Natural Intelligence: What we and AI can learn from nature

0:00This message comes from AT &T. Staying connected matters. That's why AT &T has connectivity you can depend on. or they'll proactively make it right. That's the AT &T Guarantee. Terms and conditions apply. Visit att.com slash guarantee for details. This is the TED Radio Hour. Each week, groundbreaking TED Talks. Our job now is to dream big. Delivered at TED conferences. To bring about the future we want to see. Around the world. To understand who we are. From those talks, we bring you speakers and ideas that will surprise you. You just don't know what you're going to find. Challenge you. We truly have to ask ourselves, like, why is it noteworthy?

0:44And even change you. I literally feel like I'm a different person. Yes. Do you feel that way? Ideas worth spreading. From TED and NPR. I'm Manoush Zomorodi, and on the show today, natural intelligence. And we're going to kick things off by going back in time to the 1750s and a marshy, swampy, subtropical wetland in North Carolina where a man named Arthur Dobbs lived. Yeah, no, I like Arthur Dobbs. This is our tour guide, Greg Gage. He was the governor of North Carolina. This is back in the 1750s, still under British rule. And Arthur, he says, was a curious man, a bit of a gentleman scientist. So when Arthur started to hear a rumor about a very unusual plant growing not that far away, he decided to go check it out.

1:47And he heard these stories about this plant. He says, well, I'm the governor of this land. I should go investigate it. He went out to the swamps, and so there's a little small area, I think it's about six square miles. And then there, low to the ground, are these little tiny plants in that if you watch them long enough, a bug will eventually fall into its little, or walk across its little leaves, and it snaps shuts and it eats this bug. And so, of course, he's pretty fascinated by this. Understandable because, well, no Europeans had ever documented a plant like this before. So. He does what scientists do at the time.

2:30They take out a letter and they write to their colleagues in Europe. The great wonder of the vegetable kingdom is a very curious unknown species of sensitive. He writes, it's a dwarf plant. Leaves are like a narrow segment of a sphere consisting of two parts, like the cap of a spring purse. Upon anything touching the leaves or falling between them, they instantly close like a spring trap. To this surprising plant, I've given the name of fly trap sensitive. The fly trap sensitive, or as we call it today, the Venus fly trap. The news of this remarkable animal-eating plant took off across Europe.

3:14And 100 years later, Charles Darwin would write an entire book called Insectivorous Plants. Eventually, the great Charles Darwin got to study this plant, and this plant absolutely blew him away. He called it the most wonderful plant in the world. Here's Greg Gage on the TED stage. This is a plant that was an evolutionary wonder. This is a plant that moves quickly, which is rare. And it's a plant that's carnivorous, which is also rare. And it's in the same plant. But I'm here today to tell you that's not even the coolest thing about this plant. The coolest thing is that the plant can count. So let's pause for a moment.

3:49Greg Gage is not just an amateur flytrap historian. He is actually a neuroscientist and educator. And on the TED stage, he conducted a live science experiment. So I'm going to pretend to be a fly right now. Surrounded by monitors, microscopes, and of course, plants. And here's my Venus flytrap. And inside the leaf, you're going to notice that there are three little hairs here. And those are trigger hairs. And so when a fly lands, I'm going to touch one of the hairs right now. Attached to the flytrap, EKG sensors, measuring any electrical signals generated by the plant. Ready? One, two, three. And there, the monitor lit up as Greg grazed the hairs inside the flytrap.

4:31What do we get? We get a beautiful action potential. However, the flytrap doesn't close. That's because it's waiting to see if it gets touched again. within 20 seconds or so. Venus fly traps don't want to be hasty for several reasons. Number one is that it takes a long time to open the traps back up, you know, about 24 to 48 hours if there's no fly inside of it. And so it takes a lot of energy. And number two, it doesn't need to eat that many flies throughout the year. It only needs to eat a handful. It gets most of its energy from the sun. It's just trying to replace some nutrients in the ground with the flies.

