Noah Hutton

The Quest to Merge Human and Computer

Air Date: March 28, 2022

"In Silico” documentarian Noah Hutton discusses the project to simulate the human brain on a supercomputer.

READ FULL TRANSCRIPT

HEFFNER: I’m Alexander Heffner, your host on The Open Mind. I’m delighted to welcome to our program today, Noah Hutton. He is a writer and director of documentary and narrative films. His newest feature this spring is “In Silico.” Thanks so much for joining me today.

 

HUTTON: Thank you for having me, Alex. Appreciate it.

 

HEFFNER: And you can check it out at insilicofilm dot com. Noah, tell me about the origin of this project. How, and when you started working on it.

 

HUTTON: Well, I studied neuroscience as an undergraduate in college. I didn’t, I wasn’t in film school, but I started making films at that same time. And so as a student of this field of neuroscience, I’d heard about this project that was starting and I was starting to make films. So I kind of put two and two together and decided to make a film about what known then as the Blue Brain Project. And it was a very ambitious attempt announced by, in a Ted Talk in 2009, by its leader, Henry Markram, to simulate the human brain on supercomputers in 10 years. And that was his audacious goal. He threw that out there on the Ted stage. It kind of captured the imagination of the community at the time, it was like the biggest project in the brain science world around then. And, you know, as you can imagine, stirred a bunch of controversy too, because many scientists wouldn’t be so audacious to claim they could do such a thing in 10 years, but he did it. He said he had a way to do it. He said he had a proof of concept. He had the backing of IBM for the machinery. He had the Swiss government to fund him. And so I decided, you know, this is one of the only projects I’ve heard of to put a timeline on full understanding of the human brain. Why not make a film that was sort of calling the bluff on that timeline a little bit, you know, at the time I didn’t think of it that way, but I thought, why not just follow this 10-year project with a 10-year film? So I started to do that. I went there once a year for the last decade.

 

HEFFNER: What an amazing if arduous and complicated task over 10 years. I don’t want you to give it away, the spoiler of whether he succeeded and achieved what he wanted, because I want folks to go and stream this. But can you share the short answer of whether he ultimately succeeded in that decade?

 

HUTTON: Well, I think people could surmise that they probably would’ve heard about it in the news if he had succeeded, because that was really the goal here. He said that he would flip on the simulation, it would speak, it would have languages. It would be conscious. You could teach it, you know new languages, you could dial up a molecule, he told me and change the IQ, dial down a molecule to dial down the IQ, I mean he would that he promised that much control over this sentient being. And you know, I think people can imagine they, if that had come true in a decade, they would hear about it. Now the project has done things. They haven’t done nothing, but they have not necessarily reached that, you know. But I don’t want, I don’t want to give away too much.

 

HEFFNER: Well, we have half hour here, a little under half hour, so you’ll give away something, but not what you refuse to give away. Let me ask you this though. When you first heard of the concept, did you think it was something of a quack? Did you think it was legitimate science? What did you think?

 

HUTTON: I, at the time was, you know, an undergraduate, enamored by the work of this project’s leader, Henry Markram. I have been assigned his papers in classes I was taking in college. He really was a central figure in neuroscience in the nineties, he worked under Nobel prize winner. He published foundational papers that you know, we were all learning about in these classes. So when he pivoted and moved away from the cellular neuroscience that he was doing. He was studying circuits and rat brains, you know, really like they call it wet lab neuroscience, he’s actually studying real cells. And when he moved away from that work and announced the simulation project to take, you know, the data that he had collected and that others had collected and to feed it into computer modeling of the brain and to build up to reverse-engineer, a full model, a full simulation. First, he was going to do, or the rat and the mouse, and then he was going to move to primate and cat brain and then finally the human brain.

 

So when I was looking at that as a 22-year-old, I really was wide eyed. I was. I found this inspiring. I mean, many others did too, but you know, there were also a lot of people in those Ted comments, if I had just scrolled down a little more at the time who were, you know, calling into question whether this was possible. But I wasn’t really, I didn’t really have a mode at that point for taking in the criticism. I really was, you know, full steam ahead. I wanted to see this happen in a decade. I wanted to be the one who captured that with the film.

