The Army of Autonomous Robots Restoring Nature | Tom Chi | TED

Name: The Army of Autonomous Robots Restoring Nature | Tom Chi | TED
Duration: 18 min 53 s
Description: Impact investor Tom Chi challenges a dangerous assumption: that economic growth and ecological health are opposing forces. He reveals how advances in AI and rob

TED18:53Added Aug 30, 2026

Impact investor Tom Chi challenges a dangerous assumption: that economic growth and ecological health are opposing forces. He reveals how advances in AI and robotics are enabling a radical shift towards innovation as a force for restoration. Imagine mines that extract less, farms that regenerate soil and fleets of robots that can plant 100,000 mangroves in a single day. What if the same technologies that power our economy could actively repair the pl

Watch on YouTube →

Contributed by 刘嘉琪

Transcript

00:08So over the last decade,
00:10I've been working on resolving a profound paradox.
00:14And, simply stated,
00:16it's that if you go and talk with any person, actually --
00:20you could go grab 100 random people off the street --
00:23and you were to go ask them how do they feel about nature,
00:27you're going to end up with extremely positive answers,
00:30ranging from, you know, "nature's inspiring,"
00:35"nature is the most beautiful thing that exists."
00:39And you'll see it everywhere.
00:40They'll put it as the backdrop of their desktop and their phones, like,
00:44every little thing, absolutely.
00:48Every single person you ask,
00:49you're going to get, like, a pretty positive response.
00:53And even though, if you ask all these individuals how they feel about nature,
00:56they're really positive about it,
00:58somehow, as a collective civilization,
01:02we've come together and we are destroying nature at a planetary scale.
01:08And therein lies the paradox.
01:10How did this happen?
01:11How did a group of individuals that all love nature
01:15somehow create a civilization,
01:16create an industrial economy that is out there,
01:20effectively planetary level assault on nature?
01:24Well, I think it actually stems from a broken mental model
01:28that we've kind of unconsciously adopted.
01:30And I'd like to start dissolving that right now.
01:34Now, you can see the mental model with the X below it,
01:39and that is the one
01:40that I would like to say that we've accidentally adopted.
01:42And that's a mental model
01:44that whenever you get economic wins, they are traded off against the ecology.
01:49And if you really care about the ecology and want to make that healthier,
01:52unfortunately, you're going to have to trade off economic wins, right?
01:56It's one or the other, kind of a balancing act.
01:58And depending on how much we care about the economy,
02:01or how much we care about the ecology at any given time,
02:03the pendulum swings one way,
02:05the scale swings one way, or it swings the other.
02:08And I'm going to tell you that, sure,
02:09that is a psychological position that you can take,
02:12but it's actually not one that is particularly physically true.
02:15I'm formally trained as a physicist.
02:17I do think about what is physically true,
02:19and what is actually much more true about the economy
02:22is the economy is not versus ecology.
02:26The economy is a subset of the ecology.
02:29And maybe this is a new idea to you guys,
02:31but I can prove it to you very quickly.
02:33Actually, you could prove it to yourself,
02:35even just with the clothes on your body or the things immediately around you.
02:39Because if you think about and look at everything the economy has produced,
02:43everything in the economy is either mined or grown,
02:47which means it comes directly from nature, no abstractions.
02:50You guys might be wearing some cotton, obviously grown.
02:54You'll be sitting on a chair that's got some metal underneath --
02:57that was mined.
02:58Everything in the economy is mined or grown, full stop.
03:01And some of you guys might be thinking,
03:03"Aren't we moving to a digital economy, a virtual economy?"
03:06Well, not really.
03:07Like, every line of code that is ever run
03:09runs on a substrate that was mined or grown.
03:13Every single service you've ever used
03:15is using server architectures that are mined or grown.
03:18When I say the entire economy is mined or grown,
03:20I mean it literally.
03:22There's literally nothing that doesn't come directly from nature.
03:25And to the extent that you damage the ecology,
03:28you actually start to create problems for the economy.
03:31And this is what we're experiencing right now.
03:33And if you think about it in this balancing-act-type way,
03:36you're going to miss the right way to actually fix these problems.
03:41Now, let's talk about exactly how much we are mining and growing.
03:46This graph actually kind of tracks 50 years
03:49of how much extraction we've been taking every single year from the planet.
03:53At this point in history, it's over 90 billion tonnes per year.
03:57It comes out to about 11.5 tonnes per person per year.
04:01And if you guys feel like you don't do that much,
04:04well, I'll shock you to say that 11.5 is the average in Asia,
04:08but Europe is about two x that, and America is three x that.
04:11Yay.
04:13There's two billion people that live on less than five dollars a day
