#36 - Bill Perkins - Space and The Future of Satellite Imaging

0:00Today we're talking satellite imagery. And what better location to do this than on a boat floating just off the coast of SpaceX's Starbase, watching the launch of Starship Live with Bill Perkins, who has founded not one, but two game-changing satellite imagery companies. As you'll see, it's a blast. What used to be considered classified spy-level observation is now commonplace. Teenage boys can download it. I can tell you where the meth lab is. That's how good it is. We don't want to get to the point where people are tracking their significant others. I think there may be concerns about tracking people.

0:35But now we're stuck. Are we actually stuck? A lot of people are concerned about climate change, and it's not like I'm not concerned about climate change. But I can tell you this. If the ocean dies, we all die. It is our food source. It is a significant source of oxygen. We do not get our oxygen from the Brazil rainforest. I think this is the issue of our time. Oh! Give me, give me, give me.

1:12This is a beautiful day. A little bit windy. I hope it doesn't hold up the launch. But we're getting to see Starship 6, right? This is the 6th. We certainly are. The 6th, right? We're about to get a little sliver of history. So, yeah, I'd love to understand a little bit more. Because, I mean, I'm fully sold on the fact that space is a win-win. because one of the things that creates win-lose dynamics is too much scarcity. And so you need to find ways to unlock more abundance. And where could there be more abundance than space? The problem when we mine the Earth is that we are almost certainly, every time we do that, screwing up some kind of ecosystem.

1:49There's some sort of cost. Exactly. Whereas if you're out in space doing that, you can, as far as we can tell, extract with zero ethical dilemmas or minimal ethical dilemmas. A lot of people will agree, I'll be dead by that point. But just access to information allows us to make better decisions here on planet Earth. And so that will release some of the abundance here and maybe fix some of the wrongs that we're doing right now. Let's see whether they're... Let's start it. Oh my gosh. Right away. On time. Oh. This is better than the New York City subway.

2:24Wow. Get ready for the sound. Get ready for the sound. Oh my! Gimme, gimme, gimme. Watch, you must watch. Not here yet. Wow. You... Here it comes.

2:43Yes!

2:56Yes! Yes! Yes! Yes! Yes! Yes!

3:09Stage separation coming out soon! Not yet, right? Should be. Two dots. Or actually we would see one. And one thing flying off the edge. Check it out for stage separation. There it is! There it is! Yee-haw!

3:31Got em. So the bright one. Got em. The bright one is the starship now. That which is going off to do an orbit around the Earth. Right. And the booster you have to find. What a beautiful day. Wow. No cloud in the sky. There is only this one that we made. So we just saw a rocket take off. It was amazing. How many of this has now been for you? Like three? Three or four? Yeah, four. This is like, yeah, my fifth now. I cry every time. It says something about seeing the absolute bleeding edge of physics and engineering. Like, this is the actual edge of humanity's capabilities right there in, like, an iterative process.

4:17And there's just nothing more fucking beautiful. I think earlier, though, before it was taking off, we didn't expect it to just go that fast. It did not. But we were talking about, you were talking about the long-term benefits of space, and I was talking about the short-term benefits that we get immediately from knowing more about our world. And I think information is power, and we can use that information for lots of good here, and unlock, you know. That's what I love what you're doing with Sky-Fi, because people don't realize that, I mean, I remember me, like I've always wanted to like, look at satellite imagery real-time.

4:51I'm interested in the rates of deforestation, in the Amazon rainforest. But the access to satellite imagery, I always assumed was just completely unavailable to someone like me who's not part of a university or is not part of some kind of formal institution or I'm not a member of the military or whatever. And you guys are actually democratizing the information about the planet of Earth. We make it trivial. We make it like looking at a TikTok video. And so now... It's like the Uber. It's like you're providing an Uber service to satellite imagery. And what people don't appreciate is like so many of our problems are like, they're so complex.

5:27We need as many perspectives and like viewpoints and just as many brains on the problem as possible. And by democratizing that information, now we're actually giving ourselves the best shot at solving these issues. Oh, 100%. We get the hive mind working. So all the use cases, I'm not bright enough to think of all the use cases. We're closer to shore. But they're out there. Oh, we're going to have to pull anchor because we're closer to shore. But Igor, you got to do your job. Okay. Just cruising along here. But we're talking about unleashing the hive mind on the world's problem with new data. Data that was inaccessible except for the richest of companies or governments.

6:05And now what used to be considered classified spy level observation is now commonplace. Teenage girls can download it. Teenage boys can download it. All types of students, universities. There's other things to download than other things to download. Yeah, there's other things to be downloaded. I think these things and they can look at these problems whether it be erosion, deforestation, city planning, rescue, environmental issues, economic issues, everything. So talk me through some of the, because one of the cool things you guys are doing, it's not just the visual spectrum that you can use, you know, a visible light to take imagery.

