298 | Jeff Lichtman on the Wiring Diagram of the Brain

0:00:00.0 Sean Carroll: Hello, everyone. Welcome to the Mindscape Podcast. I'm your host, Sean Carroll.

0:00:03.0 SC: There are about 85, maybe 86 billion neurons in a normal human brain. Meanwhile, there's about 8 billion human beings on Earth, and each neuron in the brain is actually pretty complicated. So if you imagine assigning each human being on Earth the job of understanding 10 neurons in the human brain, we still wouldn't quite have enough work effort to be able to do that, and that wouldn't be nearly enough.

0:00:32.6 SC: Even understanding the neurons is not enough, because the real action happens in how the different neurons are connected to each other. Not all the action, the individual neurons are doing something interesting also, but it's the wiring diagram, the way that all the neurons are talking to each other, where the neuron gets input from, where it sends its output to, sometimes called the connectome of the brain. That appears to be crucially important for understanding how we learn, how we have memories, things like that.

0:01:00.5 SC: So obviously, scientists are hot on the trail of mapping the connectome in the human brain, but you can imagine it is an overwhelmingly large task. We're not quite anywhere close to finishing it yet. We have mapped the connectome of some simple organisms, drosophila, the fruit fly, is the most recent success. Before then we did the C. Elegans, the roundworm and so forth. These are much, much tinier than the human brain, but we're plucky, we're going to try to do it.

0:01:28.6 SC: And one of the leaders in that effort is today's guest, Jeff Lichtman. He's a neuroscientist at Harvard, and one of his pioneering achievements is a way of actually imaging neurons in the brain, called brainbow. This is, of course, a pun on the word rainbow, because what brainbow does is it takes a little slice of the brain. I mean, you have to slice the brain. So you take a dead mouse or something like that, take a little slice of its brain, and then you can use brainbow to light up the different cells in different colors using fluorescent proteins. It's a very non-trivial trick that allows neuroscientists to actually gather this data that tells us, for example, how the wiring diagram is actually wired.

0:02:12.3 SC: And these days we're moving on from fruit flies, mice, et cetera, up to at least little bits of the human brain. There's lots of deep conceptual philosophical questions about the mind, consciousness, agency, things like that, as well as the down-to-earth scientific questions about neurons and how they're wired together. Turns out it's actually a little hard to separate these two categories. Even if you say all you want to do is understand the wiring diagram and how that influences behavior and so on, you run into bigger conceptual questions such as, is it even possible to say that we understand the human brain, in the sense that the brain is trying its best to do very complicated tasks with limited resources it has.

0:03:01.1 SC: So if there were any way to do those tasks much more simply, much more directly, the brain might just do it that way, sort of in some sense, the brain is the simplest thing it can possibly be to do the task that it's assigned to do. And therefore understanding it might not be much more than figuring out that entire wiring diagram, which is going to take way more data than we have stored in every hard drive on Earth right now. But we're working toward it bit by bit. That's how science goes. And what of course you find is that along the way you discover some pretty amazing things about these brains that give us our thoughts and our behavior and who we are. So let's go.

[music]

0:04:03.3 SC: Jeff Lichtman, welcome to the Mindscape Podcast.

0:04:06.4 Jeff Lichtman: Thank you for having me, Sean. Looking forward to it.

0:04:09.8 SC: I hope so. I think, you know, the brain, it's exactly what I point to when I say this is the most complex system that we know about in the universe. This is why I became a physicist because it's way too complex for me. I want to understand very simple things, but give us the super high level picture here. We seem to be making an enormous amount of progress in neuroscience, at least at the experimental level. Are we truly coming closer, you think, to understanding what goes on there in the brain?

0:04:36.6 JL: No.

0:04:38.7 SC: No?

[laughter]

0:04:42.7 SC: You have tenure. You can say that.

0:04:46.9 JL: No, I mean, it's an enormous amount of effort. And I don't want to make fun of this, because this is the field I've been in for my entire professional career, and I'm old. And we have learned an enormous amount. The question is, if you're climbing Mount Everest and you've gone three feet, have you made a lot of progress? You have made an infinite amount of progress relative when you started and you had gone zero. But we are still very far from having a deep understanding of how the brain works. And I will probably say at some point, I'm not sure that is what we should be aiming for anyway.

0:05:27.6 SC: Well, it's good to have lots of unanswered questions. For young people out there it's a vast territory of questions to be asked. But let's just start ourselves, again very simple. So the brain has a bunch of neurons in it. Is the important part of the brain just the neurons? Can we focus on the neurons or do we need to think about other cells as well?

0:05:50.2 JL: Well, the brain has neurons and it has supporting cells, typically called glial cells, G-L-I-A-L, or glia in plural, G-L-I-A. But the neurons are carrying the information. And the brain is basically taking information that you get from your sense organs, your ears, your eyes, your tongue, your nose, and all the receptors on your skin and receptors in your muscle, and then taking that information in. And in a directional pathway, those nerve cells that are responding to the outside world are taking that information in and turning it into an electrical signal that passes from one nerve cell to another, gets highly processed.

0:06:37.3 JL: And ultimately, the nervous system is designed to respond to sensory signals. So if you see a looming eagle and you're a mouse getting bigger and bigger heading towards you, your legs start moving very quickly in a direction that gets the mouse away from the eagle. And that's through a complicated path of interconnected neurons from the sensory side in the eyes and probably the ears of mice, all the way to their foot and arm muscles that make the mouse run very fast.

0:07:14.8 JL: And virtually everything a human does is the same way, sensory in, motor out is basically the way you think about this. It's sensory motor processing, sensations come in, then you do something in the middle, and then out comes a behavior which is a reaction to the sensation. Humans do a little more in that we, because we store information that we have learned in the past, we can ruminate with even in a dark room where nothing is stimulating us. And that can cause us to do something. Like, I'd better get up and pee 'cause I feel like I need to pee in the middle of the night.

0:07:51.1 JL: You know, where you will ruminate or you might say, I forgot to do something, I'm going to go down and wash my underwear so I have clean underwear tomorrow, or something, even though nothing stimulated that other than a thought about something going on in your previous life. So that's the purpose of the brain. It's to take sensory information in and turn it into a motor code, which is a reaction to the sensation.

0:08:19.5 SC: Some of my philosophy friends, not all of them, probably not even most of them, but some of the good ones are skeptical that we can truly understand what a human being is by simply thinking about the mechanistic pieces of information flowing around the brain. We might need something more than just that, more than the underlying physical stuff. I presume that most working neuroscientists are more or less physicalists in this way rather than some kind of non-physicalist, we need more stuff attitude.