04:16doing substantially less than that --
04:18that's why it all balances out.
04:20But this is exactly how much we're mining
04:22and exactly how much we're growing.
04:24Now, in the process of mining and growing this much,
04:28and using it to power everything in the economy --
04:31because, like I said, literally everything in the economy is mined or grown --
04:35we've been using really old industrial ideas and industrial processes.
04:39Most of how we're growing today was invented about 50 years ago.
04:43Most of the ways that we've been mining, refining metals,
04:46all that sort of thing, were invented about 100 years ago, 150 years ago.
04:50These are not technologies that we've updated recently.
04:53And with the arrival of new robotic and AI tools,
04:57I think it's the right time to go ask new questions
05:00about whether we could be mining and growing differently,
05:03in a way that starts to honor this idea
05:06that the economy is a subset of the ecology.
05:10Now, this is where it overlaps into my world,
05:12because my entire career has been built off of doing new inventions
05:17and robotics, artificial intelligence, advanced algorithms.
05:21And I've shipped everything from Microsoft Office --
05:24sorry about that --
05:25(Laughter)
05:26you know, to web search -- I think that one was fine --
05:29to self-driving cars.
05:31So I've worked on relatively sophisticated things,
05:34and given that, I have an interesting background, perhaps,
05:38to be able to go look at these problems
05:40and see if we can take a different swing at them.
05:42And I'm going to share a number of examples with you today.
05:47Now, these examples fall into three major shifts --
05:50remember, everything's mining or growing --
05:52and they're represented by these three images here.
05:55So on the left-hand side, we have a bunch of mined materials,
05:59and what we need to be doing
06:00is we need to figure out more and more ecological ways
06:04to be able to go mine materials
06:05and get the most of the ores that we extract
06:08so we do the least disturbance of earth and watersheds in the process of mining.
06:14In addition,
06:15what is even better than mining more ecologically and mining less
06:21is to not mine at all.
06:22And to the extent that we're able to go close the loop
06:25through really skillful mechanical or chemical recycling,
06:29we can have a larger and larger proportion of the feedstock for industry
06:33move over to closed-loop materials,
06:36as opposed to virginally extracted materials.
06:39The second major shift has to do with the way that we grow.
06:42A lot of the way that we've been growing currently is very unsustainable.
06:47It basically is damaging soil function.
06:49And little by little, we've been kind of wearing down topsoil
06:52in agricultural lands all across the world.
06:55And Gabe Brown is pictured here.
06:58He's a friend who has taught me a huge amount
07:00about regenerative agriculture,
07:01and I've really learned from him that if you invest in soil function,
07:05you can actually make it easier to grow, cheaper to grow,
07:08a higher margin to grow every single year,
07:10and do so in a way that is regenerating soil function,
07:13giving more services to biodiversity,
07:16and even healing the hydrological function of those soils.
07:19And lastly, we need to be thinking about large-scale repair,
07:24because we've been at the Industrial Revolution
07:27for a couple hundred years now,
07:28and there's a lot of landscapes that we've heavily degraded.
07:31And if we are serious about the task of renewing the ecology
07:35in order to support a vibrant economy going forward,
07:37then we're going to need better tools for large-scale repair.
07:41So let's jump into all three of these.
07:44So what's pictured here is a company I have the privilege to work with.
07:48And they are a great example of moving closer to that closed-loop world.
07:53So what you’re seeing here is actually an image from inside
07:57the largest lithium NMC battery recycling plant in North America.
08:02And this is an advanced form of chemical recycling
08:04that is able to go bring all of these used battery materials,
08:07because most lithium batteries kind of have a 10-year life;
08:10they don't go too many years beyond that.
08:12And after that's the case, you know, you can't use it in the car,
08:16or you can't use it in the consumer electronics device anymore.
08:19You want to be able to recover those materials.
08:21The process that they do here is about two times cheaper
08:24than the next closest process,
08:26and is able to return the material to complete virgin quality.
08:30It's better than the stuff
08:32that you would have been able to mine out of the ground in the first place.
08:35And if we get really skillful about closing these loops,
08:38what's great is the car battery doesn't just evaporate and disappear.
08:41It's a relatively large object that you can go handle,
08:44and you can do a reverse logistics supply chain
08:46and pull these things together.
08:48And whether it's robotics and AI to do advanced mechanical recycling --
08:52or, in this case, advanced chemical recycling --
08:54then there are really skillful ways, with our new technologies,
08:57to be able to go close the loop
08:59and make it so that a higher and higher fraction
