WEBVTT - Smart Talks with IBM: NASA and AI: Decoding Our Universe 0:00:00.160 --> 0:00:02.960 Hey everyone, it's Robert and Joe here. Today we've got 0:00:02.960 --> 0:00:05.240 something a little bit different to share with you. It's 0:00:05.280 --> 0:00:09.400 a new season of the Smart Talks with IBM podcast series. 0:00:09.720 --> 0:00:13.520 This season on Smart Talks with IBM, Malcolm Gladwell is back, 0:00:13.520 --> 0:00:15.760 and this time he's taking the show on the road. 0:00:15.880 --> 0:00:19.160 Malcolm is stepping outside the studio to explore how IBM 0:00:19.239 --> 0:00:22.880 clients are using artificial intelligence to solve real world challenges 0:00:23.079 --> 0:00:25.440 and transform the way they do business. 0:00:25.480 --> 0:00:29.560 From accelerating scientific breakthroughs to reimagining education. It's a fresh 0:00:29.600 --> 0:00:33.040 look at innovation in action, where big ideas meet cutting 0:00:33.120 --> 0:00:33.960 edge solutions. 0:00:34.320 --> 0:00:37.160 You'll hear from industry leaders, creative thinkers, and of course 0:00:37.280 --> 0:00:40.920 Malcolm Gladwell himself as he guides you through each story. 0:00:41.440 --> 0:00:44.680 New episodes of Smart Talks with IBM drop every month 0:00:44.680 --> 0:00:48.279 on the iHeartRadio app, Apple Podcasts, or wherever you get 0:00:48.320 --> 0:00:52.240 your podcasts. Learn more at IBM dot com slash smart Talks. 0:00:52.880 --> 0:00:56.920 This is a paid advertisement from IBM. 0:00:57.080 --> 0:00:59.880 If I were to go back, I don't know thirty 0:01:00.080 --> 0:01:04.200 years in Kenya. What's the difference between then and now 0:01:04.280 --> 0:01:07.880 in terms of tree cover. I'm talking to Philip thego 0:01:08.440 --> 0:01:10.080 Special Technology envoid. 0:01:10.080 --> 0:01:13.200 To the Kenyan president, let's speak as if you think 0:01:13.240 --> 0:01:16.280 about we aren elevent trol posts and previously we were 0:01:16.319 --> 0:01:20.120 more than twenty percent. So we are cutting trees more 0:01:20.160 --> 0:01:21.480 than we're planting them. 0:01:21.840 --> 0:01:25.559 In thirty years, Kenya last half its tree cover half. 0:01:26.280 --> 0:01:29.759 And here's why that matters. Kenya is a mountainous country. 0:01:30.280 --> 0:01:33.119 Dotted throughout the highlands are dozens of what canyons call 0:01:33.280 --> 0:01:39.479 water towers natural reservoirs, densely forested areas capable of absorbing 0:01:39.520 --> 0:01:41.760 the enormous amount of water that falls on the country 0:01:41.840 --> 0:01:46.440 during the rainy seasons. The tree roots and undergrowth secure 0:01:46.560 --> 0:01:50.200 and capture moisture, then slowly release it into the rivers 0:01:50.200 --> 0:01:53.560 that flow down into the country's low lying coastal areas. 0:01:54.200 --> 0:01:59.080 But in recent years the water towers have depleted, settlements 0:01:59.240 --> 0:02:03.040 have encroached on them, trees have been chopped down, thousands 0:02:03.080 --> 0:02:06.800 of acres cleared, the natural reservoirs ceased to hold nearly 0:02:06.840 --> 0:02:10.400 as much water, so now Kenya is prone to extremes. 0:02:10.800 --> 0:02:13.320 Too much water flowing down from the highlands in the 0:02:13.400 --> 0:02:17.280 rainy season and too little water left during the dry season. 0:02:17.600 --> 0:02:19.680 So you have a couple of hours of water then 0:02:19.680 --> 0:02:22.000 you have a couple of hours with no water that 0:02:22.080 --> 0:02:24.840 it tops off to be dry by the city authority. 0:02:25.040 --> 0:02:28.120 So that's the significance of the water towers we have 0:02:28.240 --> 0:02:29.360 when they cannot hold water. 0:02:29.800 --> 0:02:34.000 Kenya desperately needed to restore its water towers by planting 0:02:34.320 --> 0:02:37.799 as many trees as humanly possible. So in the fall 0:02:37.840 --> 0:02:41.680 of twenty twenty three, the Kenyan government took action. It 0:02:41.800 --> 0:02:46.359 started a national holiday, national Tree Growing Day, a day 0:02:46.400 --> 0:02:49.119 to allow the citizens of Kenya to go out into 0:02:49.160 --> 0:02:52.480 the forest to dominate the Kenyan countryside and plant as 0:02:52.560 --> 0:02:56.600 many trees as they can, and the government decided on 0:02:56.680 --> 0:02:57.200 a number. 