0:08:52.5 JL: Yeah. I don't think there's any magic in there other than enormous complexity, but I would say that they're right in my view, and this is not shared, I would say, by every neuroscientist, that if you're trying to understand a human being, it really depends what you mean by the word understand. And you will hear me say many times in our little discussion here, we can describe the brain in enormous detail now, but understanding it is another kettle of fish altogether. And I don't, I'm not sure everybody fully understands what they mean by the word understand.

0:09:31.6 SC: That's why we have philosophy.

0:09:31.7 JL: Because understanding is, I mean, for me, and I think most people understanding means that there's a shorthand, there's a compressed version of some complexity, that once you have the gist of that argument, you don't need the details anymore because now you have it. And I would just pose as an alternative that there are certain things in the world, maybe a brain as an example, where it is the most concise way. There is no simplification. If there were, the brains would've been simpler.

0:10:04.7 SC: Fair enough. Are human brains, do we have more neurons than other animals?

0:10:10.1 JL: We have bigger brains relative to our body size than most animals. So we are encephalized. And that certainly is part of the magic ingredient that makes us smart. But we by far don't have the biggest brains. Elephants and whales have much bigger brains than we do, but they're not smarter than us. But they have a lot of body that they have to move around in the world, and they have a lot of sensations that they have to bring into the brain from their very large bodies. And part of their largeness is to deal with much larger muscles and many more sensory organs.

0:10:48.9 JL: So it's not that alone. It is that in our brain, there is an addition to being big for its size, there's a lot of what we call association cortex, which if you remove, if a patient has a tumor, which is not a good thing to have, but if you have a tumor in association cortex, it can be removed and the patient is still intact. They've lost something, but nobody knows exactly what they've lost. Probably they've lost some code of some of the memories they've had, but memories are coded in so many different ways. Maybe these things can be lost and not noticed. We have more of that kind of association cortex than any other animal.

0:11:27.2 SC: Okay. Do the elephants and whales literally have more neurons? Or are their neurons just bigger?

0:11:34.1 JL: They have more neurons.

0:11:36.8 SC: Okay.

0:11:37.7 JL: They are bigger neurons, but they have more. But even the number of neurons is a little surprising. I think crows, corvids, these corvids, these...

0:11:47.5 SC: Corvids.

0:11:49.9 JL: Very smart birds have actually more neurons than we do.

0:11:54.2 SC: Okay.

0:11:55.5 JL: But they're very small.

0:11:55.6 SC: Yeah.

0:11:55.7 JL: They're really packed in there, so I don't think number... Bigger is not necessarily better. It's necessary but not sufficient for you to be a human being.

0:12:07.8 SC: And remind us the basic story of a neuron. There's some inputs and then there's an output. Some of the details might be fuzzy to me and to the listeners.

0:12:15.0 JL: Yeah. So you can think of a neuron like you can think of a human brain. A human brain, you get input, as I said, from your sense organs, you ruminate on that information, and then you decide to do something. A teacher might ask a little kid in first or second grade, how much is five plus eight? And the students think it through. And if they know the answer, then they send a signal to their deltoideus muscle to put up their arm and oscillate their biceps and their triceps. So that's what the whole brain is doing.

0:12:48.8 JL: And you can think of... In microcosm, that's what every single neuron does. It gets all these inputs. And if they're strong enough, and that might be if enough of them are firing at the same time, or the ones that are firing are especially powerful, that neuron will take that information and decide it is worth sharing with its outgoing partners who are listening to it, and then it'll send that information out to process.

0:13:16.7 JL: So just to make this a little more technical, the inputs which are coming in at synaptic sites are coming into the dendrites of a neuron, which is a local, very complicated branching structure. And that when you normally see a picture of a neuron with all those wires coming out of it, you're looking at the dendrites. If you look carefully at some of those drawings, you'll see one wire is not so branched and it goes a very long distance. And often we don't even know where it goes, 'cause it goes so far. It can go to the other cortex. It can go down the spinal cord. And that's the axon that's sending the information to other cells. So the information comes in on the antennas of the cell called the dendrites, and the output goes from the cell body out the axon, which goes to other cells.

0:14:04.4 SC: So you've, this is great. You've anthropomorphized an individual neuron a little bit, that it gets some inputs and it decides to give an output. How much do we know about exactly, if I give a neuron certain inputs, it will fire the axon?

0:14:19.9 JL: So we know that a cell has a threshold, that is, it has to get a certain strength of input before it pays attention. And it's very similar to a human being. A certain amount of noise in the background in a normal day is not going to get your attention, but a very loud noise will. However, at night when everything is very quiet, even a very small noise will get your attention. And neurons are sort of the same way. They're looking for signals that are above the noise at that particular time. And those signals can be amplified, that the neuron, if it sees that signal and sees its salience and its importance, it'll then send a signal out to all of its target cells that this thing happened to it and you should pay attention.

0:15:09.8 JL: It's sort of like a Twitter or X feed, you get lots of messages in if you, I don't do it, but I'm assuming you get lots of messages in. And if one of those messages is really interesting to you, you then send it on to your followers. And we say, well, that's a metaphor, but it isn't a metaphor. It's actually your brain. It's just a bunch of neurons. That's what's happening. The information is coming in through your sense organs and your neurons notice its salience and then send it out to your fingers so you can send it on to your other brains. So the network of interconnected brains is very much like the network of interconnected neurons in a brain.

0:15:48.9 SC: I know that the connections between the neurons are really like super important, and that's what we'll be spending most of our time talking about. But do individual neurons all have the same algorithm for firing when they get certain inputs? Or are different neurons programmed differently? And can those programs change over time?

0:16:09.3 JL: Well, neurons are definitely programmed differently. There are certain neurons when you excite them, will just respond with a single what's called action potential. That's the electrical signal that leaves the axon and goes down to talk to other cells. There are other neurons when you excite them with input onto their dendrites, they respond with a burst of multiple action potentials. They encode the strength of the signal with the frequency of action potentials. So there's sort of an amplitude input that's being encoded is in the output, in frequency.

0:16:49.4 JL: So FM and AM, if you think of radio. Amplitude modulation on the input side and frequency modulation on the output. And some of those frequency modulations in some cells are quite different than other cells. And it's very complicated, and it's related to the specific kinds of channels in the membrane that allow the ions, the charged ions that change the cells electrical potential, what kinds of ions are allowed through and how linear or non-linear those responses are. It's an extremely non-linear system generally.

