This is a clip from a conversation with Paola Arlotta on the Artificial Intelligence podcast from Aug 2019. You can watch the full conversation here: http://bit.ly/2O4ncha or listen to it here: https://lexfridman.com/paola-arlotta/ and subscribe on your podcast app. You can watch other clips on the clips playlist: http://bit.ly/2JYkbfZ or full episodes on the full episode playlist: http://bit.ly/2EcbaKf Please note, I select clips with insights from these much longer conversation with the hope of helping make ideas from the podcast more accessible and discoverable. Ultimately, this podcast is a small side hobby for me with the goal of sharing and discussing ideas. For now, I post a clip every day when there is not a full length video. I did a poll and 92% of people either liked or loved the posting of daily clips, 2% were indifferent, and 6% hated it, some suggesting that I post them on a separate YouTube channel. I hear the 6% and partially agree, so am torn about the whole thing. I tried creating a separate clips channel but the YouTube algorithm makes it very difficult for that channel to grow unless the main channel is already very popular. So for a little while, I'll keep posting clips on the main channel. I ask for your patience and to see these clips as supporting the dissemination of knowledge contained in nuanced discussion. If you enjoy it, consider subscribing, sharing, and commenting.

Full episode: http://bit.ly/2O4ncha
Full episodes playlist: http://bit.ly/2EcbaKf
Clips playlist: http://bit.ly/2JYkbfZ
Podcast website: https://lexfridman.com/ai

Paola Arlotta is a professor of stem cell and regenerative biology at Harvard University. She is interested in understanding the molecular laws that govern the birth, differentiation and assembly of the human brain’s cerebral cortex. She explores the complexity of the brain by studying and engineering elements of how the brain develops.

Subscribe to this YouTube channel or connect on:
- Twitter: https://twitter.com/lexfridman
- LinkedIn: https://www.linkedin.com/in/lexfridman
- Facebook: https://www.facebook.com/lexfridman
- Instagram: https://www.instagram.com/lexfridman
- Medium: https://medium.com/@lexfridman
- Support on Patreon: https://www.patreon.com/lexfridman

Do you think you can maybe talk through the first few months and then on through the first 20 years and then for the rest of their lives, what is the development of the human brain look like?

Yeah, at the beginning, you have to build a brain, right?

What's the very beginning, Which beginning are we talking about?

As the embryo is developing in the womb, in addition to making all of the other tissues of the embryo, the muscle, the heart, the blood, the embryo is also building the brain.

And it builds from a very simple structure called the neural tube, which is basically nothing but a tube of cells that spans sort of the length of the embryo from the head all the way to the tail, let's say, of the embryo.

And then over, in human beings, over many months of gestation from that neural tube, which contains a stem cell like cells of the brain, you will make many, many other building blocks of the brain.

So all of the other cell types, because there are many, many different types of cells in the brain that will form specific structures of the brain.

So you can think about embryonic development of the brain as just the time in which you are making the building blocks, the cells.

Are the stem cells relatively homogeneous, like uniform, or are they all different types?

You start with a more homogeneous, perhaps more multipotent type of stem cell.

Multipotent means that it can, it has the potential to make many, many different types of other cells.

And then with time, these progenitors become more heterogeneous, which means more diverse.

There are gonna be many different types of the stem cells.

And also they will give rise to progeny, to other cells that are not stem cells, that are specific cells of the brain, that are very different from the mother stem cell.

And now you think about this process of making cells from the stem cells over many, many months of development for humans.

And what you're doing, you're building the cells that physically make the brain, and then you arrange them in specific structures that are present in the final brain.

So you can think about the embryonic development of the brain as the time where you're building the bricks, you're putting the bricks together to form buildings, structures, regions of the brain, and where you make the connections between these many different type of cells, especially nerve cells, neurons, right, that transmit action potentials and electricity.

I've heard you also say somewhere, I think, correct me if I'm wrong, that the order of the way this builds matters.

If you are an engineer and you think about development, you can think of it as, well, I could also take all the cells and bring them all together into a brain in the end.

So the cells are made in a very specific order that subserve the final product that you need to get.