0:02:58.080 --> 0:03:03.160 The presidents focus right around how to ensure that we 0:03:03.200 --> 0:03:05.959 do not lose more forests was in this very ambitious 0:03:06.000 --> 0:03:07.800 campaign around fifteen billion trees. 0:03:08.320 --> 0:03:11.400 That's right, fifteen billion with a B. 0:03:12.400 --> 0:03:14.679 So imagine that number will tell you the ambition, not 0:03:14.800 --> 0:03:17.400 as he tells you the deficit. It has to be 0:03:17.440 --> 0:03:19.760 fifteen billion in the next eight years. 0:03:20.080 --> 0:03:23.359 Fifteen billion trees over eight years averages out to more 0:03:23.360 --> 0:03:27.440 than five million trees per day. That's a lot of trees. 0:03:28.000 --> 0:03:30.600 But with such a massive goal, how can you track 0:03:30.639 --> 0:03:33.320 your progress? How do you know where to plant those 0:03:33.360 --> 0:03:36.480 trees so they'll have the most impact. How do you 0:03:36.600 --> 0:03:40.520 monitor where older trees are still being cut down? Well, 0:03:40.520 --> 0:03:44.200 the answer to those questions came from IBM and a 0:03:44.240 --> 0:03:50.680 little space agency called NASA. That's right, folks, Smart Talks 0:03:50.800 --> 0:03:54.960 is going to space. My name is Malcolm Glawell. You're 0:03:54.960 --> 0:03:58.480 listening to the latest episode of Smart Talks with IBM, 0:03:58.840 --> 0:04:01.600 where we offer our list a glimpse behind the curtain 0:04:01.720 --> 0:04:05.520 of the world of technology. In this season, IBM has 0:04:05.560 --> 0:04:10.440 gone inside elementary school classrooms, toured formulation labs at Loreel, 0:04:10.760 --> 0:04:14.640 and spoken with the fan development team at Scuderia Ferrari HP. 0:04:15.640 --> 0:04:20.159 In this episode, how IBM is partnering with NASA to 0:04:20.200 --> 0:04:24.760 build geospatial models using data from satellites to better understand 0:04:24.760 --> 0:04:30.640 our Earth and Solar system. 0:04:30.680 --> 0:04:39.119 Five four three two one zero all engine running liptoff. 0:04:39.160 --> 0:04:42.279 We have a liptoff thirty two minutes past the hour 0:04:42.640 --> 0:04:44.000 liftoff on Apollo eleven. 0:04:44.680 --> 0:04:48.320 IBM has worked on space related projects since before I 0:04:48.400 --> 0:04:49.000 was even born. 0:04:49.839 --> 0:04:51.479 Im all for man. 0:04:52.480 --> 0:04:55.760 A team of four thousand IBM engineers helped create the 0:04:55.800 --> 0:04:58.600 Saturn five rocket that took Neil Armstrong to the Moon. 0:05:00.360 --> 0:05:01.040 Buy up plate. 0:05:02.920 --> 0:05:05.520 And when I think of NASA. I tend to picture 0:05:05.520 --> 0:05:07.840 the moon landing, or the team of people back in 0:05:07.920 --> 0:05:12.000 Houston guiding the Apollo mission, or the Hubble telescope, or 0:05:12.040 --> 0:05:16.320 astronauts aboard the International Space Station. What I didn't think 0:05:16.360 --> 0:05:19.880 about until now are NASA's geographers. 0:05:20.680 --> 0:05:22.599 In order to go places, you need at map things. 0:05:23.080 --> 0:05:27.120 This is Kevin Murphy, chief Science Data Officer at NASA's 0:05:27.240 --> 0:05:28.599 Science Mission Directorate. 0:05:29.320 --> 0:05:32.320 But I think that there's an assumption that NASAs all 0:05:32.360 --> 0:05:36.159 about rockets and astronauts, and certainly that's a really large 0:05:36.160 --> 0:05:37.480 part and important part of NASA. 0:05:38.279 --> 0:05:41.640 NASA sends people to space and looks out of the stars, 0:05:42.120 --> 0:05:45.720 but NASA also looks down at the Earth. The agency 0:05:45.720 --> 0:05:52.640 has about one hundred and fifty satellites that use radar, lightar, landset, Aquaterra, cloudset, AURA, 0:05:53.000 --> 0:05:57.920 low Earth orbit, Medium Earth orbit, geostationary orbit, on and on. 0:05:58.600 --> 0:06:02.760 In one sense, NASA makes hardware to build rockets and 0:06:02.800 --> 0:06:06.279 spacecraft and all those satellites that circle the Earth. But 0:06:06.400 --> 0:06:12.640 fundamentally NASA also collects data. It's scientists and engineers people 0:06:12.640 --> 0:06:16.160 like Kevin want to make the best use possible of 0:06:16.200 --> 0:06:20.280 all the information gathered by all those many dozens of instruments. 