 

HEFFNER: Well, you mentioned IBM support and it is a natural extension, if far more sophisticated in its final product, a natural extension of Watson, in at least as we conceive of it. So it didn’t seem like impossible, or it would be some feat of a miracle to produce. So, you know, to the extent that that we have learned more about neuroscience and how our brains can be projected in the fashion he was seeking to, if not in the format he was looking for, have there been specific advances you want to identify over this decade?

 

HUTTON: Yeah. Well, the advances of the project have stayed at, have not gone beyond a mouse brain, in terms of what the amount of brain that they’ve simulated yet. So they have yet, to date, to simulate a full mouse brain connected to a bot, let’s say a, a simulation of a mouse avatar, which is what many of the project’s critics have been looking for this whole time. So they’ve not really reached that milestone yet. The, the cognitive capacity, even of a mouse is vastly yet beyond our abilities to simulate in a, you know, biologically realistic organism. So it’s a little bit like Watson and the AI efforts of a company like IBM. And, you know, there are other companies obviously doing that these days. Google is maybe the leader in that these days, in that arena with Deep Mind.

 

But, you know, these are a bit of different trajectories of work. And many critics of AI will point out that many of these results are being fitted to the modalities at hand. So you’re trying to design a system that just can beat people at the game of Go or chess, but, you know, there’s something called general intelligence, which is the sort of holy grail of these systems, which has not been achieved yet. And even language AIs, you know, break down at a certain level. But you know, not to not to diminish the achievement, it’s over the last decade since I began the film, which have been remarkable and people I’m sure have read about them and seen them. But this project was doing something different, which was not to retrofit an artificial intelligence system and to train it on problem sets and to have it beat someone in a certain game, or, you know, like Watson, you know, to be able to be used by companies or business enterprises for different solutions, whatever, it’s not doing any of that. It’s just literally reverse engineering cells in a brain. And I’m saying, and the promise was that if you do reverse engineer the very seat of intelligence from the bottom up, everything else will emerge. And that is the great promise of this bottom-up approach that Henry Markram has been, you know, the sort of spearhead of over the last decade. And, you know, there there’s many people who still believe that that will yield fascinating results and may even transform AI and, and eclipse what we’ve been talking about. But to date, it’s not necessarily had the output that was promised in a time span, not to say it won’t in another couple decades

 

HEFFNER: Would not a fear of this be that if you were simulating the human brain on a supercomputer that you would be in empowering that electronic device to basically combine the inanimate with the animate, you know, with the inanimate, with the human and sort of have a Terminator-like function that can then have, you know, supremacy over us homo sapiens?

 

HUTTON: Yeah, well that, I guess that is a little bit of why this bottom-up approach we’re talking about does have a little bit more of a glimmer of that science fiction fear that we, you know, the Terminator fear, because when you do it this way, when you build it from the bottom-up, and again, we’re, we’re, we’re in dream one land here, because this hasn’t happened yet. It’s not even close to happening. They’re not even at a full mouse brain yet, but imagine that you build it this way bottom up, and if you’re just building the cellular structure of an organism and it does somehow, you know, go above and beyond and achieve some sort of intelligence that is spooky and eerily like our own or maybe surpasses them, yes, it’s a, although you ostensibly do have a window into the interior of that brain that is simulated, many scientists I interviewed for the film also pointed out that we might have a perfectly pristine view of the model in action of a simulated brain doing something, but we might still not even understand how that the cells we see firing are achieving the output that that simulated avatar is, is performing. So the opacity of our, of the brain may not be affected very much, even though we can see it fully, you know. So that I found that to be very interesting, you know, will this actually, would that situation actually lead to a greater understanding of the cellular basis of consciousness let’s say? Or are we sort of creating another type of a black box in very much the same way that many of the most complex AI systems these days are somewhat opaque. So even the people who are programming them, they train them on massive sets of data, but it’s not, it, they can’t… it’s so complex inside the algorithm at a certain point that it’s actually hard to even understand how it arrived at that next move it made in the game of Go, you know. So I, there’s, there’s an interesting conundrum there at the same. And you know, but I would actually say that we’re so far from that and that the line in the sand became very clear to me over the decade of making this film, that, and it’s something I became very invested and interested in the film, which people will see if they check it out, is that, you know, the, to a degree, these are perfect simulations of an organism that is deeply imperfect. We run on mistakes and tiny mistakes in our biology is what got us here through, through evolution, produce the changes, the mutations that led to, you know, so the, the fact that these are ultimately perfect machines trying replicate, I perfection to me is a, perhaps a paradox that will prevent them from ever eclipsing us.

