Cybornetics: Hugh Herr and Janna Levin in Conversation
A prodigious teenage climber, one of the best in the world, is overwhelmed by a storm after a technically difficult ascent. He loses both of his legs below the knees to frostbite. There are few events that define any of us so starkly. Hugh Herr, seventeen at the time of the amputations, recovers and strives and comes to lead the Biomechatronics group at MIT’s Media Lab where he designs his own legs. He can be shorter or taller. He can climb again. He builds prosthetics that can be controlled by thought alone. Herr’s Biomechatronics group aims to redefine what it means to be human in the wake of cybornetics. Hugh joined me in conversation at Pioneer Works to consider if it is better to be a cyborg than to be an ordinary human. He recounts the harrowing details of his ordeal, his return to climbing, and the potential for individual expression through a human-machine hybrid during what he coins “the age of body alchemy.”
This is a transcript of our conversation, edited for print.
Hugh, so great to have you here. Your work, it's part engineering, obviously, part mechanical engineering, part electrical engineering, part neuroscience, right? Computational science, physics, a lot of physics. And design. Where, as a scientist, do you see yourself? Where do you see yourself as a scientist in that space? Do you feel like, "I'm an engineer with an interest in biodesign." Or...
I don't feel a need to define myself. [laughs]
I tend to be the person that becomes passionate about some problem, and then I just think, "What do I need to learn?" And I drive towards that. The MIT Media Lab is... I would say it's grounded in the idea of creativity. Innovation is spawned from an intense intellectual diversity. So by design, every single research group in the lab, and there's over thirty, is completely distinct from any other group. A lot of the interesting problems and solutions occur at the boundaries between formal disciplines. So by doing that you increase the edges, if you will... between different ways of thinking.
The contact edges. So tell me, what does bio mechatronics mean? Did you make that word up?
I don't remember.
Even better. Do you fight over who first came up with the term?
Yeah, what does it mean? It's a longer, more academic word for bionics. I actually prefer the word bionics.
It's cleaner, more concise. But Hollywood stole it and I'm trying to steal it back.
We're all going to sign a petition for Hugh to get bionics back.
It's merging and integrating the design world with the physiological innate world.
Which brings us to your work in bionics. These are your legs.
Those are my legs.
You have many legs?
I do. Yeah, it gets annoying when people say, "You don't have legs, are you depressed?"
Yeah, you're like, "I have tons of legs." Your closet must be very interesting.
It's very interesting. You have a lot of shoes, I have a lot of legs.
I saw an interview in which you put on one pair of legs to run to the site to climb, you changed your legs to climb, you changed them again coming down.
During the climb, my leg fell off.
When you're climbing and you pull a rock off you yell, "Rock," right? So I actually yelled, "Leg." I nearly killed that poor camera person down below.
But they got an excellent shot. It was the shot of a lifetime. So tell me about these legs, because they're quite fancy.
Very fancy.
They're computer-controlled, of course.
Right before we came on stage, you were pressing buttons and we're like, "What are you doing?"
I was rebooting my leg. I'm basically a power tool. This is the battery. I charge in forty minutes.
And how many microprocessors do you have in those legs?
Three per leg. Mini computers and twelve sensors.
The sensors can tell if your foot is... where it's facing relative to the ground?
Right. Position speeds, accelerations, forces, torques, temperatures.
And it sends all of that to the microprocessors and what do they do with that information?
One thing my lab does is study how the biological body works and then we take principles from that science and that informs the design. I call it stealing from the cookie jar of nature. We studied how the biological complex works. And we discovered, for example, the calf muscle, how it's controlled by spinal circuitry, and there's a... the dominant reflex is a positive force feedback when the foot's in contact with the ground, which means the greater the force is on the muscle-tendon of the calf and the Achilles tendon, the greater is the neural activation of the muscle. So you get this explosive burst of torque and power. On these small microprocessors, we program in that biological-like reflex. So, the synthetic motor is controlled in the same way as the biological tissue.
You're saying biologically, we also have a feedback loop.
Yeah.
And your legs are emulating that to some extent.
Right. The emergent behaviors of positive torque feedback are remarkable. The faster I walk, the greater torque and power I receive. When I go downhill, I don't get net work, as should be the case. When I go uphill, I get a lot of energy output from the devices. And the device has no idea what the world looks like. It has no idea that there's a hill and I could be on carpet or wood, it has no idea of texture. These reflexive behaviors have very powerful emergent behaviors.
Now you designed these yourself.
Plus hundreds of people. I'm just on top of this big rocket.