0:06:20.920 --> 0:06:24.640 Right now, we gather around twenty five petabytes of new 0:06:24.800 --> 0:06:27.920 observational data per year. In the next couple months, we're 0:06:27.960 --> 0:06:33.920 about to launch a high resolution global radar. When that launches, 0:06:34.360 --> 0:06:37.400 will double how much we collect every year to about 0:06:37.440 --> 0:06:39.000 fifty petabytes of information. 0:06:39.680 --> 0:06:44.080 Actually, since we recorded this conversation, NASA launched that global radar, 0:06:44.279 --> 0:06:48.640 what they call NYSAR. So NASA is already generating new 0:06:48.720 --> 0:06:52.159 data at the rate of fifty petabytes each year. To 0:06:52.200 --> 0:06:55.679 put that in perspective, a single petabyte could hold about 0:06:55.680 --> 0:07:00.159 five hundred billion pages of standard printed text. You know, 0:07:00.200 --> 0:07:03.000 can anyone sort of apply to use this data. 0:07:02.880 --> 0:07:05.400 Is they don't even have to apply. It's free and 0:07:05.480 --> 0:07:09.400 open data. It advances how we understand what we do 0:07:09.480 --> 0:07:12.920 on Earth and how we see ourselves within the universe. 0:07:13.280 --> 0:07:16.200 People can take it for so many different downstream applications. 0:07:16.440 --> 0:07:18.720 So you can go to our websites today, you can 0:07:18.920 --> 0:07:23.160 search through our tools and you can download information from 0:07:23.240 --> 0:07:26.280 the Mars rovers, you can download information from the Lunar 0:07:26.320 --> 0:07:29.480 Reconnaissance Orbiter or any of the Earth Science Data satellites. 0:07:29.760 --> 0:07:33.120 And give me an example of a really cool application, 0:07:33.840 --> 0:07:36.120 a really cool use that someone I don't know in 0:07:36.200 --> 0:07:39.080 academic or whatever has used your data for it is there? 0:07:39.120 --> 0:07:42.360 It okay. So one of the really kind of cool 0:07:42.480 --> 0:07:46.320 but unexpected observations that we had is that we launched 0:07:46.360 --> 0:07:50.720 a pair of satellites in their early two thousands called Grace, 0:07:51.120 --> 0:07:54.000 and these satellites orbit the Earth and they can measure 0:07:54.120 --> 0:07:56.840 very precisely the distance that they're away from each other 0:07:56.840 --> 0:07:59.880 as they orbit the Earth, and as you go into gravity, 0:08:00.440 --> 0:08:03.280 you can actually see a satellite accelerate and the other 0:08:03.320 --> 0:08:07.280 one accelerate after it, right, And using that information, we 0:08:07.320 --> 0:08:10.640 were trying to map kind of the gravity fields of Earth. 0:08:10.880 --> 0:08:13.280 What what they found is that they can actually map 0:08:13.440 --> 0:08:16.960 below kind of the mass of Earth to where water 0:08:17.000 --> 0:08:20.440 storage is. For instance, so aquifers, right, so you can 0:08:20.840 --> 0:08:25.600 monitor through gravity how much water is being depleted or 0:08:25.760 --> 0:08:28.880 added to an aquifer or the density of glaciers. 0:08:29.440 --> 0:08:32.800 So, just to back up for a moment, the presence 0:08:32.960 --> 0:08:37.280 and density of water deposits below the Earth's surface have 0:08:37.360 --> 0:08:42.640 an effect on gravitational fields that are being measured in space. 0:08:42.679 --> 0:08:43.040 Correct. 0:08:43.559 --> 0:08:47.120 Yeah, And so does that tell you presume you learn 0:08:47.240 --> 0:08:49.160 things like where there's an aquifer where you didn't think 0:08:49.160 --> 0:08:50.600 there was an aquifer. 0:08:50.640 --> 0:08:52.920 Or if it's being depleted faster. 