0:17:26.3 JL: So it's very hard to say all neurons are the same. It's wrong to say all neurons are the same, but even to model a single neuron, the easiest way to see what a neuron is doing is to record from that neuron while you do step function depolarizations from stimulation and looking how it responds. If you just looked at the channels in the membrane that carry these ions, it would be too complicated actually to model.

0:17:49.8 SC: And we have 86 billion of these in our brains chattering away.

0:17:52.3 JL: That's right.

0:17:52.8 SC: All the time, so...

0:17:54.0 JL: And each of them is receiving input from on the order of 10,000 different neurons. And each of them are talking to 10,000 yet different neurons on the output. It's just, this is why I say describing it may be possible, understanding it, that's another matter.

0:18:12.6 SC: And are the neurons sort of purpose-built, or could the programming inside an individual neuron be reprogrammed over time?

0:18:20.7 JL: Well, certainly we learn. When you ask a child this question and they respond by activating their deltoideus and raising their arm back and forth, that clearly was not a design feature of the nervous system genetically, they didn't understand the language until they learned the language, and until their teacher said, don't shout, raise your hand, enough times that the student learned to raise their hand. So lots of our wiring that allows us for these behaviors in mammals and especially human beings must come from experience, as opposed to from a genetic program. And that is one of the deep, deep mysteries of how experience modifies a wiring diagram.

0:19:05.2 SC: Sure. But I guess my impression, which maybe was wrong, was that learning those behaviors, those responses to stimuli was a matter of the wiring between neurons rather than the programming of individual neurons.

0:19:20.0 JL: It's both.

0:19:20.9 SC: It's both. Okay, got it.

0:19:23.0 JL: Wiring definitely is what we can study now. But there are definitely situations where a neuron's sensitivity seems to be able to change based on experience. And then therefore synapses themselves become more sensitized or depressed based on experience. Even though the wiring hasn't changed, the strength of the connections can modify. So everything. If it's possible to be useful, evolution has taken advantage of it just to, but it wasn't designed to be understood. That's why it doesn't matter how complicated it is. So all it has to do is work.

0:19:58.1 SC: Yeah. You don't, we don't expect someone to go debug the program, right? It's just supposed to do its job. So, okay, let's get into these connections. We have 86 or so billion neurons, like you said, they're connected to thousands each, and the whole shebang, the set of all those connections, that's the connectome, right?

0:20:17.2 JL: That is what we would call the connectome. Yes.

0:20:19.7 SC: And the idea of the connectome, or at least maybe just the label of it, is relatively recent. It's not a hundred years old?

0:20:28.0 JL: No. I mean, it really is a technical challenge to even imagine getting a full description of all the wires. And that has only been possible relatively recently. But the inspiration for all this began with the beginning of neuroscience itself, with the first bona fide neuroscientist, Ramón y Cajal, this professor in Spain who used a stain that was made by another professor, an Italian professor named Camillo Golgi, that stained randomly a very small subset of nerve cells. And that would seem to be a bad feature of a stain that only stains 1% of the cells. But it was random which 1%.

0:21:18.1 JL: But because of that, Cajal could see, Ramón y Cajal is his full last name, could see the exact connectivity of one cell in terms of where its axon was going and what its dendrites looked like. And he found that there was this directional network I mentioned, where the inputs come in on the dendrites and the output of the cell is the axon. And from that, he inferred the brain must be made up of such circuits. And it has just been very difficult to come up with strategies that would allow you to see all of the connections, not just a very small number.

0:21:57.8 JL: And most of my growing up in neuroscience, Cajalian-style drawings of sort of stick figure diagrams of the way nerve cells were connected, which was based on sparse labeling techniques, is how we grew up. And I think many people assume this is the brain until you actually can look at everything. When you look at everything you see, not surprisingly, it's infinitely more complicated than the stick figures.

0:22:27.1 SC: And when was this Cajal?

0:22:30.2 JL: When did we start doing it in different...

0:22:31.5 SC: When did Cajal do his dyeing of the...

0:22:34.1 JL: Oh, Cajal did his work in the 1880s to the 19-teens or so. He got a Nobel prize in 1906, I believe.

0:22:42.4 SC: Oh, one of the early ones. That's great.

0:22:43.9 JL: With Golgi. Yeah.

0:22:45.1 SC: Yeah. And Golgi.

0:22:45.9 JL: One of the early Nobel prizes.

0:22:47.4 SC: So the connectome, then, if we already have these billions of neurons, the connectome is clearly going to be a challenge just to list. Is that how we imagine the challenge? Like at some point we will have mapped the connection between every neuron and every other neuron in the human brain?

0:23:04.1 JL: Yeah. I mean, it's already been done in a small nematode, the roundworm known as Caenorhabditis elegans, which only has 300 nerve cells. And the first attempt to do that took about 10 years with electron microscopy. More recently, people have been doing this in fruit flies, and there was a whole issue of Nature in October that has many articles about the fruit fly connectome. Again, it's minuscule. We've been doing a whole cubic millimeters of human brain, which are bigger than those other two data sets, but it's, a cubic millimeter is like a millionth of a human brain. It's nothing. But that is 1,400 terabytes of data. 1.4 petabytes is the proper terminology for that, it's just insane. We published that just a few months ago.

0:24:05.9 JL: So yeah, I think the trend to do this is now possible, but at the end, what you would have is an enormously complicated wiring diagram, but it would be a digital wiring diagram and it would be one that is amenable to analysis. And for example, our dataset is available online at Google. Google was our colleagues in making this wiring diagram, and all 1.4 petabytes and all of the segmented wires are visible for anyone who wishes to see them.

0:24:41.1 SC: So if drosophila, the fruit fly, that's the largest connectome that we've completely mapped. Is that right? So far?

0:24:47.3 JL: We're working right now, we haven't published it yet, on several zebrafish connectomes. A zebrafish is bigger, about the same number of nerve cells, interestingly, than a fruit fly, but a larval zebrafish is about a couple millimeters long. So it is a bigger, it's got a bigger brain. But, yeah. So, but it, we are in a very steep, climbing slope now towards bigger and bigger animals.

0:25:15.5 SC: So, but does that mean that we understand the fruit fly? Like we could put the fruit fly in a computer and tell you what it's going to do?