I've heard you say sometimes you make yourself taller, sometimes you make yourself shorter. You've played around with that. You use the word biomimicry. You talk about emulating nature, but sometimes you also completely defy it.
Yeah, when I can. I can be as short as four feet five or something… I don't know, I'm as tall as… so a funny and actually true story: My first week or two in undergraduate school, I decided to do an experiment. Every day that I went to school, I would increase my height by an inch and I wanted to see how long it would take for someone to notice. And I got really tall. Granted, I was not very popular. I was almost eight feet tall and someone says, "What is going on? You seem to be getting taller." And I set it up. I knew this would happen. I said, "Of course, college is a growing experience."
You were a very prodigious climber, beginning at around seven. You started climbing with your older brothers?
Correct, I'm the baby. Two brothers and two sisters.
You all went out and bought gear and decided to just climb?
It's amazing we're alive. We saw rock climbing on the television and we said, "We need to do that." And then we got a how-to manual, I think from REI or something, and opened it up and realized, "Oh, we need a rope. Okay, let's get a rope. Okay, we need crampons, okay." And we just went out without any instruction whatsoever and we lived to tell the tale.
But my understanding is that you really excelled in particular among your siblings and by seventeen you were an internationally known, prodigious climber. What does a seventeen-year-old kid have to do to become an acknowledged prodigious climber?
By the age of ten, I was climbing things that only adults climb. And then by the age of, I don't know, thirteen, I was climbing rocks that no adult could ascend at the time. So I gained this reputation of being Boy Wonder, child prodigy and-
Did you love it?
I loved it. It's all I did. And because of that, I was a terrible student. My passion was to leave the academic environment and I actually went to vocational high school because it afforded me three hours during the day where I could train and go climbing.
Then there's this really defining moment when you were seventeen and I know you're asked about this all the time, and I hate to ask you again but it is such an incredible story. We all think about those defining moments in our lives. You're seventeen and you go out with a friend in New Hampshire. Who is the friend you went out with? Jeff?
Yeah, yeah, Jeff. This was when I was seventeen. During the winter, I typically would do extreme rock climbing. I was just bored and thought, "Okay, let's go ice climbing." So we were ascending what's called Huntington's Ravine, many of you probably have been up to the whites in a particular gully system called Odell’s. I think, 600 feet of ice, slightly less than vertical.
To entertain myself, I climbed it without protection. We got to the top and we were protected by this gully system and we decided to just walk, three, five minutes above Huntington's towards the summit, and just reassess the conditions. The conditions became horrendous. The winds were very, very high.
I think we went five minutes, and we turned around. We descended rapidly into an adjacent gully system that funneled us to a different side of the mountain than Huntington's.
You were lost? Did you know where you were?
It's more accurate to say we were trapped. Once we hit the tree line again, it was getting dark. The average depth of snow was to the waist, sometimes it was to the chest. The tree limbs came to the edge of the surface of the snow. To make progress, you'd actually have to tunnel through the snow. So, we had to follow the banks of this river and of course, occasionally, we fell into the ice. At one point I was nearly swept under the ice, and Jeff extended his bisax and slowly pulled me out.
We were getting wet, and it was minus 20 degrees Fahrenheit, and frostbite and hypothermia quickly set in.
You were building ice caves to weather the storm?
We would try to get two hours of sleep a day, and during the day, we're probably doing at least one marathon of effort. We slept during the night by... We found very large house-sized granite boulders, and we would dig the snow out from the edge of the granite, where the boulder meets the snow and create a cave where the ceiling of the cave was the rock and the floor was the snow.
That's already pretty industrious under those conditions… For two seventeen year-old kids.
We took tree limbs and whatnot and created a bed to take our bodies off the snow.
You stayed warm together, I think I heard you saying? You would not have survived without each other.
Yeah, it's critical when you're in such conditions to be with another human being. The phenomenon is, you're with someone else and you hug each other, and you increase the power supply of the heat source by two but the surface area doesn't go up by a factor of two. So it's a very effective way to keep your organs functional. But your limbs of course go very quick.
How many days?
It was nearly four. Yeah.
I think you said you weren't aware that it was four days?
Right. With hyperthermia, one does not think clearly. Even though I saw sunlight, nighttime, sunlight, nighttime, if you were to ask me how much time progressed, I would say, "Oh, less than a day.
Which is kind of merciful, actually.
Right.
You get within a few miles of a road, and then you can't walk anymore? Your limbs are severely frostbit and you're frozen.