5:06And the third thing is, it only opens, then closes the traps a handful of times until that trap dies. So therefore, it wants to make really darn sure that there's a meal inside of it before this fly trap snaps shut. So how does it do that? It counts the number of seconds between successive touching of those hairs. I'm going to touch the Venus fly trap again. I've been talking for more than 20 seconds. And then if I'm a fly moving around, I'm going to be touching the leaf a few times. I'm going to go and brush it a few times. And immediately the flytrap closes. So here we're seeing the flytrap actually doing a computation.

5:45I mean, that sounds like this plant is pretty smart. Yeah, it's, well, I mean, it definitely is competing. I always think that plants are kind of cool because, you know, humans, if there's a rough situation, we can just run away, where we can just kind of like get out of dodge. But these plants are stuck there. They're in the ground, right? And so they've got nothing to do except for try to feed themselves. They come up with very incredible ways of doing that. We hear a lot about the powers of artificial intelligence. But all around us, nature continues to find extraordinary ways to survive and communicate that we are still just beginning to understand.

6:27So today on the show, natural intelligence. New findings about the brilliance of dragonflies, our immune system, and whales, and how they are influencing human behavior. First, though, back to Greg Gage. He brought another surprisingly sensitive plant onto the TED stage. The mimosa, not the drink, but the mimosa pudica. Greg, during your stage experiment, you also had this other plant that kind of looks like a fern, and it's called a mimosa. And this is a plant that's found in Central America and South America, and it has behaviors. And you just lightly touch the leaves, and it kind of like wilted away from you.

7:11If I tap the leaf, the entire branch seems to fall down. And my dad actually had a mimosa plant when I was a kid, and I was fascinated by this response. Like, I would touch it, and it would pull away from me. Like, get off me, you know? Yeah, so it's actually very, very similar to hematryl. So there's like a touch receptor, just like we do in our skin. And when we press down on something, we feel it because an electrical impulse is being sent back up to our brains. And we interpret that impulse as the feeling of touch. Inside of this leaf, very similar cells are within there that will send an electrical current.

7:46Only this time, instead of using muscles, it again uses water. It flushes the water out and makes the plant move. But why? Is it always about survival? I suspect so. And it's funny because with the mimosas, there's a couple of theories of why. If an animal brushes past, it doesn't look as, I'm looking at mine right now, they don't look very tasty, right? Maybe I would eat another different plant if I saw that one. Or maybe it kind of freaks out some insects that would want to land on it. So that's a learning mechanism that it can do, but that's kind of boring compared to some other experiments that have been done.

8:22Experiments where you can take a peapod and you let it grow in the dark. you have like a little tube that goes up and it kind of goes like a Y. It's called a bifurcated tube. And if you shine a light in one of the tubes, say on the right side, and you blow a fan on the other one, you can do this protocol where you kind of blow the fan on the left and then you shine the light on the right. And then the next day you blow the fan on the right, you shine the light on the left. Each day it's kind of growing up this tube, trying to get to the top of it, right? And then on the experiment day, right where it's about to make a decision, you flip it again, but you only blow the fan.

9:01And then you ask the question, which way will the plant grow? Will it grow where it last saw the light, which makes sense, right? Or did it figure out the relationship between the fan and the light and grow to where the light would have been if it's going to come on soon, right? You get a majority of the plants that kind of figure it out. They kind of figure out that they have to go in the other direction and go away from where it lasts on the light, but where the fan was indicating where we go. To me, that's kind of flexible. And that's starting to show some decisions, right? That's starting to show a little bit of intelligence.

9:34So it's the ones that made it are the ones that can sort of live on to tell the story, right? So then those genes get inherited and move forward. so we have covered plants which seem like a more simple life form than say mammals yet kind of display some very seriously smart behavior but you are also really interested in um it seems strange to say this but you're really excited about slime mold which is not a plant but no but a mole is a cell right a slime mold is a single cell uh they they kind of sit in a petri dish and you feed them, you feed them little oats and they kind of wrap around it and they kind of suck up the nutrients and they kind of can go dormant for many, many years and they can come back alive again.

10:20And so, uh, yeah, they're absolutely fascinating, but they're, uh, you can see them though, even though they're single cell? They could be, yeah, they can be a single cell could be like, you know, a meter long there, they're, they could be big. So they're macro. And so, uh, we, we were doing an experiment. One of the experiments we did was we put a piece of food in there. We watched how often would it go towards the food? Not surprisingly, it goes towards the food a lot, right? But slime molds are kind of, they don't want to be dried out. They don't like sunlight. And so if you shine a light there, then they will shy away from it.