6:45You have like what's called hyperspectral. Can you explain a bit what that is? So our eyes obviously are red, green, blue sensors, and we can see some shades of that, but a computer can see every shade of red. So it doesn't see one type of red. Maybe we see five types of red, but a computer will see 200 shades of red, 200 shades of blue. So it's quite a granularity. Yes. And so when you look at an apple, you are actually a chemical detector. You can tell whether that apple is good or bad based on the color of the apple. If it's brown, the gaps in the chemicals, right, as it transitions, right, that's a chemical sensor.

7:27And so imagine your resolution and a computer's memory to basically be able to chemically identify every single pixel. I'm trying not to hit anybody while we're talking here. We got the cops there.

7:45So when we use things like computers and their sort of supervision, their ability to chemically fingerprint every pixel, we now have a barcode reader from space. So Planet Labs just released, along with Carbon Mapper, released a satellite that is a hyperspectral satellite, and it's mapping every single methane plume in the world. And they've released this publicly. and they've already alerted a company to a methane leak and they've remediated it and it was a significant amount of gas and there are so many methane leaks around the world from oil and gas wells or from maybe uh even uh garbage dumps etc and now we would have just going on oblivious to this this information was we just didn't know unless you walk by and you can't smell methane right at the wellhead it's colorless it's odorless and now we see them all and now this issue is being addressed and methane is I think 26 to 30 times more heat trapping than CO2 and so this is one way that our ability to get payloads up into space cheaply is paying dividends right now.

8:53What are the um because presumably there must be like some areas that are very sensitive you're not allowed to take imagery of like how how do you navigate that like or is it is there they said actually nothing like you know for example china i imagine they're not thrilled at the idea of americans being able to take satellite imagery of their country so china's part of the commercial market but one of the things chinese will say is is that you can't image over china with our satellites okay but they'll let you image other places and sometimes they won't let you uh image over conflict zones because they may have they don't want they don't want people doing uh bad actors doing things in conflict zones or they may have a stake in this conflict etc and each operator has its own rule sets and the u.s government has its rule sets the u.s government just sets a resolution limit and then the operators set their own limits depending on what country they're from or on their own ethics like panel has its own ethics guideline on what they will allow people to image obviously they will make some exceptions for news organizations and reporting but there's a there's a little bit of a vetting process so with any tool it can be used as a weapon whether it's a hammer a saw satellite imagery or even education right so So people are looking at ways to mitigate that being used as a weapon against other people.

10:10Have you personally run into any examples where you're like, I feel like these people are trying to use our platform for ill? We have a pretty stringent vetting platform and checks for bad actors, but any system can be gained, right? Like Israel has satellites. They don't want Iranians using their satellite imagery to image Israel to then target some sort of infrastructure or facility. but it's you know the cat's out of the bag right in terms of satellite imagery and so and it's not necessarily the best tool for doing nefarious things right nation states already have that capability and to use satellite imagery for let's say some sort of nefarious individual basis a drone is better right right you don't you don't need 25 square kilometers like exactly 2 000 feet with a high-res camera and like high zoom or a small camera right like a small a small plane in a camera just for the resolution limit so it's actually one is it's it's asymmetrically tipped in the in the direction of benefits that doesn't change the paranoia of nations and other people for for competitive reasons and we basically follow all the laws and we follow the ethics of the of our providers um there are far far far far uh more uh positive things to use satellite imagery for that will have massive benefits to to study i was just recently talking to someone about using satellite imagery to help track whales right so whales have to surface uh they're there and they're visible from uh satellite imagery and so instead of like wasting a bunch of gash running around or getting next to the acoustic sensor you can tell them here they are this is where they're going this is the route and we've tracked them along this route and then you can then protect those migratory patterns, maybe give warnings to shipping vessels, right?

11:56So that's just one environmental of many that I'm thinking of. Also crop yields, whether or not there's a crop infestation, using hyperspectral to look at the chemical analysis of your crops to see, hmm, maybe you have a worm infestation, maybe it's overwater, maybe it's underwater, right? There's so many use cases, and I'm being repetitive, that one person cannot possibly think of them all. If you remember the beginning days of the internet, if you look at any eclipses, like, well, what can we use this for? And they would be like, you can read the newspaper and check sports scores, right? Right.

12:28It was limited by imagination. It wasn't that they were wrong. It's just that it was so myopic and so limited, right? And now the internet is used for everything, right? And so this is, that was access to terrestrial data that was on the planet Earth. This is access to data about the Earth in real time. So I think there is many, many wonderful use cases that are going to unfold the more people have this data. And one of the main things, one of the main drivers of coordination failures, these collective action problems, is usually because there's an asymmetry of information. Correct. Or a lack of flow of information between the systems.

13:04So it makes it hard on people to coordinate. if there's actually accessible, you know, for people who are skeptics of the rate of deforestation or desertification or whatever, you know, latest environmental tragedy that's unfolding, it's like, oh, you don't believe me? Go have a look yourself. Go order the, like, you know, there's the imagery. And, like, it's, it makes it, it's far easier to promote Let me add to that. It's nice to show, here it was then, look at it now, right? A picture's worth a thousand words. Do I need to read this long report? Or do I need to go, oh my gosh, look what's happened here.