0:25:24.8 JL: Yeah, I think the hope is that you would make a digital twin of the wiring diagram and then send in sensory input to the sensory fibers and out you'd get motor behavior. Of course, it's not that simple because of all the other things that we talked about, the strength of synapses, the non-linearities of the response of cells, and most especially the timing of when the different inputs are activating the cell. There are both excitatory inputs and inhibitory inputs plus modulatory neurotransmitter inputs. And all of that is latent but not physical in the wiring diagram. The wiring diagram just shows you where the information could flow, but it doesn't show you how it flows through those wires.

0:26:08.1 SC: Got it. So even in something like C. Elegans where there's only 300 neurons, we know the complete connectome, but there's still a lot more data yet to be collected to truly reproduce or simulate an artificial roundworm.

0:26:21.2 JL: Right. I think that is the hope, and maybe it's realistic there, but it's a very nonlinear system in that animal. It's a highly-evolved animal, you might think 'cause it has 300 nerve cells it's primitive. It is not primitive. It has figured out how with only 300 neurons to have a complete behavioral repertoire that allows it to survive for a billion years, it's an old animal compared to us newbies, us humans, with 300,000 years.

0:26:46.0 SC: It takes me back, because this is episode, more, roughly 300 of the podcast. And one of the early episodes I did was with Coleen Murphy at Princeton, who studies longevity, and she studies C. Elegans, she messes with its DNA to make it live longer. So yeah, that little roundworm is plucky. It's going to teach us a lot, I think.

0:27:05.6 JL: Yeah. It's a fantastic animal. I think one thing that is clear is just looking at its wiring diagram, again, it is way more complicated than it should be, and if a human had to design a 300 nerve cell behavioral network, it wouldn't look like C. Elegans. It has way more wires, and most of them, we don't really know why they're there.

0:27:24.5 SC: Now, for C. Elegans, I'm assuming that, well, I shouldn't assume these things. I should ask questions of the experts that I'm talking to here. Is every single roundworm possessing the exact same connectome? Because I'm guessing that's not true for every single human being.

0:27:42.9 JL: Well, funny you should ask that particular question. In a paper I published in Nature a couple years ago where I was one of the authors, from Mei Zhen's lab and RV Samuels' lab and my lab, we did the wiring diagram of, I think it's 12 animals, and they're isogenic, they all have exactly the same genome, to see how variable the wiring diagram was. And one of the results was that about 40%, 40 percent, of the connections between one nerve cell to another was unique to each animal. And these are animals that are identical. And these are animals that don't learn anything, or much, so clearly they have largely a genetically determined wiring diagram.

0:28:36.1 JL: So there's a lot of interesting variability that doesn't seem to have any purpose, but the connections that are functionally most important often are from nerve cells that are connected by many synapses. And those were in every single animal, they were, the ones that were very variable, were the super weak connections where one nerve cell made one synapse on another one. So this is some background, either noise, or it serves a purpose where it just doesn't matter whether that particular nerve cell is connected to that other nerve cell weakly because it's just part of the background that maybe keeps the cell active, but it doesn't really tell you anything that's important for behavior. Not what we expected, but that's...

0:29:22.0 SC: That's great. No, I mean, it's almost like junk DNA, right? We have non-coding parts of our DNA, we have junk connectome a little bit. [laughter]

0:29:30.5 JL: Right. And we did the same in a mammal looking at the connections to muscle, a particular muscle that is mirror symmetric in the back of the head of a mouse. And again, every wiring diagram was different. But in a, I would say rule-based way, they all had the same range of sizes of axons in terms of the number of muscle fibers they connected to. But the particular place in the muscle where each axon went was variable from one animal to the other. So the system sort of self-organizes in worms and in mammals to work. But there's a certain amount left to, I would say, chance or to things that ultimately don't matter for survival, otherwise they would've been highly constrained, which gives rise to variability.

0:30:23.2 SC: So every listener of this podcast right now, their neurons are very, very slightly being rewired because they're getting some information from our conversation. I mean, it almost makes me wonder like, how do the neurons know how to rewire in such a way as to make this information be stored? But is that something that we have clues about or is that a big open question?

0:30:49.0 JL: Well, I can do this sort of hand-waving about this, but I think about this a lot. I think of a nerve cell as a living creature. It's a single celled organism, like a paramecium or an amoeba, but it's living in a very weird pond, which is your head. It doesn't know that it's inside your head. It doesn't care about whether you're eating a sandwich or listening to a podcast. It's just there. And it has to do certain things to stay alive. And so all the things it does are for its own survival because it's a single celled organism with a will to survive. And those things end up generating learned based wiring diagrams. That from their perspective, they don't know that that's what they're doing. They just know that if they don't do that, they're going to be punished and die.

0:31:42.9 JL: And they're like us. They don't want to die. And that, and why are they like us? Because we are made up of neurons, and that's why we don't want to die, because our nerve cells don't want to die. Why do we eat chocolate chip cookies? Because our neurons want the glucose. We're just the embodiment of all our neurons. So we're just doing what our neurons are saying.

0:32:05.8 SC: I mean, I completely both sympathize and agree with that perspective. I'm wondering how specific we can be about the journey of sound waves impinging upon my eardrums to neurons deep inside my cortex strengthening a synapse here and there.

0:32:23.9 JL: I mean, do we know that, is that what you're saying?

0:32:28.3 SC: Yeah.

0:32:28.6 JL: No, we're inferring that it must be that way. We certainly know in the visual system that in animal models where you record from the visual cortex in the back of the brain, the occipital cortex, while animals are seeing particular spots of light or lines of light, we see the cells respond. We know that the cells are responding to visual information. And in fact, many of the modern artificial neural networks, the convolutional neural nets, are based on the original studies of two Nobel laureates, Hubel and Wiesel, on how visual information of the world is broken into these very small ideas that then are put back together to give us perception. So it's been very influential what is known so far, but it's just the tip of this gigantic iceberg that no one knows how big it is.

0:33:27.2 SC: I would love to dig a little bit more deeply, roll up our sleeves, as it were, into the experimental methodologies here. I mean, you mentioned what was going on in the late 19th century, but if I were to visit Cambridge and walk into your lab, or are you across the river? Are you in Cambridge or are you in...

0:33:42.7 JL: I'm in Cambridge.

0:33:43.9 SC: You're in Cambridge. What's going on there? What are the most important techniques you're using to look at neurons and their connections?