And so for example, all of the nerve cells, the neurons are made first, and all of the supportive cells of the neurons, like the glia, is made later.

And there is a reason for that, because they have to assemble together in specific ways.

But you also may say, well, why don't we just put them all together in the end?

It's because as they develop next to each other, they influence their own development.

So it's a different thing for a glia to be made alone in a dish than a glia cell be made in a developing embryo with all these other cells around it that produce all these other signals.

First of all, that's mind blowing, that this development process.

From my perspective in artificial intelligence, you often think of how incredible the final product is, the final product, the brain.

But you're making me realize that the final product is just, the beautiful thing is the actual development process.

Do we know the code that drives that development?

First of all, thank you for saying that it's really the formation of the brain.

It's this incredibly choreographed dance that happens the same way every time each 1 of us builds the brain, right?

And that builds an organ that allows us to do what we're doing today, right?

And this is why developmental neurobiologists never get tired of studying that.

Well, it's millions of years of evolution of really fine tuning gene expression programs that allow certain cells to be made at a certain time and to become a certain cell type, but also mechanical forces of pressure, bending.

This embryo is not just, it will not stay a tube, this brain for very long.

At some point, this tube in the front of the embryo will expand to make the primordium of the brain, right?

Now the forces that control that the cells feel, and this is another beautiful thing, the very force that they feel, which is different from a week before or a week ago, will tell the cell, oh, you're being squished in a certain way, begin to produce these new genes, because now you are at the corner, or you are in a stretch of cells, or whatever it is.

And so that mechanical physical force shapes the fate of the cell as well.

So from my perspective, biology is this incredibly complex mess, gooey mess.

How different is like a computer or any kind of mechanical machine that we humans build and the biological systems.

Have you been, cause you've worked a lot with biological systems.

Are they as much of a mess as it seems from a perspective of an engineer, a mechanical engineer?

Yeah, they are much more prone to taking alternative routes, right?

So, if you, we go back to printing a brain versus developing a brain.

Of course, if you print a brain, given that you start with the same building blocks, the same cells, you could potentially print it the same way every time.

But that final brain may not work the same way as a brain built during development does because the very same building blocks that you're using developed in a completely different environment, right?

Therefore, they're going to be different just by definition.

So if you instead use development to build, let's say, a brain organoid, which maybe we will be talking about in a few minutes.

Yes, so if you use processes of development, then when you watch it, you can see that sometimes things can go wrong in some organoids, and by wrong I mean different 1 organoid from the next.

While if you think about that embryo, it always goes right.

So this development, for as complex as it is, every time a baby is born has, with very few exceptions, the brain is like the next baby.

But it's not the same if you develop it in a dish.

And first of all, we don't even develop a brain, you develop something much simpler in the dish.

But there are more options for building things differently, which really tells you that evolution has played a really tight game here for how in the end the brain is built in vivo.

So just a quick, maybe dumb question, but it seems like this is not, the building process is not a dictatorship.

It seems like there's not a centralized, like high level mechanism that says, okay, this cell built itself the wrong way, I'm gonna kill it.

It seems like there's a really strong distributed mechanism.

And if you think about, for example, different species building their brain, each brain is a little bit different.

So the brain of a lizard is very different from that of a chicken, from that of a, you know, 1 of us, and so on and so forth, and still is a brain, but it was built differently, starting from stem cells that pretty much had the same potential, but in the end, evolution builds different brains in different species, because that serves in a way the purpose of that species and the wellbeing of that organism.

And so there are many possibilities, but then there is a way, and you were talking about a code.

Nobody knows what the entire code of development is.

We know bits and pieces of very specific aspects of development of the brain, what genes are involved to make a certain cell types, how those 2 cells interact to make the next level structure.

That we might know, but the entirety of it, how it's so well controlled, it's really mind blowing.

So in the first 2 months in the embryo or whatever, the first few weeks, months, months.

So yeah, the building blocks are constructed, the actual, the different regions of the brain, I guess, and the nervous system?

Well, this continues way longer than just the first few months.

So over the very first few months, you build a lot of these cells, but then there is continuous building of new cell types all the way through birth.