0:08:53.240 --> 0:08:53.480 Yeah. 0:08:53.640 --> 0:08:56.280 Yeah, So who's using that kind of data? 0:08:56.800 --> 0:08:59.800 All sorts of different organizations, whether they're you know, NG 0:09:00.640 --> 0:09:03.960 or government agencies or people that are planning a large 0:09:03.960 --> 0:09:04.960 agricultural product. 0:09:05.040 --> 0:09:07.200 How did you Was that an intentional decisisse? 0:09:07.200 --> 0:09:08.520 It wasn't. It was accidental. 0:09:09.240 --> 0:09:15.800 It was accidental. NASA has assembled a historically unprecedented mountain 0:09:15.800 --> 0:09:19.640 of data about the physical world, free and open to anyone, 0:09:20.040 --> 0:09:22.800 and the possibilities for how that information can be used 0:09:22.880 --> 0:09:28.679 are so vast that even NASA is still uncovering them. 0:09:29.080 --> 0:09:31.760 When I was a kid, I loved legos. I had 0:09:31.800 --> 0:09:34.880 a huge bin full of them. At the time, legos 0:09:34.960 --> 0:09:38.520 were really just colored bricks of various sizes. They weren't 0:09:38.520 --> 0:09:41.440 as complicated as they are today. And what I realized 0:09:41.480 --> 0:09:44.040 even then was that there were more possibilities in a 0:09:44.080 --> 0:09:46.960 box of legos than I could ever imagine on my own. 0:09:47.800 --> 0:09:49.440 I played with my brother and he would show me 0:09:49.480 --> 0:09:51.760 something that hadn't occurred to me, And I go to 0:09:51.760 --> 0:09:54.080 my friend Bruce's and see that he was off on 0:09:54.120 --> 0:09:57.600 some legos tangent that I'd never even thought of, like 0:09:57.640 --> 0:10:00.400 a cool bridge or a castle or a truck. I 0:10:00.559 --> 0:10:04.080 use legos one way. Bruce used his legos in a 0:10:04.160 --> 0:10:09.640 completely different way. NASA's data treasure trove is like a 0:10:09.760 --> 0:10:13.080 very very big box of Legos. And here's the question. 0:10:13.640 --> 0:10:18.800 With so much data, containing so many possible connections, could IBM, 0:10:19.080 --> 0:10:24.960 and specifically IBM's artificial intelligence help NASA scientists uncover patterns 0:10:24.960 --> 0:10:28.400 and connect systems in a way they've never done before. 0:10:30.880 --> 0:10:33.360 Everything started with a question, right. 0:10:33.400 --> 0:10:37.520 I'm talking to one Bernabe Moreno, director of IBM Research 0:10:37.559 --> 0:10:38.040 in Europe. 0:10:38.840 --> 0:10:42.840 As we advance AI, we have new tools to understand 0:10:43.520 --> 0:10:46.560 the around this, understand the world, understand the language, and 0:10:46.720 --> 0:10:49.760 understand our planets. And the question that we were asking 0:10:49.800 --> 0:10:52.920 ourselves was all these new advances that we see in language. 0:10:52.960 --> 0:10:56.280 It was a post GPT moment. Could we apply the 0:10:56.360 --> 0:10:59.800 same idea and the same architecture and technology to a 0:11:00.240 --> 0:11:01.240 dour planets? 0:11:01.760 --> 0:11:05.120 The advent of AI created a new opportunity. What if 0:11:05.120 --> 0:11:08.680 all of NASA's mountain of data could be organized, analyzed, 0:11:09.000 --> 0:11:14.280 understood by artificial intelligence. The original idea was to create 0:11:14.360 --> 0:11:18.000 a geospatial foundation model for the Earth and from there 0:11:18.440 --> 0:11:23.200 create additional specialized models for other scientific priorities of NASA, 0:11:23.920 --> 0:11:27.600 and finally quit an AI system that can understand all 0:11:27.720 --> 0:11:31.480 the data across those specialized models in order to uncover 0:11:31.600 --> 0:11:36.760 hidden insights and relationships. Together, these models could unlock an 0:11:36.920 --> 0:11:41.520 infinite number of potential applications. I asked Kevin Murphy at 0:11:41.600 --> 0:11:44.840 NASA about the beginning of these Earth models. 