you end up with a house. So it's backwards. Like, wait a minute. This is completely stupid. It must cost so much effort and time to do it like that. Why doesn't nature build things the easy way? No. So this is not how the brain. This is now how nature did it does it by. And this is where the idea of pruning comes in. It first builds a gigantic neural network, and then it takes out all the neurons that it doesn't need and all the connections that it doesn't need. So it's hard to believe, but little kids have a much bigger brain in terms of neurons or connections that we have. And as we go through development, particularly through puberty, well, first there's a huge massive cell, the in which we probably lose about half of the neurons. And that happens even while we're still in utero or right after birth. But then there's a massive pruning of connections, which happens mostly during puberty, which we eliminate. We still don't know, but maybe about half of the connections of our brain, they disappear. So that's what I mean by the sculpture, that nature is sort of working backwards from a large neural network, taking out the things that it doesn't need. And guess what? Human engineers have invented a way to do that too. And these are these special type of chips where use. Let's say that you try to make a chip, and for A particular job. So then if you know the job, you design the logic in the chip and you just build it ready to go. But imagine that you don't know exactly what the job is and say, well, so we're not going to find out what the job is until we use the device in this particle electronics in the field. And engineers, human engineers, have solved this problem with one particular design, which is called field programmable gate arrays. And this is a long name, but what it means is essentially an electronic circuit in which has all the possible connections. And then you go in, you figure out what you have to do, and you eliminate all the ones that you don't want. So that's why I talk about the brain and the cortex in particular. Sensors, biological, a pga. Not that it's as if nature is shipping us out into the, into the world with a big brain chip that can do all kinds of things. In principle, you could use the speak Chinese or speak English. You end up being born in the U.S. well, you don't need to speak Chinese. There goes those connections. So of course, that becomes a problem if you want to learn Chinese when you're older because you got rid of those connections already. But that's essentially the idea behind how the brain develops.
- [33:43]
Gregory McNeff
Okay, again, some great themes there. Plasticity, the critical development period and learning in adulthood, and possibly the link to mental diseases. So I want to follow up there, but I want to ask you about this quote in the wiring of the brain. You write the second wiring diagram involved, starting with the distributed connectivity. That is where a neuron is indiscriminately connected to many other neurons with no specificity at the beginning. But then the connections get pruned up and down through learning and experience. So you end up with a selective neural circuit. In other words, random goes in, then learning, then outcomes, precision. Is that effectively what you're saying? We start with everything and then we prune it down in our early environment?
- [34:27]
Rafael Yuste
I don't want to say, well, it depends. So it's a whole process, a cascade of burning. And that probably starts in utero, even. And the first whack is getting rid of neurons. And we don't know exactly what percentage of neurons we lose. But in some parts of the brain, like in the spinal cord, it's estimated that we lose half of the neurons during late gestation, early postnatal development. But this is just the beginning of a gigantic pruning job. And they're probably not just one single process, but many different mechanisms that are going to start taking out Neurons that are not needed and connections. Once the neurons, once you, you reduce the number of neurons, then you work on the connection, the connections. Connections come later and there again we don't know the number but. And it could be many processes. And this happens during what's called a critical period of development. So this is a period in the life of an animal which is a postnatal period during which whatever happens to the animal changes the connectivity of particular parts of the brain. And by changing, I mean it actually turns out things and then it crystallizes the connectivity into a circuit that is then used for the rest of the life of the animal. And the critical periods occur early for certain things like walking, for example, seeing vision in the sequel period. And then those happen in the early development, the first maybe five years of the life of the, of the person. Then come the currigore periods for speech and language. And those are a little bit later. So normally by the end of the first decade you finish with your critical period for language. And as you know painfully well, if you try to learn a language after the critical period, it's extremely hard compared to learning a language during your critical period. And then comes the later period. Critical period have to do with more sophisticated things like socializing, behaving in, in a group, human relations. And those come later. So. And that has to do with the frontal cortex. So that has the, the, the later acal period probably in the early adolescence. That's when it happens. Adolescent. And all of that ends with puberty. Pretty much after puberty, that's it. You're from the point of view of nature, you're good to go. There's no more tweaks. Go out and reproduce. That's the whole point. Yeah.