Yes. When you're frozen from the knee down, there's no sensory information going to the brain regarding balance and whatnot. Your only indication that you're falling is your eyes, I guess your vestibular and they're too slow. So you just fall over, it's the strangest feeling in the world.
Now you're in what you think is the final of your ice caves and at this point are you thinking, "Well that's it. We're not going to make it out of here." Or are you just delirious?
No, we were thinking this was it. And it's very interesting we did not feel sad for ourselves but we imagined our loved ones feeling sad and that made us sad if that makes sense. As an extreme climber, and a very young person, I had a very romantic notion of death. As a teenager, I was free soloing rock faces that are thousands of feet high in just a few hours so I had this fantasy about death and when I was there I was like, "Okay, this is it. Here we go."
You decided to stop hugging each other?
Yeah, we were in a fair bit of pain.
So, we just said, "Let's not hug each other, let's accelerate the death process."
There's this very serious search for you underway, but somebody just happens to stumble on your footprints in the snow, who wasn't part of the actual organized search.
Yeah, someone was out snowshoeing just for fun. She heard about the rescue happening way over at Huntington's and came across what appeared to be human tracks, and said, "I'll follow them." Those tracks led to our cave.
When you are ultimately rescued, you're very severely frostbitten. You sent me this picture of you in the hospital. Look at that beautiful boy of seventeen. You describe the moment they realize they had to amputate your legs. You describe much later with the vision of an adult and your understanding of biomechatronics that the amputation is as crude as it would have been in the Civil War.
Right, since the US Civil War, if you look at anatomical surgical textbooks from the Civil War era, and you compare that to modern textbooks, you'll note that what is being taught in terms of how limbs are amputated, really hasn't changed. So even though we have these extraordinary mechatronics, the way we're handling the tissues, and the surgical procedure is completely outdated. So recently, my lab asked the question, "Could we apply creativity to manipulation of tissues and create a tissue environment that's friendly to electrical neural interfacing as well as mechanical interface."
You've said about yourself to some extent you are bionic, but you said, "I'm not a cyborg yet." Can you explain your distinction there?
The word bionic and the word cyborg are not well defined, but I have my own definition. Bionic is... typically you design something out of synthetic materials. That's through the lens of science. You might learn something about how a foot works, and then apply that principle to designing a synthetic foot. Or you may design a sailboat and you study how whales work, hydrodynamics, so it becomes a bionic boat. Bionics is this interplay between design and human bionics and biology.
Cyborg can be that but not always. Cyborg does not have to emulate nature or biological function. In the cyborg, there's a bidirectional communication between the nervous system and something designed. The biological brain can inform the design part and the design part that has synthetic computation can inform the nervous system. That bidirectionality, when that exists, then it's a cyborg entity.
How quickly after your accident, after you're released from the hospital, let's say, are you designing your own prosthetics?
The mountain climbing accident was in late January. My legs were amputated in March, I believe, and then I went home for a month or so to heal. I was in a wheelchair. Then in about May I went to a rehabilitation center where I was fitted with artificial limbs for the first time. The rule was at this rehab center that during the weekends a patient is allowed to go home. The first weekend, they did not allow me to take my legs home, because I was like, "Can you fit the rock climbing boot on the artificial foot?" The second weekend, they slipped up and they allowed me to take my legs home and I went climbing.
Were they naturalistic prosthetics?
The 1980s, they were wood, foam, rubber.
No microprocessors.
No sensors.
No computational intelligence.
And exceedingly painful as well.
The fit?
Right.
I also have this spectacular picture. That's you. You're back on the mountain.
Right, I'm three meters high. Look, the distance from my hip to my toe. It just ain't right.
So you made these, that you're using to climb?
Right.
And you're just like, "Oh, what do I need naturalistic prosthetics for?" You mentioned that you realized, "Hey, I don't really need a heel for climbing. Heels aren't great for climbing." You've become an expert climber again and an avid climber again.
Right. Yeah, very, very quickly, just a month or two after the surgeries I made the intellectual leap that I'm synthetic from the knee down. It's a blank slate. I can create anything in that space. It can look human or not, and I decided that at that early point to just think about function and not worry about what it looks like. So I asked the question if you were to optimize a foot for the vertical world, what would it look like? And I ended up with baby feet that are at an optimal angle with no heel. Baby feet because you can get the center of mass right against the vertical wall and get a lot of your body weight through the toe.
You also designed specifically for your friend Jim Ewing?
Yes, I've known Jim for thirty-seven years. He's a climber as well. A few years back, he fell rock climbing, he fell fifty feet and hit a very rocky surface. His rope failed to catch him, and he broke many bones in his body, punctured both lungs. One leg in [articular had a lot of pathology, and after healing in rehab, every step he took was extremely painful.