0:11:45.440 --> 0:11:48.280 Has some colleagues, and we were investigating a number of 0:11:48.320 --> 0:11:53.560 different avenues of using AI with our data, but also 0:11:53.800 --> 0:11:56.400 kind of the management and stewardship of the data, so 0:11:56.480 --> 0:11:58.600 not only like the observations, but how we make it 0:11:58.640 --> 0:12:02.320 available to people, make it discoverable. And they said, hey, 0:12:03.000 --> 0:12:05.559 we see these transform architectures. We think that they can 0:12:05.600 --> 0:12:09.760 be applicable to some of the sequential observations that we make. 0:12:10.320 --> 0:12:12.400 We'd really like to work with IBM on that. And 0:12:12.440 --> 0:12:16.679 I was like, I'm really skeptical, but because I hadn't 0:12:16.760 --> 0:12:22.880 seen those types of tools really produce results that were 0:12:23.520 --> 0:12:26.520 commensurate with the amount of effort you put into them, right, 0:12:26.600 --> 0:12:28.960 So we were getting some really good results and deep 0:12:29.040 --> 0:12:32.240 learning approaches, but they took a lot of effort. 0:12:32.559 --> 0:12:34.160 But Kevin came around quickly. 0:12:35.000 --> 0:12:39.680 When we typically develop a new data product or an algorithm, 0:12:40.160 --> 0:12:43.880 it takes anywhere from you know, twelve months, eighteen months, 0:12:43.880 --> 0:12:49.280 twenty four months to go from data and hypothesis to 0:12:49.480 --> 0:12:53.439 results which is validated. We were able to get approximately 0:12:53.480 --> 0:12:58.560 the same precision for some well known types of benchmarks 0:12:58.960 --> 0:13:01.040 with and I think it was about four four months. Oh, 0:13:01.040 --> 0:13:02.120 instead of starting the work. 0:13:02.200 --> 0:13:05.640 Yeah, yeah, so was it happened faster than you thought, 0:13:05.840 --> 0:13:10.480 much faster. In twenty twenty three, IBM and NASA launched 0:13:10.520 --> 0:13:15.079 a foundation model trained on NASA's harmonized landset sentinel to 0:13:15.320 --> 0:13:19.720 satellite data across the continental United States. They named the 0:13:19.760 --> 0:13:24.600 model Prithvi, the Sanskrit word for Earth. The first version 0:13:24.600 --> 0:13:28.920 of Prithvy used only Earth observation images and just that 0:13:29.120 --> 0:13:32.480 was enough to totally change Kevin's idea of what foundation 0:13:32.679 --> 0:13:37.320 models could do. But they didn't stop there. IBM and 0:13:37.400 --> 0:13:40.760 NASA were encouraged at how well Prithvy worked for Earth 0:13:40.880 --> 0:13:45.200 observation tasks, so they decided to create a more complex 0:13:45.280 --> 0:13:49.000 version of Prithvy that could understand whether and climate data. 0:13:49.960 --> 0:13:53.080 They hoped this new version of Prithvi would allow researchers 0:13:53.120 --> 0:13:55.920 to answer new questions about the Earth, from short term 0:13:55.960 --> 0:14:00.680 weather forecasting to longer term climate effects. Imagine you have 0:14:00.720 --> 0:14:05.360 a map of all the different temperatures, pressures, clouds, rainfall, 0:14:05.640 --> 0:14:09.720 and more from around the globe. With this map, IBM 0:14:09.800 --> 0:14:13.960 and NASA could implement advanced tasks. They could track the 0:14:13.960 --> 0:14:16.599 formation of El Nino or predict how the path of 0:14:16.640 --> 0:14:19.720 a hurricane would change if the ocean temperature went up 0:14:19.800 --> 0:14:21.680 by half a degree. 0:14:21.960 --> 0:14:25.040 I would always remember this moment was when we created 0:14:25.080 --> 0:14:30.160 the Weather and Climate Foundational Model. The senior methodologist of NASA, 0:14:30.640 --> 0:14:33.480 it was like, I cannot believe that it has changed 0:14:33.680 --> 0:14:35.800 the way I think about the AI And ever since 0:14:35.840 --> 0:14:38.800 he's been kind of preaching with this A samples. 0:14:38.960 --> 0:14:41.440 One and his team then took the model and decided 0:14:41.480 --> 0:14:45.560 to test it, really tested. They took away ninety nine 0:14:45.600 --> 0:14:48.840 percent of the data points and ran the experiment again. 