 

HEFFNER: Wow. in this “Archive 81” pandemic era we’re living through, I don’t know, I’m guessing that you’ve seen “Archive 81” or heard about it on Netflix. Maybe not.

 

HUTTON: I have not, no.

 

HEFFNER: It’s funny. When someone’s association with something to me is so “Archive 81” on Netflix, our viewers may be familiar with it. It is kind of dubbed a horror film, but it really is in effect trying to recreate an environment in which, like I said, the inanimate and human collide and you’re basically generating some mysterious form of intelligence: spooky, spiritual. But you’ll have to watch it and then we’ll do another episode. But the point being that I find it interesting you’re saying, you know, that the jury isn’t going to be out for a while. But isn’t that what we said about the pandemic? And on this program, I’m proud to say, although I lament this too, we’ve been asking leading scientists about the potential for the next pandemic, not if, but when, for many years leading up to 2020, I mean it. So, you know, and in the same vein, I don’t know if we’re being naive to think that, that black box isn’t going to be discovered. In other words, you know, somewhere in some lab, be it in the states or in some country around the world, there is a genie that’s going to get out of the bottle and going to basically enable the human brain and technology to collide in some omnipotent yet unseen way, right?

 

HUTTON: I can understand that in terms of the pandemic because in a way in way viruses or, you know, operate at that spooky gap between non-living and living matter. So, you know, in a way it’s almost easier to understand the sort of mechanistic steps that would need to happen for a pandemic, like the one we’re in to happen. But, you know, there’s such, there is such an impasse when it comes to consciousness. There’s such a desire to capture it the genie. And yet, I, I, there there’s been no indication in, in any of the, you know, any of the work I’ve done with, with this film, the scientists I’ve talked to, the reading, I’ve done, that you know, we’re, we’re at any, any degree closer to cracking the mystery of consciousness.

 

So I know that there it’s, it occupies an endless limit of, of science fiction, potential. I mean, there are, there’s this show you’re mentioning. But there are so many of, you know, shows, especially these days, I’m noticing this there’s a glut of, of content out there about like doubles, you know, someone getting cloned, or their consciousness being uploaded into a machine. And then, you know, they put into another body or an avatar or something. This is, is really you know, on people’s minds and certainly on the minds of people who, who make movies and TV shows. But I just don’t, I guess I’m a profound skeptic when it comes to that, that line in the sand about human consciousness. I do think that there are faculties of the brain. You know, the brain, I really look at the brain as a combination of a lot of a modality is not, not one light bulb that goes on or off, but it sort of arises from a bunch of low-level functions, so called you know, that our animal brain up to the highest level, right stuff.

 

But there are functions of the brain that I think will be easily captured by technology. But I think that there will be ultimately a bit of a line in the sand when it comes to the high order consciousness that we prize that we, we, we consider makes us uniquely human and separates us from other organisms.