Jim came to me and said, "I can't live like this. I'm in such pain." And he's like, "Does amputation make sense?" His timing was immaculate because we had been inventing this new paradigm for how to amputate limbs in a better way.
It turns out Jim was the first human volunteer for this procedure.
He was the first sort of cyborg where it was going in both directions for your lab?
Yeah. So like myself, Jim wanted to return to climbing so we built him a climbing leg. His has an ankle and a subtalar joint, and it's brain-controlled with afferent neural proprioceptive feedback. When the joint's moving, he can feel within his nervous system its positions and speeds as if it's part of his own body.
You had this really wonderful scene when you were filming him where he got something stuck on the bottom of his shoe right after you had put the prosthetic on and he naturally just went to shake it off his foot. Had you anticipated such naturalness in the first few minutes of being on his leg?
Right. Yes, we've observed and we continue to observe very interesting human-machine behaviors. When you turn the feedback on where the synthetic computation of the limb is inputting information into the nervous system, it seems the person no longer views it as a separate device from their body. They no longer view it as a mere tool. They view it as part of them, and I call that neurological embodiment.
Jim and other people that have received this surgery and neural intervention, they say and do things that are very unusual. As you stated, Jim accidentally stepped on a roll of electrical tape and it's sticky on the side, and it stuck to his bionic sneaker, and he did something that's so natural. You don't reach down like this when something's stuck to your show. You just shake it the hell off, and that's exactly what he did. He had been neurally connected for three hours. Then at the end of the sessions, he sits down and he’s chatting like this, and he's just gesturing with his foot. You don't see that in a prosthesis that's not linked to the brain.
On the one hand, you're thinking about biomimicry. How can we study nature to make these as natural as possible? On the other hand, you're rejecting this dated notion of realistic-looking prostheses, which have none of the realistic motions or experiences, and going to something that's maybe just abandoning the realism of the visual altogether. Will there be a desire to go back to a naturalistic appearance?
No.
Or do you think that that's over? That's just dated.
It's neither. There's no beginning or end. I believe in individual freedom. Something I've done in the past is attempt to design the bionic limb with a biologic shape, but in no way does it look biological. And that's an interesting framework because one day a person may be in such a mood and they want to explore and express human-like beauty in human-like forms. And the very next day, they might want to explore a machine-like beauty, and they can do that by just peeling off a sock or a synthetic skin to expose the machine parts. I think people should just be free to express all forms of beauty.
You have also said that you believe within this century it will be completely natural for most people to have some machine parts, if not required, by choice.
Yes. I believe as we march through time there will be more and more designed content in our bodies. Not only synthetics, with biologics as well. And less and less innate biological material. And what will end up is at a point where every human being has a plethora of ways of augmenting or modifying themselves, like me in a crude sense. Every person will be able to sculpt their cognition, their sensory experience, their physicality, and create their own identity that they seek. I call this the age of body alchemy. Using technologies to magically transform ourselves and our expressions.
You also described implantation as part of progress. It's not just attaching an external-
Oh, sure, sure.
Yeah, mechanism but actually implanting.
Yeah, generally the tech will go on the body and inside the body as we progress in time.
Now, how does that work in terms of feasibility with the body rejecting foreign tissue... You want the tissue, if I understand correctly, and the technology to begin to grow together so that they're not such starkly separate systems with a closed tissue system and an external attachment. But rather that you implant carbon fibers or titanium materials, or wires, electronics, electrodes, and the tissue grows around it. Is that something you think is still feasible or is there a lot of rejection of the hardware?
That's the Luke model of the future. So Star Wars, Luke, the scene with the cut and the gears inside. That model is fifty years from now, the prosthesis is going to have biological skin over it but when you cut the skin you'll see all these gears and motors. That's one model of the future. Another model of the future is like this. That I don't have biological skin that surrounds my mechatronics.
It feels like the biological skin's the shortcoming. After all, electronics, we haven't really tested to last that long. Gadgets conk out really easily. They have very short lives. If it's inside your body, how are you going to replace those wires and those electrodes? That seems to be a very hazardous problem.