0:33:53.4 JL: A few bit years back, maybe 15 years back, we thought maybe the solution to the Cajal-Golgi stain, which was sparse, was to do exactly the same thing, but label every cell but a different color. So instead of having a big brown mass where you can't trace any wires, if every cell is a different color, but looks like the Golgi stain, then you should be able to trace every cell, even though they're densely labeled. And we called this brainbow, and it's really been tremendously useful as screensavers and things like that, [laughter] posters.

0:34:28.3 JL: But, and for certain parts of the nervous system where I work in the peripheral nervous system it's fantastic and very informative. But in the brain itself, the wires are so densely-packed that this fluorescence-based approach, at least with standard diffraction, limited fluorescence microscopy, doesn't have the resolution to see the wires. So we delved into a different approach, which was serial-section electron microscopy. People, I'll explain what that means in a second. People have been doing serial-section electron microscopy for a long time, but not with the aim of getting wiring diagrams until relatively recently. That's, this is a somewhat newer idea.

0:35:10.6 JL: What serial-sectioning means is that you have a block of brain, at some point, we hope a whole mouse brain, but not quite yet. You have a block of brain and you slice it very, very thin, each slice might be about a thousandth the thickness of a human hair. So I'm talking about really thin, 30 nanometers, or even now with techniques, perhaps 10 nanometers thick slices. And you have, the block of brain is in a heavy resin plastic, and it's been impregnated with heavy metals that electrons respond to, like osmium, lead and, believe it or not, uranium. Not highly radioactive uranium, spent uranium, but it's still heavy metals that have big nuclei filled with lots of protons.

0:36:01.5 JL: An atom has nucleus and electrons, the nucleus has a lot of protons in it, and electrons zipping by are attracted to that nucleus and will change their trajectory. And that's how electron microscope works with these heavy metal stained samples. So we have a very thin section, and osmium especially loves the membranes of cells, the boundaries of cells. And so when electrons get near the boundaries, they're deviated in a way that you see a signal coming from the membranes. You can see the membrane of the nucleus, you can see the membrane of the mitochondria, you can see the membranes around synaptic vesicles, and you can also see the membrane around the whole cell. And they're beautiful. Electron microscopy is quite beautiful.

0:36:49.4 JL: And you take a, in image of one of these 30-nanometer sections, you see where everything is, but you're just seeing a thin slice through this big bowl of spaghetti, with meatballs being maybe the cell bodies and the spaghetti being all the wires. And so all you're seeing is just a bunch of cut little wires. You can't trace them 'cause it's 30 nanometers thick. But the next section, which is from the same brain, is just 30 nanometers deeper into the same volume. And so the meatball might get a little bit bigger, and a piece of spaghetti, if it's going straight in, will be in the same place. But if the spaghetti is running diagonally, you'll be in a slightly different place than the next one.

0:37:32.5 JL: And you just do that over and over again for thousands of sections. And then you trace out, you can do it by hand. And then once you do it by hand, you can just color in the same object from section to section to section and have a three-dimensional version of it. And then you can train a classifier. So machine learning can take over for humans. And that's just what we, with the help of Google, do.

0:37:56.8 SC: So this is implicit in what you just said, but that 30-nanometer slice is smaller than the size of the cell.

0:38:04.1 JL: It's smaller than the size of a synaptic vesicle.

0:38:06.8 SC: Right.

0:38:09.9 JL: Yeah. The cell... The cell is about 30 microns, the cell body.

0:38:12.0 SC: Got it.

0:38:13.6 JL: Thirty nanometers is a thousandth of that.

0:38:18.2 SC: This is a very sharp knife that you must use to slice...

0:38:20.5 JL: It is. It is a diamond knife.

0:38:22.3 SC: Right through the cells.

0:38:23.9 JL: It is a diamond knife. It's the only type of knife that can do this, is a diamond knife, which is harder than osmium, and lead, and the resin, and so... But even those knives get junked up, so after about 3000 sections or so, we have to change the knife, and the knives are expensive and we can re-sharpen these knives, but this is not... It's not for the faint of heart. It's not like you just start and you just say, Okay, fine, we're all done, you just have to do this over and over and over again, and then you have this data set where you've taken pictures, and a single section, if it's large enough, can be half a terabyte of data, one section. And if it was a whole mouse, it would be a terabyte of data per section, and you have to do this all over and over and over again. You can imagine it adds up, a terabyte here, a terabyte there. Eventually you get to petabytes or even exabytes.

0:39:18.6 SC: Pretty soon you're talking real memory. Yes, absolutely. And I presume that much of this is automated or is this just some very plucky set of graduate students?

0:39:28.7 JL: No, no, no. Well, it was... It began... It was very pained graduate students and everyone putting in shifts day and night to make this work. Now it's automated, but things break so often that it's still... There's a lot of manual interruptions, but we have a very, very fast electron microscope now that Carl Zeiss made that has multiple beams scanning different regions of the sample at the same time, which saves us a lot of time, but it's very expensive to have these. These are multi-million dollar microscopes, and they didn't exist before, and they were built expressly for this purpose, and we have the first one of these devices in our lab. Now they're selling them, they're not going off the shelves at a crazy speed, at 5 million a pop, but there are probably 10 or 15 of them in the world.

0:40:29.9 SC: And my impression was that slightly earlier times, people were doing kind of a coarse-grained wiring diagram thing. They would look at little bits of brain which were larger than cells and ask how they were wired together, but now we're clearly right in at the individual cell level.

0:40:48.5 JL: Yeah, there has always been possibilities to map groups of cells, something we call projectomics as opposed to connectomics, where you look at the projections of a whole group of cells that run as a fascicle from one place to another. That's been known for a long time, but the individual connectivity, how one cell sends its branches of its axon to other places, and how all the axons that are talking to that cell's dendrites are arranged, that's connectomics. There's no other way.

0:41:23.4 SC: Okay. And which is the bigger interest right now, like to do something like a mouse where maybe you could do the whole thing, or is it to just whack away at the human brain, even though that's a larger term project, but we'll get there eventually?

0:41:39.6 JL: Now, we are definitely thinking mouse is next. We're funded. There's a consortium of labs that are working together to do a proof of principle of doing a whole mouse brain, which will be about 1000 petabytes or a million terabytes, an exabyte, to do. So we're not ready to do a whole mouse brain yet, but we're building the tools now over this five-year period that we're in right now to show that if you just scale up what's being done now, this group of laboratories, again, with the assistance of Google, we could probably start a whole mouse in about five years and maybe finish a whole mouse, at least the imaging of a whole mouse, within about five years.