- [37:17]
Gregory McNeff
Sounds like college is, I guess college is still relevant then, right? When you say adolescence, they're still able to learn into the early 20s.
- [37:26]
Rafael Yuste
Is that. Yeah, later adolescent. But it's starting to become harder and harder because again the, the, the way Kahal used to talk about the concrete of the brand, it's settling.
- [37:41]
Shopify Advertiser
When it's time to scale your business, it's time for Shopify. Get everything you need to grow the way you want. Like all the way. Stack more sales with the best converting checkout on the planet. Track your cha chings from every channel right in one spot. And turn real time reporting into big time opportunities. Take your business to a whole new level. Switch to Shopify. Start your free trial today.
- [38:10]
Rafael Yuste
USAA knows dynamic duos can save the day. Like superheroes and sidekicks or auto and home insurance. With usaa, you can bundle your auto and home and save up to 10%. Tap the banner to learn more and get a'@usaa.com bundle restrictions apply.
- [38:25]
Gregory McNeff
Now I absolutely want to ask you about Cajal, but I want to follow up with this because on the one hand you suggest why shouldn't our brains continue to be, quote, plastic during adulthood? And on the other hand you say as a species we are notoriously adept at learning even after adulthood, but there may be some tie in with mental diseases. Could you talk about that?
- [38:46]
Rafael Yuste
Yes. So actually my thesis adviser, Thorsten Bissell, discovered pretty gold period. He got the Nobel Prize and he was studying vision. And one of the earliest questions is why does the group be close? Why cannot we have a brain that keeps learning, like things like as if we were kids? The. The kid's brain is like a sponge. So why can't we just continue like that our whole lives? It's a huge mystery. We don't know. But for whatever reason, evolution decides that no, no, we're going to. It's like this circuit I was talking about. The FPGA is not. So maybe you need to be really good at what you do, at seeing, at talking. So that's why we're going to print out all the unnecessary stuff, so that you can do one job, but do it really well. So then the questions are, well, what if we could reopen the critical period? What if we could infuse plasticity? Or what if we didn't prune? What if we kept the brain super connected? And, and that's. We don't know why, but there is a link with mental disease. It turns out that the psychosis and particular schizophrenia are associated with deficits in pruning. So one current hypothesis to explain schizophrenia is that these patients have an abnormal period of brain development in which they don't prune enough so their brains remain too connected. And because of that, this is all a hypothesis. You can imagine that they live in their heads because the connectivity internal dominates versus the external influence. So their model of the world is essentially through tree running is completely perfect. It works well, but it's not connected to the upside. And that's what happens with many of these patients when you talk to them. And as a doctor, I had a lot of early experience with schizophrenics. I said, well, these people have a completely rational view of the world, except it's not the same world that I live in. They have their own world and that generates all the problems. So, so then so it's possible that we have to prune so that we're not, we don't end up being crazy. That our, our models of the world are so powerful that we have, we run the danger and we, I mean humans, as opposed to other animals, humans have a particularly the price that we pay for having such a gigantic model of the world that we can use to calculate the future and think about what's going to happen after climate change, for example. We worry about things that are, we're not even going to see in our own lives. And the price that we pay is that a certain proportion of. Of of us become psychotic because we're at the border of. Of. We're all at the border of being psychotic. No being having a model of the world that gets disconnected from the world. So that's the, that's, that's one hypothesis to explain schizophrenia. But then comes another question and which is, well, if schizophrenia is so, so bad, why hasn't evolution eliminated it? Isn't the purpose of evolution to make better and better brains so that animals are sharper and sharper and they can reproduce better and they can survive better? So if evolution ends up giving us brain that makes us crazy, isn't