10:50They'll try to go around it, but they try not to go through the light. That would make sense if it was an animal. You're like, okay, well, their eyes are seeing it. The eyes are sending the message back to the brain and the brains are telling the muscles which way to go. But then you realize, wait a minute, There's no brain, right? There's no, it's just, it's just cells, right? So they have these sensors on the outside of the cell that's sending back information, but that information is being processed inside the cell itself. And so I think there's a lot to be said about the, about the cell. I mean, the cell is basically a little computer.

11:23It's got a little Turing machine inside of it and it has goals, right? It tries to do things. And so, uh, and it has a lot of, a lot of things and its resources to be able to do. I recently went to a lecture about the crows that are the only birds that we know that use or fashion and then use tools. Oh, corvids, yeah. They're amazing. And just every so often you hear this incredible story that shouldn't blow our minds, but does about how the natural world is so incredibly smart. And I guess I'm wondering, you know, should it expand our definition of intelligence? In simple terms, how do we define what intelligence is?

12:08Yeah, I think the simplest way, given what you've got, right, you've got to figure out how to be able to get to what you want, right? And so that's kind of my kind of the back of the envelope sketch of intelligence is being able to get to what you want given what you have. And that's what intelligent things do. You look at a dog trying to get through a door. It'll kind of go check the other door. It's trying to figure out what it needs to do. And so you can look at these plants as, you know, in the case of these plants that are trying to find the light they're doing. They're taking what the information they have to figure that out.

12:41You know, you can look at single cells doing that. I think the joy of intelligence is really in every living thing. I think every cell is intelligent. I think everything that comes from cells is intelligent. That's Greg Gage. He's a professor at the University of Michigan and co-founder of Backyard Brains. a company that builds neuroscience experiments for kids. You can find all his talks at TED.com. On the show today, natural intelligence. I'm Manoush Zomorodi, and you're listening to the TED Radio Hour from NPR. We'll be right back.

13:32This message comes from NPR sponsor Walmart. Did you know that Walmart is one of Fortune's 100 best companies to work for in 2025? It's no surprise. Walmart is investing$1 billion in training and development to help associates build lasting careers with or without a degree. In fact, 75 % of Walmart management started in hourly roles. With benefits like health care, paid parental leave, and tuition support, Walmart isn't just a place to work, it's a place to grow. Learn more at walmart.com slash spark. This message comes from NPR sponsor Informatica. Everybody's ready for AI except your data. Informatica offers the only end-to-end AI-powered cloud data management platform that connects, manages, and unifies your data across virtually any multi-cloud or hybrid environment.

14:27Informatica accelerates your AI readiness and scale at a pace that allows you to pay for only what you use without compromise. Get your data ready for AI at informatica.com slash AI. Informatica, where data and AI come to life. Hey, before we get back to the show, I want to tell you about our next bonus episode for TED Radio Hour Plus. So we hear a lot about having empathy for others. But why is it totally exhausting us? And is putting ourselves in other people's shoes actually making the world a better place? What's the difference between empathy, sympathy, and compassion anyway? Well, renowned psychologist Paul Bloom and author of the book Against Empathy is back.

15:13He's hilarious and thought-provoking, and he's on our bonus episode coming Wednesday. If you're not a Plus supporter yet, please join your fellow listeners to get all kinds of bonus content and all our episodes sponsor-free. Plus, you'll be supporting the show. Just go to plus.npr.org slash TED or give it a try right in the Apple Podcasts app. It's the TED Radio Hour from NPR. I'm Manoush Zomorodi. On the show today, natural intelligence, which is actually where some computer scientists are looking for inspiration to design the next generation of artificial intelligence. Yeah, so African dung beetles, they roll up balls of feces and balls of dung and roll them away as quickly as they can.