 

HEFFNER: There are known mechanisms to man and woman to recreate or reproduce that consciousness. I mean, we’ve talked about cloning. It is a real thing that, you know, has happened in animals. You know, so it’s when we talk about the human brain, it’s a distinct category of consciousness. But I think you’re alluding to the consciousness of other species. Obviously, we know, you know, Dolly had a consciousness. We have consciousness too. It’s in a, you know, different orbit. I’m not so sure. So I’m not sure it’s different you know, different enough that we’re unable to access it or, you know, understand it. I think that that remains to be seen, that the question would be, you know, where are we, where do we want that endeavor to go? Right. I mean, but…

 

Absolutely. But, but I would, I would say, you know, we didn’t, we didn’t create Dolly. We didn’t, we didn’t, we, we basically, we cloned an existing organism. It’s sort of the easy shortcut to saying we created a new, you know, a consciousness, right.

 

HEFFNER: Right.

 

HUTTON: We didn’t, we didn’t reverse engineer that animal and put it back together again, we don’t know how to do that,

 

HEFFNER:  Well said.

 

HUTTON: And in fact, and I would actually not, I would, I would, I would not separate humans from other animals. I’m saying that I’m saying that it is a big lift to even do a mouse and have it act like a real biological mouse. I will be blown away when I see that you can, you can get me back on and be like, they did a mouse, like, and, and I’ll, I’ll be the first to fess up with my skepticism. I was wrong, but I think we’re even far away from that.

 

HEFFNER: But, if we’re far away from that, what are we on the precipice of with this?

 

HUTTON: So yeah, here we go. So I think we’re, we’re actually, what’s going on. What we’re doing is we’re creating intelligences for what we humans want the intelligence to do in our world. We we’re not necessarily going to recreate, what’s been the product of millions of years of evolution. And, you know, nor do we necessarily want to though, or need to. I think that what we’re on the precipice of doing is creating a lot of very advanced intelligences to do specific functions in our world. And, you know, people have already seen examples of that with beating people at various games or, or creating, you know, specific intelligences that business uses to either get rid of human workers and make their systems more efficient or, you know, anything. But we’re going to see more and more of that for, for all sorts of business applications. And, you know, the assistance that we have in our daily lives, in our pockets, I think that these specific applications of intelligences to solve human needs is really what’s happening. And it’s less of mastering what we, we think we’re mastering evolution or human, you know, millions of, years of biological evolution. I don’t think that’s, what’s going on.

 

HEFFNER: Noah, do you think that the likelihood of the next stage, is going to come from, you know, the AI as the metaverse, you know, basically some kind of computer designed technology, or do you think it might come from outer space? And I don’t say that glibly, I mean, I think that people underestimated the potential for drastic transformation in their lifetimes on the scale of, on the scale of human catastrophe, that we may not have experienced until this pandemic. So I asked that not tongue in cheek, I asked that quite seriously. There is the notion that our knowledge about consciousness is actually going to come from outer space, not from technology we’re building here.

 

HUTTON: Well if, if we were to get some, some tremendous advance in our knowledge of consciousness, it would have to come from, from outer space to break through the level I think we’re talking about, which is, I’m not sure that our brain, our brains even collectively are advanced enough to understand themselves at the level maybe science fiction thinks that they will soon or already thinks that we, they do. I mean, I’ve heard people talk about, you know, this person has invented a new, a computer chip that’s as powerful as the human brain. We’re constantly hearing the language of biology sucked up by Silicon Valley to explain their latest inventions, as if they’ve cracked neural networks. And, you know, these terms are just thrown around. And even a word like learning. I mean, these systems are learning, but they’re not learning in the way that we learn. So I just, I bristle sometimes at the speed at which we appropriate the language of biology to explore, to sort of hold in, in fascination the, our own, at the tools we’ve created. I think that we’re, we’re creating tools, but I very much would like to separate what millions of years of evolution have created from the tools that we humans create with silicon. I think that they’re, you know, that they’re just serving different purposes and they’re fundamentally different whether, whether a break in that line comes from outer space. Now we’re back in science fiction and mean, yeah. Who knows? I think it, I think, yes, it, it certainly would rock the boat.

 

HEFFNER: Well, in the couple minutes we have left, let me go back to that first question about whether the notion of simulating the human brain on a supercomputer was, and the vision that you set out to possibly document 10-plus years ago, whether that was… it’s clear it wasn’t realistic, but, but was it as authentically the scientific method as you thought it might be, in arriving at answers that have integrity, that have gone through tried and true methods of scientific processes and discovery?