09:01comes from a post-consumer or post-industrial waste stream,
09:04as opposed to from the ground.
09:07Now, moving over into the regenerative growing side,
09:10we're actually at a really compelling point in history,
09:14because there’s a mini renaissance in regenerative agriculture
09:17that's happening right now,
09:18with different farmers around the world discovering the benefits
09:22of agroforestry, intercropping, you know, no-till agriculture
09:27and a lot of other practices
09:29that really help to establish healthy soil function
09:33and a healthy soil microbiome.
09:34And what you're actually seeing on the left-hand side
09:37is a new technology,
09:38which also uses machine learning to be able to go disambiguate.
09:41It's a new form of Raman spectroscopy
09:43that allows folks to be able to listen to the soils in this really skillful way.
09:49So effectively, they're able to go and measure
09:52all these compelling compounds from the soil,
09:55so they're able to have the soil speak to them
09:57in ways that the soil can basically tell them,
10:00“Hey, here’s the next couple of things that you should do to make me healthier.”
10:04Instead of it kind of being a black box that needs to get interpreted,
10:07now farmers can have a direct relationship with their soils
10:10and be really skillful in the management toward greater and greater health,
10:14fewer inputs and higher margins every single year.
10:20Moving on into other ways
10:22that artificial intelligence and machine learning
10:25might be really useful for agriculture,
10:27what you see pictured here is the development of corn or maize.
10:31And it was an Indigenous project
10:34that happened over the course of hundreds of years.
10:37And they started with, basically, an inedible bit of grass,
10:40because corn is actually a type of grass.
10:42And by selective breeding over generations and generations,
10:46they went through lots of different varieties,
10:48until we got to the "lots of calories per grow cycle" version of corn
10:53that is now feeding a huge percentage of the calories around the world.
10:57Now, this was an Indigenous activity that happened over hundreds of years,
11:01and I'm really thankful for it,
11:02because most of the foods that we eat today were selectively bred
11:06to be as large and healthy and nutritious as we experience them.
11:11But using artificial intelligence and machine learning,
11:14we've been working with a company
11:15that has been able to rapidly speed up this process,
11:18and not through genetic modification.
11:20What they do is they're able to take the sequence information
11:24from all the existing commercial crops,
11:26plus a bunch of native varietals that are not in current circulation,
11:33and work out what the different gene functions do,
11:35and then map out exactly the crossbreeding pathway
11:38in order to go get the desired traits.
11:40So what we have here is adaptive sugar cane,
11:43which dramatically reduces the amount of deforestation
11:46required to get to the yield level that you want.
11:48You also have heat-resistant tomatoes
11:51that are able to grow in way hotter, way drier conditions,
11:54which is really important,
11:55because we're going to go through at least a 50-year period
11:59where we're going to be destabilizing a lot of the farmlands of the Earth
12:02as the climate destabilizes,
12:04whether that's hotter or colder, wetter, drier.
12:06It's all going to happen,
12:07and being able to have seed stock that is ready for that challenge
12:10is really powerful.
12:12And lastly, a cotton that basically is drought-tolerant as well,
12:16requires a fraction of the water, one-tenth the water,
12:19and much less pesticide and fertilizer input.
12:22And all of these things are fantastic for the planet,
12:24but they're also fantastic for the future of us
12:27having viable food and materials
12:29in a destabilized, growing environment.
12:34Lastly, let's get on to scalable restoration.
12:36And what you're seeing here is an image from the company Chloris Geospatial.
12:40And it is, of course, of the Amazon basin.
12:43But what's really compelling about this company
12:46is that they've done really deep work on sensor fusion
12:51across a bunch of satellite feeds,
12:53and they also paired that with over a decade in the jungles,
12:57meter by meter, doing ground-truthing data,
13:01to be able to go really verify how much terrestrial biomass is associated
13:05with signals that can be detected from satellites, via remote sensing.
13:09And given this, they've been able to make the most accurate,
13:13both historical and current assessment of aboveground biomass on planet Earth.
13:18And the data stretches all the way back to the beginning of the 21st century,
13:21so over 20 years of data on that front.
13:24And that really allows us to see which landscapes we're hurting,
13:28which landscapes are recovering,
13:29and if people are developing restoration projects or carbon projects,
13:33this is a fantastic way to go stay on top of how those are going.
13:38Now, this is great technology and also uses really advanced algorithms
13:43and allows the things that I've been talking about.
13:45But in some ways, it's a little bit passive.
13:47This doesn't restore the forest itself.
13:49This doesn't restore the grassland itself.