16:04Because even the smallest creatures on Earth can execute some amazing feats. So they're standing on their head, they're rolling the ball of dung with their hind legs, and they're using various cues to roll in a straight line. If they're nocturnal, they're known to use moonlight to be able to make sure that they're going in a straight line. And I wouldn't be able to roll anything standing on my head. This is Frances Chance. Sahara desert ants, when they find food and they want to bring it back to their nest, they know how to calculate the straightest path back to their nest. Frances sounds like an entomologist.

16:38Oh, and honeybees. You know, they also forage. But she's actually a computational neuroscientist. At Sandia National Laboratories, she researches how natural intelligence can help develop new security technology. For example, what missile defense systems might learn from the dragonfly. So they're very good at what they do. These graceful, flittering creatures also use a very special technique to hunt. Dragonflies are really good at hunting. We know that they fly to intercept their prey. They fly really fast, and they're very successful. It's known that dragonflies catch up to 95 % of the prey that they choose to go after.

17:24And even though they're really fast, they don't just fly straight at their prey. They fly on an interception pathway, which means they're aiming slightly ahead of where their prey are. So is that like Gretzky, the hockey great, saying like, don't go to where the puck is, go to where the puck is going to be? Exactly. We need to aim ahead of where the puck is going to be. And so the dragonfly is constantly reacting to changes of the prey's direction or the prey's speed to calculate how far ahead of the prey they need to aim. By understanding how these nearly instantaneous calculations happen in the dragonfly's brain, Frances hopes to build AI that mimics it and is just as efficient.

18:13Here she is on the TED stage. When dragonflies are hunting, they do more than just fly straight at the prey. They fly in such a way that they will intercept it. They aim for where the prey is going to be. To do this correctly, dragonflies need to perform what is known as a coordinate transformation, going from the eye's frame of reference, or what the dragonfly sees, to the body's frame of reference, or how the dragonfly needs to turn its body to intercept. And dragonflies are fast. This means they calculate fast. The latency, or the time it takes for a dragonfly to respond once it sees the prey turn, is about 50 milliseconds.

18:53So in the brain, a computational step is a single neuron or a layer of neurons working in parallel. It takes a single neuron about 10 milliseconds to add up all its inputs and respond. The 50 millisecond response time means that once the dragonfly sees its prey turn, there's only time for maybe four of these computational steps or four layers of neurons working in sequence one after the other to calculate how the dragonfly needs to turn. In other words, the neural circuit that I need to understand can have at most four layers of neurons. This is a small neural circuit, small enough that we can identify it and study it with the tools that are available today.

19:39And this is what I'm trying to do. I have built a model of what I believe is the neural circuit that calculates how the dragonfly should turn. In a computer simulation, I can predict the activities of individual neurons while the dragonfly is hunting. To test the model, my collaborators and I are now comparing these predicted neural responses with responses of neurons recorded in living dragonfly brains. These are ongoing experiments in which we put living dragonflies in virtual reality. Now, it's not practical to put VR goggles on a dragonfly. So instead, we show movies of moving targets to the dragonfly, while an electrode records activity patterns of individual neurons in the brain.

20:27If the responses that we record in the brain match those predicted by the model, we will have identified which neurons are responsible for coordinate transformations. The next step will be to understand the specifics of how these neurons work together to do the calculation. But this is how we begin to understand how brains do basic or primitive calculations. So you are building computer models based on the dragonfly's brain that can intercept things in just a few steps. The dragonfly, it must be doing really complicated math really fast. Is that what you are trying to figure out, those calculations?

21:04Yeah, so what I'm really interested in are what are the fundamental operations that neurons are capable of, or what are the fundamental operations that neurons do. And that's what I want to bring to a computer. It may be something like basic trigonometry, or it may be something that's kind of a different type of math than we're used to thinking about. But that's what we're trying to understand, because if we can understand the operations, then we can begin to understand what the algorithms or, say, the computer programs of the brain are. The way that these neurons compute may be different from anything that exists on a computer today.

21:44And the goal of this work is to do more than just write code that replicates the activity patterns of neurons. We aim to build a computer chip that not only does the same things as biological brains, but does them in the same way as biological brains. This could lead to drones, driven by computers the same size as a dragonfly's brain, that capture some targets and avoid others. Personally, I'm hoping for a small army of these to defend my backyard from mosquitoes in the summer. The GPS on your phone could be replaced by a new navigation device based on dung beetles or ants that could guide you to the straight or the easy path home.