You're exactly right. So that's why I have a bit of a bias away from Luke. It would be nice to have a world where we implant and augment an eight year old and that eight year old lives to a hundred and fifty years old, and we don't have to continually open the body up with surgery after surgery. That says, "Okay, we need to minimize the technology that's inside the body and maximize technology that's out." If we need to transfer neural information into and out of the brain, broadly defined, we should do it in a wired way. We should figure out the tissue engineering and synthetic tissue interfaces where you can run wires, run anything you want across the skin boundary without the threat of infection, with a huge bit rate and efficiency. If we're going to truly wire our brains, that's a lot of information and I think putting tons of electronics in the body and wireless communications and we'll break down.
We're actually thinking: how can we run wires across the skin? I believe that's a solvable problem. Imagine that each of us in the future has buttons that sit right on our skin and you can just plug into your nervous system.
I mean, you could do that very remotely. So I could have something like what you're describing, but be operating a limb that's in another country?
Sure.
Absolutely. Yeah, telepresence is a wonderful idea, and solutions in bionics are going to make telepresence tractable, and a realistic experience.
Are you concerned about military applications of some of these augmentations?
I was raised in Lancaster County, and my family was in the tradition or in the religion of Mennonite and the teaching of Mennonites is they're devout pacifists. They take seriously “turn the other cheek.” So I was taught of the horrors of war. Some Mennonites, they don't even vote for fear that the politician will start a war because they don't want blood on their hands. That's my upbringing and I do take money from the US Department of Defense, often to solve medical problems. A large fraction of the research budget of the DOD is to solve medical problems.
Well, we are bringing back a lot of soldiers who need prosthesis.
Right. If you plot prosthetic innovation versus time you see a spike after Iraq.
During and after every major war. Except for Vietnam because of the bad feelings in the country. World War II, for example, a massive spike in progress.
The word "disability" has become contentious.
Yeah, it's not what I'm about. For persons that have paralysis, for example, that want to walk and dance and move their bodies, I want to give them that freedom. It's about freedom. It's not about taking away from an individual.
If a person wants to have a certain physicality or certain sensory experience, or cognition, awesome. I just want to give people choices. Now people don't have choices.
What about enhancement? Are we getting to the point that nobody wants to just be what evolutionary biology prescribed?
Yeah, about half of my lab is working on assistive rehabilitative technologies and the other half is working on augmentation. I love the idea of bionics and cyborg function augmenting humans going beyond innate physicality, or cognition, or sensory experience.
I'm a designer, so I get upgraded every week, every month. As my biological body ages and degenerates, my synthetic body improves in time. The synthetic part of my body is beyond immortal. It's actually improving in time. If you think about it rationally, when we're at the point where we can design and construct a limb replacement that's as good or better than the innate biological limb, and a person wakes up and their hands hurt, their arthritic, it's the completely rational decision to upgrade to a synthetic design.
You made this very radical statement that at some point augmentation or upgrading is the only logical decision from a biological organism, purely biological to an interface. At what point do we say, "Well, then just go pure machine." At what point are you going to say that any biology is advantageous and is there... Do you think of a boundary at which you say, "Oh, just go pure robot, pure AI." Is it just the brain as the soft tissue that we should retain?
With today's knowledge, we have a fairly deep understanding of a whole set of synthetic materials. For example, we have steel, we understand the properties of steels, we can build things out of steel. There will be a day when we have that deep understanding of cells and tissues. We're not there yet, but that day, the designer will sit down and say, "Okay, I want to design something. Should I use biologics here or synthetics?" The answer's not always going to be synthetics, nor is it always going to be biologics. So many of our design entities will be hybrid, just like I'm hybrid.
Nature took millions or sometimes billions of years to figure out a really great ankle mechanism through this process of Darwinian evolution and natural selection. We, as human beings really aren't going to evolve naturally any more, presumably.
Are we evolving faster now at our own hands? Instead of external and environmental pressures, it's our own pressures of things like bionics that's happening so fast? Do you think we are undergoing human-driven evolution as opposed to nature-driven evolution?
Yeah sure. And it's so fast because it's... Technology is crystallized imagination. The innovation's becoming faster and faster and faster to the limit where if you can think it, and it's consistent with physical law. It can happen.
That's why sci-fi literature... So much of sci-fi ideas actually come into existence, because most of what we can imagine, actually, is possible.
Do you feel that we're making a new species out of ourselves? There's going to be homo sapiens, and then there's going to be post-homo sapiens.
Sure. The question is: What are we? If I replace all my limbs with mechatronics, am I still a human being? Most would say yes, but as I begin to replace and upgrade and augment at what point would you say I'm no longer a human? Again and again in this 21st century, we’ll have to ask the question: What does it mean to be human? ♦
The custom jumpsuits worn by Janna Levin and Hugh Herr were made by Andrea Lauer.
Subscribe to Broadcast