evolution going into a dead end? This is not a good species. So why has an evolution eliminated schizophrenia or gave us brains in which we don't have psychosis? And, and that this is all a hypothesis. The idea is that it is good for the species to have people that are a little bit out there. Those are the people that are borderline schizophrenics or people that are schizophrenics are some of the people that push the rest of us into a territory where we would never go? No. And this, there is a strong correlation of madness and creativity. The musicians, painters, Van Gogh, writers, scientists, mathematicians, in fact, explorers, saints. I mean if you look at many of these life. Oh my God, these people are picture perfect schizophrenics. So maybe even though it's bad for them, it's good for all of us. So this is something that it's actually this ha. This has been seen already with some genetic diseases. Falci form anemia as an example, sickle cell anemia in which it's good. I mean it's a terrible disease if you, if you get the double mutation, but if you have only one mutation as opposed to two, then you're protected against malaria. So that's, it's good for the population to have these genes floating around. Because even though once in A while there is a person that gets hit and suffers tremendously and dies. But that's the price that we paid to protect the rest of us that have a single copy and survive Malaria, for example. So it's possible that schizophrenia is something like that. It's a, it's a disease that has, is maintained at the population level because it's good for the population to have. It's terrible for the person, but it's good for the population to have to have that floating around.
- [44:45]
Gregory McNeff
Yeah, it's almost like it insulates us or I guess protects us against the brain devolving or. Can you go a little more into detail what you mean?
- [44:57]
Rafael Yuste
Well, I'm not sure I would say that it insulates or protects us. I think it enables us as a group, as the whole human race to come up with things that otherwise we wouldn't come up with. Let me give you an example. So I don't know. Imagine that Columbus was a little bit out there and he said, oh yeah, I know we have to take a boat because I know there's another island there, the continent, and we're going to reach it in four weeks. And anyone in their same mind like, no, you are going to die, man, you're going to die and I'm not jumping in the ship with you. But he was crazy enough to convince people to give him money to board three ships and he was completely wrong. But he happened to hit land because he thought he was going for to the Indies. And if you do that, if he knew math and did the number like, no, we're not going to reach if that's what the Indies are. It's going to take us not four weeks, it's going to take us 10 months to get there and we will not have water and. But he, he reach this undiscovered continent and sadly, boom, there goes modern history. I mean, it's just a revolution. So people like, like Columbus. I'm not saying that Columbus was good, Frank, I'm just. This is just a thought experiment. Imagine that he was a little bit on the borderline. And so that's why it's good for the group to have people like that.
- [46:34]
Gregory McNeff
Got it almost to push the boundaries or take risks that pay off. Yeah, that makes sense. That is fascinating. Never thought about that way that risks that pay off at a societal level where candidly, if they're wrong, they have to cover the bill.
- [46:49]
Rafael Yuste
Well, it's not just risk in this case. It's obviously the risk of dying. If you don't find Land. But imagine a mathematician, some of the best mathematician in history are schizophrenics. And they can. They've been able to come up with theorems and, and proofs that have advanced science and our knowledge. But because they had a way of thinking which was not the. The same way of thinking as the rest of the group. No.
- [47:17]
Gregory McNeff
Gosh, that's amazing. You can almost argue madness is a form of evolutionary. An evolutionary product or skill set.
- [47:25]
Rafael Yuste
That's what I'm saying. That's exactly what I'm saying. That madness is. Is programmed into the human species because it enables us. Enables our species to. To survive. Imagine that whatever glaciation comes. And there's one of these shamans who probably also a little bit borderline schizophrenic, he said, oh yeah, I had a dream and we have to go south. And guess what that was. Oh, that group of humans survived the glaciation? Yeah.