0:42:27.1 SC: When you have that much data, we're talking petabytes, exabytes, new Greek prefixes, I'm intimidated by imagining what to do with it. Just sifting through that data sounds hard. I'm sure the machine learning helps, etcetera, etcetera, but this must be a huge part of the research program trying to figure out what you could actually extract from all those bits and bites on your hard drives.

0:42:50.9 JL: Yeah, I'm less worried about this than some. I think when the human genome was being postulated, there were a lot of naysayers who said they weren't sure it was really worth doing. It wasn't so clear. And in fact, genomics, the tracing of, the mapping of the genes has come a long way from its beginnings, but most of its uses were not anticipated at the time it was generated, so I'm not too worried that people won't find a use for it. If you talk to any neuroscientists and say, are you interested in neural circuits, virtually everyone will say, yes, I am, but of course, I don't have access to them, so I work at a higher level.

0:43:40.9 JL: But if you had access to them, you would use them, and the proof is in the pudding, that people who work in C. Elegans and people who work in drosophila, where there are now wiring diagrams, they use it all the time. It is just fundamental to everyone. It's transformed those fields. And I'd say that means something, if you did it in a mouse, since there are a lot more people working on mice than on fruit flies, it would be quite important, I suspect, for many scientists in ways that I can't quite imagine.

0:44:13.2 SC: Yeah, no, I'm very open to that. But when you say it's been transformative for the C. Elegans and for the drosophila, what does it help us do?

0:44:23.8 JL: Yeah, so there are lots of motor programs and sensory programs in these animals where there are... It is a deep mystery, until you have the wiring diagram, how you go from sensation to motor action. And now for the first time, one can have a reasoned rational discussion of how this is embodied in the nervous system, not a theoretical discussion anymore, but one, an empirical discussion based on actual data. And that takes what has been a theoretical subject and now makes it an actual scientific discipline where there's data to deal with.

0:45:07.1 JL: And I think humans love data in the sense that it's easy... People don't normally think of it this way, but it's much easier to generate a hypothesis based on real data than to start with a hypothesis based on a guess, and then try to see if it's true in an animal. That's the way most science is done, this deductive way, where you start with a hypothesis and then you test it, but if you go the other way, you're on safer ground, because whatever you're seeing is really there, and so your theories are going to match what you see in a better way. And I think astronomy works the same way, you have all these theories, but nothing replaces a telescope.

0:45:48.8 SC: Yeah. There's nothing quite like data. But we said earlier that the wiring diagram is only part of the story, the individual neurons are kind of complex in themselves. So is that an ongoing thing parallel to the wiring?

0:46:03.6 JL: Oh, yes. Yes. I think a lot of scientists are working, in fact, this predated connectomics, people working on the role of ion channels and excitability, there are lots of labs that do that. People have recorded with patch clamps and sharp electrodes and extracellular electrodes to look at the firing patterns of nerve cells. There's a lot of information and thousands of papers published every year on those subjects. In fact, so much has been published on that, no human being can accommodate all that information in their minuscule little zettabyte-sized heads.

0:46:45.1 SC: Well, and then speaking of human beings, we do... Again, correct me if I'm wrong, my impression is that we have this brain initiative that is aiming at some day mapping the human wiring diagram. Are you part of that? Is that still going on? I know there were some like funding worries.

0:46:58.8 JL: Yeah, yeah, yeah. I think at the moment, when it comes to mapping the entire wiring diagram at the level of synapses, mice is the pinnacle. I think there are some really complicated ethical issues about how you would get a very fresh human brain to make such a map. I don't know exactly how you would do that without breaking the law, or breaking someone's heart. You'd have to do something terrible. But with a mouse, informed consent is not required. So it's sacrificing one mouse, but that mouse will float a lot of boats later on, but yes, with a mouse one can prepare the animal in a way that there are no artifacts related to post-mortem degradation. And a human, you can't use a human into they're brain dead basically, but by that point, the whole wiring diagram is probably already impacted.

0:48:01.0 SC: So basically, yeah, you're putting it very politely, but once we're dead, our brains kind of start to decay.

0:48:06.9 JL: I think so. I mean, some scientists say you can still get very useful data out of post-mortem brains if they're not too old, but what does not too old mean? We know that after seven minutes without oxygen an adult is pretty much dead in the sense that their brain is not working anymore. But most of the post-mortem brains that are available from cadavers are 20 hours or 40 hours after death, that they're put in fixative for the first time, so it's a little gruesome to imagine you're sitting at the bedside of someone waiting for them to die and at that moment you dunk them in paraformaldehyde and formaldehyde. I just don't want to go there. I just went there, but I don't want to go there.

0:48:55.2 SC: And especially 'cause it's not the individual cells by themselves we care about, it's the connections between them, you can imagine that those would decay very quickly.

0:49:01.0 JL: That's right. There are... Some of the fine stuff will swell and things will just look different and you won't really know whether you're looking at something normal or not.

0:49:10.1 SC: All right, let's put the gruesome reality of the human condition aside and think about the mice for a second, then.

0:49:14.5 JL: And I'm sorry, one more thing, it's a zettabyte. Now, a zettabyte of data, not only... It's hard to imagine, but that's about the digital content of the world in a year. It's just like, nobody really would know how to deal with that now, even if we could do it.

0:49:36.2 SC: It throws a little wet blanket on the idea of simulating a human being in a computer, right?

0:49:43.5 JL: I think even simulating a worm in a computer is a challenge, though, with only 300 cells, so I'm not too worried. I think, as I said, the best we may be able to do is describe in complete detail what's there, but it's a big difference from saying, I now can predict its behavior, I understand it. No human being, honestly, no human being could hold this amount of information. The analogy I give is like, do you understand New York City? And you'd say, that's a stupid question. What do you mean? There's so many things happening at the same time, and there's so much complexity, and I would say if you can't understand New York City, forget about the brain. It's way more complicated than New York City.

0:50:26.2 SC: Well, let's get into exactly this issue of what can and cannot be understood. Is there a hope that along the way as we're measuring the wiring diagram, we discover principles of organization, that obviously the wiring diagram is not just random?