13:52This just helps people monitor the changing of that.
13:55But what if -- going back to this triangle for a moment --
13:58we were to get more ambitious and we were to say,
14:01"Let's challenge this linkage
14:04between the industrial machines that run our economy and nature,
14:07and instead of it having to be an accidental relationship of damage,
14:11what would it look like if it was an intentional relationship
14:14of active repair?"
14:16And I'm going to show you that right now with my last two examples.
14:19This one's a short video.
14:20You're going to hear these little ticks,
14:23and every one of those ticks is a mangrove seed being planted.
14:27The pace of these ticks
14:28is basically planting about 100 mangroves per minute from one drone.
14:33Video: (Frequent ticks over music)
14:38Tom Chi: And that's what it looks like when we start planting.
14:41And then, two months later,
14:43you see, actually, we have over 90 percent that get to germination.
14:47And 14 months later, you can see the landscape is fully established,
14:52over 85 percent full establishing of the mangroves that were planted.
14:57Now, the scale of this technology and the scale that it's capable of
15:01is incredible.
15:03Just four people are able to go plant over 80 hectares of land,
15:07representing 120,000 mangroves being planted,
15:10and over 100,000 being established in one day.
15:17When you get to robotic scale on things,
15:19all of a sudden, then human action, human intention --
15:23and if we have good intentions, we can really multiply that
15:26in ways that can completely rewrite our landscapes.
15:30And I've worked with this company for about a decade at this point,
15:34and I got really inspired by them, you know,
15:36because they have not just restored mangroves,
15:38but they've restored 20 different terrestrial ecosystems
15:41on four different continents --
15:43dry land, inland, mountainous, you know, nearshore, all these sorts of things.
15:47And I got inspired, like, "Could we also do this below the water?"
15:52And I'm going to show you something that I founded
15:56and was the original electrical engineer for.
15:58And this is this is the Reefgen robot,
16:01which is basically its own kind of planting drone.
16:04Oh, it's already going to do some planting things, that's fun.
16:08And this robot line is the first in the world
16:12to plant live corals back into a coral reef.
16:14It's the first in the world to plant live seagrasses
16:17back into seagrass meadows.
16:19And it can also plant them in seed form as well.
16:22And this robot, you know,
16:25has been able to plant 10,000 seagrass seeds in a single day,
16:30which covers an entire underwater acre with one robot in one day.
16:35And the other thing that we did
16:36is we wanted to make sure that this robot was affordable enough
16:40that we could make a bunch of them, right?
16:42Because when I went around and I talked to people
16:44about the underwater robot that I was going to go build
16:47to restore these ecosystems,
16:49they're like, "You should budget, like, two million dollars for the robot.
16:52If you spent less, it's probably not going to do anything interesting."
16:55And I remember thinking more like 5,000 dollars,
16:59and we’re not quite there, but this is more like 10,000 dollars.
17:02And in the grand scheme of things, it's way, way less than two million.
17:05And the whole point is you want this to be an accessible technology
17:09to all the communities that have nearshore ecosystems to restore,
17:12whether they be coral, whether they be seagrasses.
17:15You want something like this to be also scalable
17:18from the CapEx perspective, right?
17:20Like, a single billionaire could spend 50 million dollars
17:24and have a fleet of 10,000 of these,
17:26and that is actually meaningful scale in terms of ocean restoration
17:30of all different types.
17:31I'll show you a little bit more here.
17:33So right here is a stake,
17:35because this is actually not a seed-planting end effector.
17:38That's a seedling-planting end effector,
17:40because there's actually two ways to plant seagrasses.
17:43You can plant it from seed,
17:45but then, there's other types of seagrasses
17:47that want to be planted as a sapling.
17:49And they want to grow rhizomatically,
17:50so they send out these little rhizomes laterally,
17:53and then the grasses grow up from the lateral rhizomes
17:56that are heading out.
17:57And this is basically a stake that we put the seagrass seedlings into,
18:04and then that feeds in through a tube into a hopper,
18:07and basically, bit by bit, this current layout
18:11is able to go and plant about half an acre of seedlings per day
18:15with just one robot.
18:17And our next version of it
18:18is going to be able to do an acre to an acre and a half in a day, per robot.
18:23So we are really kind of moving into this space
18:26where, by really digging into that mental model in a different way,
18:32instead of economy versus ecology,
18:34we start taking the best tools that we're using in the current economy,
18:38and robotics and AI,
18:39and intentionally using them to support ecology
18:43so that we're able to go build both a healthy planet
18:45and a healthy economy for the future.
18:47Thanks so much.
18:49(Cheers and applause)
18:51Thank you.