22:25And what would the power requirements of these devices be like? The human brain is estimated to have the same power requirements as a 20-watt light bulb. Imagine if all brain-inspired computers had the same extremely low power requirements. Your smartphone or your smartwatch probably needs charging every day. Your new brain-inspired device might only need charging every few months or maybe even every few years. You know, computers touch us in all sorts of ways that we totally take for granted. But one resource that limits what computers can do today is being able to power them. Yeah, and you mentioned a future scenario where maybe we wouldn't need to charge our devices every day.

23:11Maybe we could go months or even years. Could we potentially scale back big time from using all the energy we need right now to run massive servers and data centers all over the world? Yeah, I think that it could have long-reaching impacts by decreasing just human carbon energy footprint on the world, definitely. As it is, a lot of these data centers need to be next to some natural resource like a river to be able to generate enough power to use these algorithms like Google search. So I think that there's a lot of potential there that we may be able to bring that cost down. So bring this back to what you do at Sandia for us.

23:58I mean, I know you can't get into the details, but when it comes to national defense, it makes me think of Israel's Iron Dome. And leaving aside the politics, is the aim to find ways to make missile defense systems like those more efficient? Well, Sandia is interested in national security missions. That requires a lot of computer power. I'm not necessarily going to talk about what those are, but if we're doing our job, you won't necessarily see the impact of that. So being able to understand how neurons do what they do for low power means that the cost of each of these individual operations or on the cost of each of these interactions of a human with, say, something in the cloud is going to come down.

24:46And so what I'm interested in is how the dragonfly brains are able to do this calculation with low power and really remarkably fast. There is a debate about whether artificial intelligence can actually be intelligent. What about the dragonfly? Do you think of it as intelligent? Yeah, so I think there are a lot of different definitions to what intelligence is. When I think about human intelligence, it's our ability to adapt or take in new information and behave differently based on new situations. For dragonflies, they're examples of neurons solving a task in what I would say is maybe even an optimized way.

25:37You know, the dragonfly has evolved to do this particular task very well, very fast, very efficiently. So I call them clever, you know, clever solutions. They're examples of what intelligence could produce. That was computational neuroscientist Frances Chance. You can see her full talk at TED.com. On the show today, ideas about natural intelligence. So far, we've heard about plants that can count and dragonflies that intercept their prey in milliseconds. But what about our own natural intelligence? The one in our bodies. How are you feeling? I'm feeling not so good. That's our senior producer, Sanaz, and her daughter, Mina.

26:30I feel a little nauseous, like there are pebbles or rocks in my stomach. And my head just occasionally hurts. Mina is homesick with a virus. And she doesn't want to do much of anything. Right now I'm sitting on the couch. I just want to watch a movie and watch something on my iPad. As we know, these symptoms are coming from Mina's immune system that is trying to fight off that virus. Leading the charge are molecules called cytokines. They're basically like the chemical messengers of the immune system. Keely Muscatel is a psychology and neuroscience professor at UNC Chapel Hill. That's where she studies the links between our physical and mental health.

27:33And she says these cytokines float around in our bloodstream looking for anything suspicious. And when they find something, they sound the alarm. It's like, oh, there's a problem here. We need to do something to try to contain this. And it's trying to signal to other immune cells to come and try to figure that out. And that causes inflammation. And that's what we tend to experience when we have been infected with a virus or some sort of other pathogen is that widespread systemic inflammation. Now, in doing this, cytokines cause the physical symptoms we commonly have when we're sick. Keely Muscatel continues from the TED stage.

28:18things like fever and achiness and fatigue. So even though we usually think of those symptoms as being caused by a virus or a bacteria itself, they're actually caused by our own immune systems activating to try to eliminate the pathogen. But in addition to those physical symptoms, decades of research in both animals and humans clearly shows that cytokines also cause changes to our mood and to our social behavior. So inflammation in the body can signal to the brain to cause us to feel down, depressed, and even hopeless. Inflammation can also make us want to socially withdraw from other people to avoid interacting with individuals in our social networks.