And then even postnatally, you know, I don't know if you've ever heard of myelin.

Myelin is this sort of insulation that is built around the cables of the neurons so that the electricity can go really fast from...

The axons, they're called axons, exactly.

And so, as human beings, we myelinate ourselves postnatally.

A kid, you know, a six-year-old kid has barely started the process of making the mature oligodendrocytes, which are the cells that then eventually will wrap the axons into myelin.

And this will continue, believe it or not, until we are about, you know, 25, 30 years old.

So there is a continuous process of maturation and tweaking and additions and also in response to what we do.


Speaker 0: Do you think you can maybe talk through the first few months and then on through the first 20 years and then for the rest of their lives, what is the development of the human brain look like? What are the different stages?

Speaker 1: Yeah, at the beginning, you have to build a brain, right? And the brain is made of cells.

Speaker 0: What's the very beginning, Which beginning are we talking about?

Speaker 1: In the embryo. As the embryo is developing in the womb, in addition to making all of the other tissues of the embryo, the muscle, the heart, the blood, the embryo is also building the brain. And it builds from a very simple structure called the neural tube, which is basically nothing but a tube of cells that spans sort of the length of the embryo from the head all the way to the tail, let's say, of the embryo. And then over, in human beings, over many months of gestation from that neural tube, which contains a stem cell like cells of the brain, you will make many, many other building blocks of the brain. So all of the other cell types, because there are many, many different types of cells in the brain that will form specific structures of the brain.

So you can think about embryonic development of the brain as just the time in which you are making the building blocks, the cells.

Speaker 0: Are the stem cells relatively homogeneous, like uniform, or are they all different types? That's a

Speaker 1: very good question. It's exactly how it works. You start with a more homogeneous, perhaps more multipotent type of stem cell.

Speaker 0: What's multipotent?

Speaker 1: Multipotent means that it can, it has the potential to make many, many different types of other cells. And then with time, these progenitors become more heterogeneous, which means more diverse. There are gonna be many different types of the stem cells. And also they will give rise to progeny, to other cells that are not stem cells, that are specific cells of the brain, that are very different from the mother stem cell. And now you think about this process of making cells from the stem cells over many, many months of development for humans.

And what you're doing, you're building the cells that physically make the brain, and then you arrange them in specific structures that are present in the final brain. So you can think about the embryonic development of the brain as the time where you're building the bricks, you're putting the bricks together to form buildings, structures, regions of the brain, and where you make the connections between these many different type of cells, especially nerve cells, neurons, right, that transmit action potentials and electricity.

Speaker 0: I've heard you also say somewhere, I think, correct me if I'm wrong, that the order of the way this builds matters.

Speaker 1: Oh, yes. If you are an engineer and you think about development, you can think of it as, well, I could also take all the cells and bring them all together into a brain in the end. But development is much more than that. So the cells are made in a very specific order that subserve the final product that you need to get. And so for example, all of the nerve cells, the neurons are made first, and all of the supportive cells of the neurons, like the glia, is made later.

And there is a reason for that, because they have to assemble together in specific ways. But you also may say, well, why don't we just put them all together in the end? It's because as they develop next to each other, they influence their own development. So it's a different thing for a glia to be made alone in a dish than a glia cell be made in a developing embryo with all these other cells around it that produce all these other signals.

Speaker 0: First of all, that's mind blowing, that this development process. From my perspective in artificial intelligence, you often think of how incredible the final product is, the final product, the brain. But you're making me realize that the final product is just, the beautiful thing is the actual development process. Do we know the code that drives that development? Do we have any sense?

Speaker 1: First of all, thank you for saying that it's really the formation of the brain. It's really its development. It's this incredibly choreographed dance that happens the same way every time each 1 of us builds the brain, right? And that builds an organ that allows us to do what we're doing today, right? That is mind-blowing.

And this is why developmental neurobiologists never get tired of studying that. Now you're asking about the code. What drives this? How is this done? Well, it's millions of years of evolution of really fine tuning gene expression programs that allow certain cells to be made at a certain time and to become a certain cell type, but also mechanical forces of pressure, bending.