- [48:00]
Gregory McNeff
No, I mean, going to the moon or Mars or something. I mean, you've got to be mad. But I guess that's ultimately how we grow and thrive. Rafa, I want to ask you. You've talked a lot about the brain and what we know, obviously, and I noticed throughout the book used the word mystery a few times. The function of spines is still a mystery. Garba Ergon interneurons is still a mystery. How neural circuits work or operate are still a mystery. Just ballpark, how much of the brain do you think we really understand?
- [48:32]
Rafael Yuste
Bori? That's an impossible question to answer because we would need to know how much we need to know, and then I'll tell you how much we know. But I can tell you that, for instance, in this hundred years of, of the old model of the neuron theory doctrine, actually, the neuron doctrine, we've learned a lot about how individual neurons work, how neurons look like, and that we're starting to learn how they're connected. So that part I would say we know a lot. I don't know whether we know everything that's important, but would say we're. I'm pretty. We're pretty solid there. But then when you enter into neural circuits, again, we're using a metaphor from electronic circuits, but by neural circuits, I mean how neurons are connected with each other and the kinds of things that they do when they're connected, these neural ensembles, then it's. We're just in the very early days now, we know very little about that. It's almost like uncharted territories. And the reason is because we didn't have the Right. Methods. We had methods to record the activity of neurons one by one, but we didn't. Until very recently, we didn't have methods to look at the activity of groups of neurons. The minute that we developed, invented these methods, and I was actually involved in a lot of this technical development, guess what? We found that neurons do group work in groups. And there's a whole world out there that we never imagined of how neurons interact. So that's where we're at. We're essentially hacking away now at this neural circuit. Very exciting results. Almost every day there's a new discovery, and some of these results are really showing us that, hey, this could be the key level, the key to unlock the secrets of the brain could be this idea of neuronal ensembles and neural circuits. So I think we're getting close to the finish. Or by finish, I don't mean that we'll. We'll understand everything that there is to learn about the brain, but we'll have a general theory. I think just like in, in molecular biology, they were lost in the weeds until they came up with a model of double helix of DNA. I think I would say that neuroscience were still lost in the woods, but that this model of the brain that I discuss in the book, again, which is not my model, it was developed over generations of scientists already started starting a long time ago, almost 80 years ago. This new way of thinking of the brain would be the equivalent to the double helix. And going all the way back to Kant, maybe. Kant was the early proponent of this new double helix of the brain.
- [51:25]
Gregory McNeff
Why does Yaak's monad say repression is the secret of life?
- [51:30]
Rafael Yuste
Yes, well, it's fascinating question. You know, he proposed this for gene regulation and he was right. But these ideas that I've been telling you about how the brain develops resonates very well with this way of thinking, that the way evolution works is not by building things, but by taking things out. It's not by activating things, turning things on, but by removing the brakes. So imagine you're driving a car and you can accelerate by pressing on the gas. Or another way to drive is to always pressing the gas and then accelerate by removing the brake. In fact, that's how. I don't know, I think that's how they, they, they, they in formula one, that's how they start the car. When they say go, what they, they call the accelerator at maximum, they just remove the brake. So this, there are parts of the, of the brain, for instance, the, the basal ganglia. I have a chapter in the book about them, which work like that, as if what the brain is doing, it's always put the brake on and everything, and then it's removing the brakes on the things that. On the behaviors that the animal wants to do. So I don't know why it seems a silly way to do things as a human, but maybe there is. It's more intelligent somehow.
- [53:09]
Gregory McNeff
What is the TV problem in the context of how the brain builds images?