0:14:49.520 --> 0:14:51.600 What they were trying to figure out is if the 0:14:51.640 --> 0:14:54.680 model had learned enough about the basic principles of the Earth, 0:14:55.080 --> 0:14:58.040 the underlying physics of the way the planet works, to 0:14:58.160 --> 0:15:01.480 fill in the blanks on its own with just one 0:15:01.560 --> 0:15:05.120 percent of the original data, would it still be accurate 0:15:05.200 --> 0:15:11.760 in its predictions. What happened The model crushed it so 0:15:12.000 --> 0:15:14.440 it was able to extrapolate on the basis of one 0:15:14.480 --> 0:15:18.160 percent of the data what the entire picture looked like yes, 0:15:19.600 --> 0:15:23.280 because pre learned everything right, Yeah, it learned the kind 0:15:23.280 --> 0:15:27.920 of principles of exactly. Yeah. Oh wow, that's very very impressive. 0:15:28.040 --> 0:15:30.880 So at that moment when you realize you could do that, 0:15:32.840 --> 0:15:35.600 and just curious about your emotional I mean, did you 0:15:35.680 --> 0:15:37.240 jump up and down? What did you do that? 0:15:37.320 --> 0:15:40.760 So he's like, wow, it was a very emotional meeting 0:15:40.800 --> 0:15:46.160 because you know, having this person say now I'm convinced right, Yeah, 0:15:46.280 --> 0:15:49.160 it was kind of a quite a special moment. These 0:15:49.160 --> 0:15:50.840 moments make your life as a researcher. 0:15:52.280 --> 0:15:55.480 Ibm And as a launch prith Fee for Weather and 0:15:55.520 --> 0:15:59.000 Climate in twenty twenty four, and while ibm And as 0:15:59.000 --> 0:16:02.920 a scientist could use Privy to run interesting experiments, they 0:16:02.960 --> 0:16:06.040 were even more excited about how Prithy could help people 0:16:06.520 --> 0:16:13.960 in the real world. So let's go back to Kenya 0:16:14.280 --> 0:16:18.560 Ambassador Philip Diego and the country's great tree planting project. 0:16:19.480 --> 0:16:22.440 So on those initial months there was a massive effort, 0:16:22.520 --> 0:16:25.560 including a couple of national holidays. 0:16:25.400 --> 0:16:26.400 For tree planting. 0:16:27.280 --> 0:16:30.760 Yes, where the entire cabinet was sent. 0:16:31.080 --> 0:16:33.640 Ah, did you plant trees as I did? 0:16:33.720 --> 0:16:35.600 Oh my god, I said, The entire cabinet plus someone 0:16:35.800 --> 0:16:36.600 we have to be seen. 0:16:37.040 --> 0:16:38.840 Are you good at the planet two weeks ago? 0:16:39.200 --> 0:16:41.080 Well, it's very easy to go hole put a tree 0:16:41.160 --> 0:16:42.760 in the ground show. 0:16:42.800 --> 0:16:46.600 Well wow, what planting a tree is easy? But remember 0:16:47.040 --> 0:16:51.960 it has to happen fifteen billion times. IBM research has 0:16:52.000 --> 0:16:56.320 been operating in Nairobi since twenty thirteen, and what Kenya wanted, 0:16:56.440 --> 0:17:00.880 at least in the beginning was straightforward. Fee model that 0:17:00.960 --> 0:17:04.200 IBM and NASA built could be used to essentially make 0:17:04.240 --> 0:17:08.520 the world's greatest map, and Kenya, with IBM's help, could 0:17:08.680 --> 0:17:11.719 use that model to make the world's greatest map of Kenya. 0:17:12.800 --> 0:17:15.000 The first step was to lay a grid across the 0:17:15.080 --> 0:17:19.240 topography of the country, break the forest into manageable bite 0:17:19.240 --> 0:17:22.560 sized pieces, each of which could be analyzed separately. 0:17:23.400 --> 0:17:25.679 So, because our forest is massive when you look at 0:17:25.720 --> 0:17:28.360 it in terms of green hite, but only lay it, 0:17:28.600 --> 0:17:31.240 you're able to break it into pieces, like into boxes. 0:17:31.480 --> 0:17:35.119 And for us that was important because then it's easy 0:17:35.200 --> 0:17:37.960 to tackle it when it's in a greed system than 0:17:38.119 --> 0:17:40.600 just as a massive forest. So that was also what 0:17:41.280 --> 0:17:42.680 the model was able to do. 0:17:43.119 --> 0:17:46.680 Then the model painstakingly sorted through each of those boxes 0:17:47.080 --> 0:17:50.280 and look for what Philip calls hotspots, so. 0:17:50.200 --> 0:17:52.879 You can see, for example, very quickly, which other areas 0:17:52.880 --> 0:17:56.080 are being eroded very fast, and that you need to 0:17:56.160 --> 