0:50:41.5 JL: Yes. I mean, I personally think that there are probably some rules of connectivity, maybe you would call them motifs or something, that ultimately are useful, but it's sort of like the genome. You have all these genes, every one of them is unique, and they're unique because they evolved for a particular purpose. In our brains, we all have memories that are unique, that evolve related to our particular experiences. I'm not sure there's any way to compress that. Maybe one could ultimately understand how experience generates a wiring diagram that's compatible or is an embodiment of that experience, and once you had it for one example, you could then generalize that idea, but we are very, very far from having that deep insight.

0:51:39.1 JL: That is a genuinely hard problem, something I think a lot about, these problems that despite humans thinking about them for a long time, we still have not gotten even to first base. And it is how you can take an experience and turn it into something that's like a reflex that came about through, in a reflex's case, came about through genetic mechanisms that build a brain with a particular wiring diagram that can build a nest or an animal can understand a particular sound of another bird as being a conspecific.

0:52:12.3 JL: I don't understand how that's done, but at least you can imagine all the pathway from the genes to the nervous system. But how do you turn an experience into a wiring diagram that is also stable and is learned and it's unique to you because you learned this fact or this memory that no one else has but you? That's just a really hard question.

0:52:34.0 SC: Well, it's at least bumping up against what philosophers call the hard problem of consciousness, which I'm sure you're familiar with.

0:52:41.0 JL: I'd never heard it called the hard problem. I know it's a hard problem, I often say I'm not sure what people mean by consciousness, and again, I'm a scientist more than a philosopher, but I would just ask the question, if a living organism responds to its environment and does something in response to something in its environment, some stimulus, is that thing conscious?

0:53:13.7 SC: Sure.

0:53:14.5 JL: And I would say yes, and if it's conscious then every cell is conscious.

0:53:20.4 SC: Yeah.

0:53:22.0 JL: If every cell is conscious, then what exactly are we talking about when we're talking about consciousness? And so I think some people have this idea of consciousness being sort of a running commentary on what's going on using language. Then of course, only humans have that, 'cause we're the only animals with language, but when I look at my dog who wants to go out and he's just staring at me, I have a feeling he's aware of what he's doing and what he wants for me, he's conscious. And I think when I look at an amoeba, and you put light in part of the amoeba's pond on a slide and you make it very hot, and the amoeba moves away from the heat, it's a conscious, it's aware of what's doing. So, I'm not actually sure what the word means, honestly, I'm not trying to be facetious here, I'm deeply not sure as this being a big, profound problem for philosophers, why?

0:54:15.9 JL: Every living thing is conscious, I would say. And I know philosophers say, you don't know what you're talking about, 'cause that's not what consciousness means, but I think it's a slippery slope between what we have and what an amoeba has.

0:54:29.7 SC: I think there's definitely a progression there, a spectrum, right, in between, but the phrase in philosophy circles, the hard problem of consciousness, is actually supposed to be somewhat tongue-in-cheek in contrast with the easy problem of consciousness. The easy problem of consciousness is supposed to be, how do sensory inputs give rise to motor reactions and behavior. That's the easy problem. The hard problem is supposed to be, how do we get our inner experiences, how do we know what it's like to see the color red or taste something spicy. And the joke is, of course, that the easy problem is very hard and the hard problem is impossible.

0:55:15.2 JL: Yeah. I think part of the problem is the language trap, that we humans are so fixated on describing the world with language that we end up with puzzles that are linguistic problems more than brain problems. I think this is... You know, fundamentally, brains existed before language, they don't care about language. And so I'm not a big fan of a lot of linguistic arguments about this, 'cause they're just sort of self-negating in some cases, you generate paradoxes.

0:55:56.4 JL: A good example of just a paradox is that is a light particle a particle or a wave. It seems like a real problem, like the light particle is just constantly in this problem, in this turmoil. Am I this or am I that? No photon cares, it's not a problem for the photon. It's only a problem for language, for humans.

0:56:20.5 SC: I was teaching my students about this earlier today, so I think this is a good analogy here, but okay, back down to the slightly less philosophical, more nuts and bolts question, though, is it a reasonable aspiration to think that understanding the wiring diagram of the brain will help us understand how we get sensations or how we do cognition, or how various other mental aspects of human beings are related to biochemical things going on in our brains.

0:56:51.4 JL: Yeah, I mean, I think all of the things you just said require a wiring diagram and it will give us some insight into how processing takes place and why, for certain questions, it takes a long time before you get the answer, and for other things, it's almost instantaneous. It's the same nervous system, but certain parts of the network have to work extra hard to answer certain questions. We don't fully understand what it means to think about something before you come up with an answer, but you see it in animals and you see it in people that there are questions you ask where a person has to pause.

0:57:32.6 JL: And is going on here in the brain? Where is the information going, that the pause is necessary? And it may be trying out a whole bunch of different repertoires of the answers to decide which is the most reasonable one, it's maybe searching for the path that makes the most sense among many paths that are weakly activated. I think about these kinds of things all the time. The example I give my students is, think of a man, and then I say, think of a man wearing a hat, think of a man wearing a hat that's a tall hat. Think of a man wearing a hat that's a tall hat that has a beard. And imagine the man that you're thinking of that has these things is himself quite tall.

0:58:19.9 JL: At some point in this, the word Lincoln might pop into your head if you're an American and you know your presidents. And what happened? When I said, think of a man, you probably did activate Lincoln weakly along with 10 million other men. I think of a man with a hat, then you're activating a smaller subset, and then finally you give enough stimulus that finally Lincoln reaches threshold, and that's what happened. I don't know, I mean, it just seems like to me that it is just a bunch of axons that activated your Lincoln circuit and finally brought it to threshold and then Lincoln popped into your head.

0:59:03.6 SC: Well, this is what... This is what I find fascinating, 'cause you just used the phrase Lincoln circuit. You didn't say Lincoln neuron, there's not one neuron in my brain devoted to Abraham Lincoln. We were told a while back about the Jennifer Aniston neuron, but... Do you know about that story? Can you explain what was going on there?

0:59:24.1 JL: Oh, sure, I teach it.

0:59:26.7 SC: Okay, good. Tell us about... 'Cause some of our listeners are young enough to have never heard about the Jennifer Aniston neuron.

0:59:30.8 JL: Yeah, no, it's quite amazing that in human beings undergoing surgery, often for epilepsy, the patient has to remain awake during the surgery in order for the neurosurgeons to know whether the region of brain they took out will stop the seizures. These are patients who have repetitive seizures. And while the patient is awake, other scientists, and maybe the same neurosurgeons, but are stimulating various nerve cells or recording from them, and in this particular patient, they were recording from the part of the brain that is where faces are stored, facial information.