- [53:14]
Rafael Yuste
Yeah. Okay, so we're running out of time, but I'm going to tell you briefly, what is this TV problem? So this is an idea that Francis Crick suggested after discovering the double helix of DNA and, and breaking the. The genetic code with Brenner, he jumped onto neuroscience. I spent the last decades of his life working on trying to understand how the brain works. And the TV problem is the fact. And this was. He was criticizing the neuron doctrine. He was saying, imagine that the brain is like a TV screen and we have 86 billion neurons. Imagine you have a TV screen with 86 billion pixels. And imagine that what the brain is doing, it's just like in a TV screen, the pixels turn on and they build you an image, or they build you be looking at a text, build you a letter, or build you a shape that you recognize. Imagine that that's what the brain is trying to do, work like a tv. So now if that's the way the brain works, then we're never going to understand it by looking at the individual pixels one by one, which is what we've done for a hundred years with the neuron doctrine. We've been recording the activity of individual neurons of the brain, one neuron in one animal, and correlating that activity with the behavior of the animal or with the pathology of the patient, say. But if the whole point of the brain is to build, like an image, to build the correlations of neurons in space and in time, we're just doing the wrong experiment is futile. We will never understand how it works, because the. The image in a TV screen is what scientists, what we call an emerging property, which is a property of a system that has many parts that emerges from the interaction between the elements and by definition, is not present in the individual elements. So if you're looking at the screen right now, you see my face. My face, by definition, is not present in any of the pixels that form my face. You have to look at them all together, and then you see Rafa's face. So this is the TV problem. So Rick said, guys, you've been doing the wrong experiment. Experiment. We're dealing with an emergent system like a TV screen. And you've been trying to watch a movie in a tv, in the TV of the brain, by looking at the pixels one by one, and you'll never get it. And this is why we need methods to look at all the pixels together, all the neurons together. And then you see these neural ensembles. The neural ensemble is an example of something that you will never. You, never. You. You will miss. If you're recording from the neurons one by one, you have to record from many neurons at once to realize that they're firing together.
- [56:25]
Gregory McNeff
Rafa, I have two quick final questions. You suggest the whole point of the nervous system is to implement the concept of the self. What do you mean here?
- [56:35]
Rafael Yuste
What I mean is that in this big, gigantic model of the world in which we have different groups of neurons that are symbolizing this bottle, this room, there's one critical group of neuron that is symbolizing you. You are in there too. Yourself, your consciousness. And that's a key group of neurons, because. Why is it key? Because you need, in that model of the world, to manipulate the world. You need to know who. Who are you? So the whole point of manipulating the world has to do with you. Why do I want to manipulate this. This model if it's not in relation with myself? Yeah. So if you follow that logic, you said, oh, wait a minute. Maybe the entire point of building the model of the world, the first and most important piece in that model is you, is the self. And maybe that's what happened in evolution. It invented the neural network to symbolize things, and the first thing it symbolized is the animal itself. And once it had that, evolution took off. And the animals, that species that have brain, guess what? We became the. We took over the planet. Not just first Cnidarians, bilaterians, and then vertebrates, mammals, and primates.
- [57:59]
Gregory McNeff
Last question. What do you hope readers, and especially young students who are fascinated by the brain, take away from your book?
- [58:07]
Rafael Yuste
Well, I end the book saying that we have a lot of work to do and we need you. So I. I hope that my book will inspire some. Some people, students or don't have to be young. Young or old, doesn't matter. And to get involved in neuroscience, to get involved in science in general, is a. It's an incredible task to be a scientist. The scientists are the people that pull forward humanity. And now we need science. Now more than ever in the world, the world that we live today, we need clear signs to solve all the problems that we're generating with our stupidity. So I hope that this book will serve, will serve to inspire people or maybe a new generation to become scientists, to become neuroscientists, and to help understand how this machine works, this organic computer.
- [59:17]
Gregory McNeff
I have no doubt it will. Rafa, this has been a fascinating conversation, lectures in neuroscience. It's a wonderful tour of an amazing structure. Thank you so much for joining me today to discuss your book.
- [59:28]
Rafael Yuste
Thank you, Greg. Thanks for your great questions. It's been a little fun talking to you. And yeah, looking forward to inspiring more people.
- [59:41]
Gregory McNeff
Sounds great, Rafa.