0:17:59.359 quickly protect. Yeh, because you sometimes and that's where you 0:17:59.359 --> 0:18:01.159 want to target, right, I mean it's not possible to 0:18:01.160 --> 0:18:02.800 do everything at the same time. 0:18:02.960 --> 0:18:04.760 Do you have a definition of a hotspot? And how 0:18:04.760 --> 0:18:08.119 many hotspots are there according to that definition? H, there 0:18:08.160 --> 0:18:08.480 are a lot. 0:18:08.560 --> 0:18:11.880 So we have more than forty water towers, and I'll 0:18:11.880 --> 0:18:14.880 tell you all of them have hotspots. And the hot 0:18:14.880 --> 0:18:19.040 spots in my definition areas that are being degraded faster 0:18:19.160 --> 0:18:22.160 and in a very unusual way. Right. You can literally 0:18:22.200 --> 0:18:25.600 see how human activity is seriously degrading that particular area 0:18:25.960 --> 0:18:28.160 that if you do not have a direct intervention, we'll 0:18:28.160 --> 0:18:31.560 lose the entire forest. So that's the hot spot for us, 0:18:32.119 --> 0:18:34.080 because you think about cutting one hundred trees a day 0:18:34.160 --> 0:18:36.159 and cutting a million trees a day, So that's a 0:18:36.200 --> 0:18:39.240 hot spot. You want to look at places where there's 0:18:39.480 --> 0:18:43.399 just unusually high activity of deforestation. 0:18:43.480 --> 0:18:45.600 In a hot spot. The size of each box in 0:18:45.640 --> 0:18:48.560 the grid was ten by ten meters, about half a 0:18:48.560 --> 0:18:52.520 tennis court. That's how closely they were examining the forest, 0:18:53.560 --> 0:18:57.800 so very crudely. The model ingests all of this satellite 0:18:57.840 --> 0:19:01.879 data and it helps you answer some very specific questions 0:19:01.960 --> 0:19:06.560 like where should we prioritize our tree planning efforts, which 0:19:07.000 --> 0:19:11.600 areas down to an extraordinary level of specificity, are eroding 0:19:11.720 --> 0:19:15.560 most quickly. You know, all those kinds of practical questions 0:19:15.600 --> 0:19:17.159 about how to direct your strategy. 0:19:17.480 --> 0:19:19.560 So if you think about a smart forest, right, and 0:19:19.600 --> 0:19:22.200 that's really for us according it smart fencing, smart forests, 0:19:22.200 --> 0:19:25.960 everything that's smart because of AI. If you think about 0:19:26.680 --> 0:19:29.359 your usual what you can see with your eyes and 0:19:29.400 --> 0:19:32.439 then the satellite layer which just zooms in and you 0:19:32.480 --> 0:19:35.320 see green. So what the model has been able to 0:19:35.320 --> 0:19:37.639 do is to create a smart layer, right, and then 0:19:37.760 --> 0:19:42.000 that smart layer you can actually see many things, from analytics, 0:19:42.000 --> 0:19:44.879 to the greeds, to a dashboard, one a lot. So 0:19:45.000 --> 0:19:49.159 able to layer to those blocks, you can quantify degradation 0:19:49.240 --> 0:19:52.960 by blocks. You can match integrations, you can match reforestation. 0:19:53.440 --> 0:19:55.480 I asked Philip to imagine what it would have been 0:19:55.560 --> 0:19:58.440 like to attempt the tree planting project in an era 0:19:58.640 --> 0:20:04.920 before AI. His answer was, plant fifteen billion trees, restore 0:20:04.920 --> 0:20:09.720 the water towers. Impossible. With Prithvie on Kenya's side, though, 0:20:10.160 --> 0:20:13.520 it's really happening. What should be clear by now is 0:20:13.560 --> 0:20:16.200 how versatile Prithvie can be. Want to know how to 0:20:16.280 --> 0:20:20.320 combat deforestation, Prithvy can model that. Want to know when 0:20:20.320 --> 0:20:22.479 the best time in the year to plant your crops is, 0:20:23.080 --> 0:20:27.280 prithvy can help predict that too. Last year, six months 0:20:27.320 --> 0:20:31.800 after IBM started helping Kenya with reforestation, Kenya needed Prithvy's 0:20:31.800 --> 0:20:34.720 help on something else. And it was an emergency. 0:20:35.560 --> 0:20:38.280 So something was happening in the world that we sort 0:20:38.320 --> 0:20:41.040 of had these flats that we didn't expect. 