1:00:14.8 JL: And they just flashed in front of a slide projector picture after picture after picture, and there was a cell they were recording from that lit up when Jennifer Aniston's picture showed up. But it was pretty quiet when Halle Berry's picture showed up, and it was very quiet when Brad Pitt and she were together, but when she was by herself, it fired. And they had another cell where Bill Clinton's picture, the president's picture, this is when this was being done, fired it, but also his signature made that cell fire, suggesting that this is a cell that's associated with Bill Clinton and there's a Halle, there's probably a Halle Berry cell, but the cell that was a Jennifer Anniston cell did not respond to the Eiffel Tower or to a spider or to a basketball player, or to Brad Pitt plus her, just her, that was the only thing worked.

1:01:13.2 JL: And this suggests that the brain is encoding these kind of information, but that's not the one cell, it's just part of a circuit of cells that get activated under that particular situation. Yeah.

1:01:28.7 SC: So there's probably plenty of other cells that do the same thing, they were only just following a very tiny number, and it's no one cell that is in charge of Jennifer Aniston.

1:01:38.3 JL: And it may not even be the case that that's the only thing this cell responds to. It's just that network of cells that respond to Jennifer Aniston. When they're all active, they get all the way to the auditory system and ultimately to your tongue and mouth, so you can say, you can hear Jennifer Aniston in your head and you can say Jennifer Aniston, but those same cells, some of them may be part of another network that leads to something else, which they just didn't happen to test 'cause they only had a few minutes of slides to shuffle before they had to move on.

1:02:12.2 SC: That was an informed consent situation there, I presume.

1:02:14.7 JL: Yeah, definitely.

1:02:16.7 SC: And I guess I'm trying to understand whether understanding the wiring diagram will help with questions like this. Will it help us say, Oh, okay, I noticed that this cell lights up when I show you Jennifer Aniston, let's go back to the wiring diagram and make a prediction for what other cells might also light up. Is that kind of a plausible hypothesis?

1:02:38.7 JL: Yes, I think what you would hope eventually is to be able to see the physical embodiment of some kind of sensory input and what it does. It probably won't start with Jennifer Aniston in a human brain, though, I think this will probably be done in worms and in fish and in flies, and then we hope in mice. But it's a very hard problem, and the reason it's a hard problem is that the cell is not only connected to Jennifer Aniston stimuli. Each cell gets 10,000 inputs. And so cells are part of many different circuits in a way that they have to store, the brain has to store everything you know, and whatever, and every behavior you can do, it has to be in there too.

1:03:30.2 JL: And we... For no animal do we have the full description of the behavioral repertoire of an animal, so not surprisingly, the wiring diagram is much more complicated than the one thing you're interested in. The hope is you look in there and you see the wiring diagram for your interesting thing, and that's what Cajal's pictures gave us, this idea that it's a very simple circuit. But that very same set of cells participate in millions of circuits, and that's why there are thousands and thousands of wires, making it very hard to understand.

1:04:00.3 SC: Well, you made a very interesting point right at the beginning about the compressibility of the brain, relating it to understanding. For the last hour, I've been thinking like, what's a good analogy here? So let me run this by you. It's kind of like asking for a summary of the encyclopedia, right, you can't...

1:04:19.4 JL: Or a dictionary.

1:04:20.5 SC: Or a dictionary, right. You need every piece of data to know what's going on and you're suggesting...

1:04:24.3 JL: There's no value saying a dictionary is a dictionary. You need more than that.

1:04:28.3 SC: I mean, certainly, there has to be some simplifications. I carry in my brain primitive models of other people, like this person is very easily annoyed, this person is very easy-going, and that gives me some information about how people behave, it's just a far cry from the 85 billion neurons.

1:04:47.7 JL: Yeah, I think this is the human condition is that we take information in and we collapse it to make conclusions about the world. That is our job, that's how we survive, we don't always make the right conclusion, and we can end up with a fixed false belief, a delusion, and depending on which side you were on the last election, half the country roughly had one delusion and half had another delusion. It's quite amazing that this is the way brains work, but we can make those kinds of stories, but they're stories, they're sort of stories with language, and they are not the brain, they are simplification.

1:05:32.8 JL: If they were the brain, it would be great, but they're not the brain. And I'm not sure that there is a way to encapsulate this into a shorter way. There are ways you can skip a lot of things. You say, this is a... I know this is a dictionary, not only that, it's alphabetical, it's alphabetical, that's a new thing I just learned, and not only that, the last letter is Z and the first letter is A. You can do all these kinds of interesting facts, but the essence of the dictionary is every single word in there, and there is no way to compress that, I don't think.

1:06:12.8 SC: You know, I like to wind up the podcast episodes with something more or less optimistic. What is your... If it is optimistic at all, what is your feeling about what this field is going to look like 10, 50 years from now? What are the big things that we should be looking out for?

1:06:32.1 JL: I think one area where it could really be very useful, this kind of wiring, is to give us our first deep insights into what is different about brains of people who are afflicted with any number of psychiatric or developmental disorders where we don't have very effective treatments and we don't have very good insight into what's wrong. Schizophrenia and autism spectrum disorders come right to mind. These are chronic illnesses, or disorders, I would say, that are not in and of themselves fatal, they're not degenerative, but they end up with a brain whose cognitive approach to the world is quite different from people who don't have these differences, and it raises the possibility that they're mis-wired in some way. But what way? Nobody knows.

1:07:28.9 JL: How would you know? You would have to do connectomics. So maybe animal models of disorders like these will be very informative, and maybe samples of brains from human beings undergoing surgery for something else, just as we did with our epilepsy patient where we got our brain samples may give us some insights into what is different about these divergent brains. And I think that the first step to cure is knowing what's wrong, and knowing what's wrong in the medical model is not the outward symptoms and signs, as doctors say, but the inner, what is the biochemical or cellular abnormalities in the body. And pathology is the field that studies the abnormal tissues of disease. We don't have that for brain diseases by and large, for cognitive brain diseases. This gives us an opportunity to go in that direction. I think that's a potentially revolutionary way of thinking about those diseases.

1:08:32.7 SC: No, that'll be super exciting, indeed. I'm glad that was a good place to end on 'cause that makes me excited for the future. So Jeff Lichtman, thanks very much for being on the Mindscape Podcast.

1:08:42.1 JL: My pleasure. It was fun talking to you. Bye, Sean.

[music]