 

HUTTON: No.

 

HEFFNER: Or is it quackery?

 

HUTTON: It wasn’t quackery in the sense that I think people who might be aware of like Theranos and Elizabeth Holmes. People, I think might see this, my film and hear about all of this and might expect that kind of a story here. And I wouldn’t suggest it to be that. I don’t, I don’t think it was that level of quackery. And I don’t think that Henry Markram is that type of figure. He’s a very esteemed scientist who has had a very, you know, legitimate career publishing peer-reviewed work. And I think that what happened here maybe, is that, and this is my opinion from the outside observing this project, is that this project was, felt like a little bit more of a Silicon Valley type startup to disrupt science. So I actually think their energy was “science is too slow.” The peer-review process that, one paper at a time, and Henry was speaking openly about this is just too slow. I mean, it’s like thousands of scientists around the world contributing a little needle in the haystack, a little piece of hay, and we’ll never find the needle, right? So he wanted to put it all together, this like massive simulation of a hay, a hay bale, and say, finally, we’ve united all the data. And we can see what the brain is actually doing. That was the disruption idea. It was like Uber for the brain or something. And so to, you know, to answer your question fundamentally, like, no, that that’s not, they were not trying to do it the normal way. And so, as a result, I did not see the normal processes going on. They did write papers. The papers were in some journals, but they were never embraced by the community in the way that maybe they had hoped. But I think, I think a lot of the critics in the film speak to the ways in which the critics feel this project fell short of proof. And that’s ultimately what a lot of the criticism in the film, which people, you know, can see for themselves, is that the project never demonstrated that this, this I’ll say this, that this is the actual right way to understand the brain. I think they’re still trying to prove to the field, but that’s the case.

 

HEFFNER: And do you think that bypassing those processes hurt them? Not just in the public perception or the perception of their peers in the field, that it actually stymied their ability to get to where they wanted to go or where they still want to go?

 

HUTTON: I don’t think so because they, it would only have hurt them if they, if they lost funding you know, because of it. But they they’ve had very loyal funding from the Swiss government. I’m talking about the core project at the center of this film, Henry Markram’s Blue Brain Project. Now in the film, they got bigger for a second. They became the Human Brain Project, and that project continues. But Henry actually, and I don’t want to give…

 

HEFFNER: In the seconds we have left, does Henry now have a new deadline? Does he say he’s going to do this in the next five years or in the next 10 years?

 

HUTTON: Not, not in the same bold, underlined way. He gave it in the Ted Talk where he said 10 years, I’m going to do it there. He does like to forecast. And there’s a bit of that in the film. But there’s not a, there’s not a new official deadline. They’re now just sort of working year to year.

 

HEFFNER: Has your project had the effect of humbling them to seeing that this may not be possible in their lifetimes or our lifetimes?

 

HUTTON: I don’t know that my project has humbled them.

 

HEFFNER: I mean, we started with the question, did they accomplish it? And you said basically, no.

 

HUTTON: Right.

 

HEFFNER: But you know, to be continued. If there’s one thing that you hope viewers take away from, In Silico, what is it?

 

HUTTON: I would like them to watch this film and next time they hear grand pronouncements maybe in the back of their mind, there’s a little bit more of a critical streak. And I think that it’s healthy. That’s not cynical or nay-sayery. I think actually it’s healthy to, for everyone to think for themselves when they hear people give grand pronouncements and use public funds to do that work. So that that’s what I hope for viewers.

 

HEFFNER: Amen. please visit Noah Hutton’s documentary and find it streaming you can learn more at, Insilicofilm dot com. Noah, I have to thank you for really stimulating and interesting conversation today.

 

HUTTON: Oh, likewise. Thank you so much for having me on. It was really fun.

 

HEFFNER: Please visit The Open Mind website at Thirteen.org/OpenMind to view this program online or to access over 1,500 other interviews. And do check us out on Twitter and Facebook @OpenMindTV for updates on future programming.