0:20:41.440 --> 0:20:44.080 In the spring of twenty twenty four, Kenya was hit 0:20:44.119 --> 0:20:48.360 with thunderstorms and torrential rain, days and days of it. 0:20:49.200 --> 0:20:51.440 And so I got a call from the Red Cross, 0:20:51.600 --> 0:20:55.000 the one of my friends, and they're like, Ambassador, we 0:20:55.080 --> 0:20:58.120 need a little bit of help on how we deal 0:20:58.240 --> 0:21:01.240 with response because what we see is unusual, right, because 0:21:01.280 --> 0:21:04.119 normally you would only have one area. All of a sudden, 0:21:04.240 --> 0:21:07.479 we had an entire country flooding. In April, we had 0:21:07.480 --> 0:21:12.679 about three eight hundred kilometers square kind of total land flooded, 0:21:12.880 --> 0:21:16.600 which is unusual for Kenyon. And so when I got 0:21:16.600 --> 0:21:18.879 this call, we were like, Okay, there's someone could do 0:21:18.960 --> 0:21:21.800 with IBM. We only did one function for the trees. 0:21:21.840 --> 0:21:24.440 It was actually a climate model, and we said, can 0:21:24.480 --> 0:21:29.880 we use this to help us better respond to floods 0:21:30.440 --> 0:21:33.520 And so that was how we started having this discussion 0:21:33.560 --> 0:21:37.520 with IBM in terms of repurposing the model to help 0:21:37.720 --> 0:21:41.320 us deal with this new challenge around floods. 0:21:41.880 --> 0:21:47.880 Again, Prithvy is versatile. Prithe could use everything it knew 0:21:47.880 --> 0:21:52.240 about the land, the forests and infrastructure to analyze how 0:21:52.359 --> 0:21:56.560 and where and when floods would occur. The Kenyan government 0:21:56.600 --> 0:21:59.000 could then use the model to help the Red Cross 0:21:59.119 --> 0:22:02.640 organize its respet pants, show areas that needed to be 0:22:02.720 --> 0:22:06.200 evacuated or safe places where the Red Cross could set 0:22:06.280 --> 0:22:10.119 up camps. That information was invaluable. 0:22:11.080 --> 0:22:13.639 Historically, what has happened is that they would set up 0:22:13.720 --> 0:22:18.919 camp based on population congregation right where people assembly is 0:22:18.920 --> 0:22:21.359 where they set up a camp, not based on any data, 0:22:21.440 --> 0:22:25.000 right simply because people are there, they will come there 0:22:25.000 --> 0:22:28.560 to provide services and emergency response. What we realize is 0:22:28.600 --> 0:22:31.080 that that model doesn't work. So what we've been able 0:22:31.119 --> 0:22:33.560 to do with IBM is be able to to sort 0:22:33.560 --> 0:22:37.120 of give Red Cause very specific locations or options where 0:22:37.160 --> 0:22:39.520 to set up camps. So if people come here, just 0:22:39.520 --> 0:22:43.159 tell them no, move here. That's the safe place you 0:22:43.240 --> 0:22:44.879 really want to go. So I think for me that 0:22:45.000 --> 0:22:46.959 was really amazing. So we're calling them were a very 0:22:47.000 --> 0:22:49.159 funny word for it, flood assembly points. We always have 0:22:49.240 --> 0:22:51.760 fire for assembly points, but now we can say we 0:22:51.840 --> 0:22:57.280 have literally flat assembly points that are safe or citizens. 0:22:57.680 --> 0:23:02.240 That's fascinating. So the model has ingested this incredibly granular 0:23:03.080 --> 0:23:09.679 picture later of the topography and weather patterns of Kenya. 0:23:09.800 --> 0:23:13.359 It's just giving you a set of useful predictions about 0:23:13.400 --> 0:23:15.360 how you should shape your response. 0:23:16.320 --> 0:23:18.800 Yes, and what we did remember is that, as I said, 0:23:18.840 --> 0:23:22.760 it was a full multistate called capability. What IBM gave 0:23:22.840 --> 0:23:25.560 us was a base map. We didn't have that before 0:23:25.920 --> 0:23:28.480 and a base model. So you cannot have these layers 0:23:28.520 --> 0:23:29.919 up on layers, up on layers to be able to 0:23:29.920 --> 0:23:31.960 make intelligent decisions. 0:23:35.720 --> 0:23:38.680 Throughout my reporting on this episode, I've been really impressed 0:23:38.680 --> 0:23:41.720