This embryo is not just, it will not stay a tube, this brain for very long. At some point, this tube in the front of the embryo will expand to make the primordium of the brain, right? Now the forces that control that the cells feel, and this is another beautiful thing, the very force that they feel, which is different from a week before or a week ago, will tell the cell, oh, you're being squished in a certain way, begin to produce these new genes, because now you are at the corner, or you are in a stretch of cells, or whatever it is. And so that mechanical physical force shapes the fate of the cell as well.

Speaker 0: So it's

Speaker 1: not only chemical, it's also mechanical.

Speaker 0: Mechanical. So from my perspective, biology is this incredibly complex mess, gooey mess. So you're saying mechanical forces. How different is like a computer or any kind of mechanical machine that we humans build and the biological systems. Have you been, cause you've worked a lot with biological systems.

Are they as much of a mess as it seems from a perspective of an engineer, a mechanical engineer?

Speaker 1: Yeah, they are much more prone to taking alternative routes, right? So, if you, we go back to printing a brain versus developing a brain. Of course, if you print a brain, given that you start with the same building blocks, the same cells, you could potentially print it the same way every time. But that final brain may not work the same way as a brain built during development does because the very same building blocks that you're using developed in a completely different environment, right? It was not the environment of the brain.

Therefore, they're going to be different just by definition. So if you instead use development to build, let's say, a brain organoid, which maybe we will be talking about in a few minutes. For sure,

Speaker 0: those things are fascinating.

Speaker 1: Yes, so if you use processes of development, then when you watch it, you can see that sometimes things can go wrong in some organoids, and by wrong I mean different 1 organoid from the next. While if you think about that embryo, it always goes right. So this development, for as complex as it is, every time a baby is born has, with very few exceptions, the brain is like the next baby. But it's not the same if you develop it in a dish. And first of all, we don't even develop a brain, you develop something much simpler in the dish.

But there are more options for building things differently, which really tells you that evolution has played a really tight game here for how in the end the brain is built in vivo.

Speaker 0: So just a quick, maybe dumb question, but it seems like this is not, the building process is not a dictatorship. It seems like there's not a centralized, like high level mechanism that says, okay, this cell built itself the wrong way, I'm gonna kill it. It seems like there's a really strong distributed mechanism. Is that in your sense?

Speaker 1: There are a lot of possibilities, right? And if you think about, for example, different species building their brain, each brain is a little bit different. So the brain of a lizard is very different from that of a chicken, from that of a, you know, 1 of us, and so on and so forth, and still is a brain, but it was built differently, starting from stem cells that pretty much had the same potential, but in the end, evolution builds different brains in different species, because that serves in a way the purpose of that species and the wellbeing of that organism. And so there are many possibilities, but then there is a way, and you were talking about a code. Nobody knows what the entire code of development is.

Of course we don't. We know bits and pieces of very specific aspects of development of the brain, what genes are involved to make a certain cell types, how those 2 cells interact to make the next level structure. That we might know, but the entirety of it, how it's so well controlled, it's really mind blowing.

Speaker 0: So in the first 2 months in the embryo or whatever, the first few weeks, months, months. So yeah, the building blocks are constructed, the actual, the different regions of the brain, I guess, and the nervous system?

Speaker 1: Well, this continues way longer than just the first few months. So over the very first few months, you build a lot of these cells, but then there is continuous building of new cell types all the way through birth. And then even postnatally, you know, I don't know if you've ever heard of myelin. Myelin is this sort of insulation that is built around the cables of the neurons so that the electricity can go really fast from... The

Speaker 0: axons, I guess they're called.

Speaker 1: The axons, they're called axons, exactly. And so, as human beings, we myelinate ourselves postnatally. A kid, you know, a six-year-old kid has barely started the process of making the mature oligodendrocytes, which are the cells that then eventually will wrap the axons into myelin. And this will continue, believe it or not, until we are about, you know, 25, 30 years old. So there is a continuous process of maturation and tweaking and additions and also in response to what we do.

See all Lex Fridman transcripts on Youtube

Human Brain Development - Paola Arlotta, Professor, Harvard Stem Cell Institute | AI Podcast Clips