13:34This is a disaster. We need to do something about it. there's a visceral attachment to it because not only can we tap into people's um intellect but we can tap into their emotions about this is a travesty look what happened to this beautiful area this deforestation right and so i i think that's a powerful narrative to to put out there for people to see is there any limits of the information though that you do think should be limited you know is there any type of information that you think should be limited i think as you technology gets better with low earth orbiting satellites and the resolution gets better.

14:09I think there may be concerns about tracking people. You know, we talked about hyperspectral basically being a barcode reader from space. And so I can tell you what factory the paint on your car came from. That's how good it is. I can tell you where the meth labs are. I can tell you, you know, these are interesting things. We don't want to get to the point where people like tracking their significant other for like you know using because of the cadence of these satellites and who knows maybe 10 years from now the the cadence will be so high and the data download will be so quick that it will be a good tracking tool um then then we'll have to worry about those issues but right now i don't think oh hold on oh hold on we're just is it out the water?

14:56We just don't want to be in a sound bar. We are pretty late but we're actually on a bar so we gotta get... Oh fuck. Sound bar. I mean they're zoom in cars. Yeah but we we're we're we got to go this way and out to where they are. We cut across. It'd be extremely funny but also unfunny if we ran a ground. We're having fun. We're fun talking about humanity but now we're stuck. Are we actually stuck? No we're not completely stuck I just don't want to get the props hitting and then we'll be fine we're gonna move over we tried to make the shortcut and we could I think a major concern that people have is is the tracking ability and quite honestly for individual tracking right not at scale right we're talking about 25 square satellites are a bad tool okay there are much better tools drones ad tech your you know your phone that's constantly monitoring where you are etc.

15:53for tracking an individual person. So I think those concerns are overblown, but you know we have to address those and let people know as they become familiar with the technology, yeah this is terrible. It's good for monitoring infrastructure, construction schedules, you know how many cows are there, are the cows feeding. This is a use case I didn't even know. Cows are generally eating, sleeping, or in the barn. So there's like a rule of thirds, you know, And if the cows are standing or not is a signal of the health of the lot. And so people are using satellite imagery for livestock. You know, I didn't have that use case on my bingo card, right?

16:33So those are the things that it's good for, right? In, let's say, 20 years' time, what's your dream best case scenario of satellite imagery? So, well, the state of stand-on-light imagery is that we are now at roughly 10-centimeter, 7-centimeter resolution at scale. So that means you can go down to 10 centimeters. Like, you can differentiate between two objects at a distance of 10 centimeters from one another. Actually, what it means is each pixel represents is 7 centimeters. So that's a ground sampling difference. Right, which is effectively the same thing because you can say there's a difference there to there.

17:08Yeah, there's a point, right. And so at seven centimeters and at that point we're really good at tracking all infrastructure. We're really good at the analytics, the ability of computers to extract information and provide valuable insights gets very, very good at that scale. And it's cheap and it's abundant. So it's not just my brain, your brain, but everybody's brains working on finding all types of use cases. Right. What I'd like to do with that, what use case am I most attached to, it's making a significant dent in illegal fishing. Right. A lot of a lot of people are concerned about climate change and it's not like I'm not concerned about climate change.

17:50But I can tell you this. If the ocean dies, we all die. That I'm sure of. Why is that? Because it is our food source. It is our it is a significant source of oxygen for us. OK, so it's not only the oxygen and the ecosystem. phytoplankton yeah phytoplankton right it's not the we do not get our oxygen from the brazil rainforest we get it or a small amount but the majority of it comes from the phytoplankton yes by far a poor uh people don't realize that yes people don't realize that and they're already starting to be these huge dead zones there are dead there used to be dead zones only around like as river estuaries and such where it's like all the night trade runoff and so on from farming and and so forth but now they're getting them in they've found some like huge dead zones right in the middle of the ocean and no one even knows why the fuck that's happening yeah and it's very very concerning it also when you have a ecosystem collapse near nations you actually cause human catastrophes where particularly in africa where you collapse the fishing industry the popular starts to move in for bushmeat and then you have a rhino and walrus and lion issue and you have conflict there and you also have uh bad actors coming in to take advantage of the newfound poverty and distress and so it causes a lot of problems uh and then you're talking about ocean plastics the fishing nets etc uh despeciation you never know when it's the last straw to break the camel's back and so i think this is the issue of our time um and i think that uh there needs to be more focus on stopping the collapse of our ocean how useful is uh ai when like you know especially like this sort of diffusion models with analyzing it's very it's excellent it's excellent because you have you have this coverage imagine i could tell you not only when you were getting robbed but when the robber was driving to your house where he got gas he's going to be there in an hour and then where he went and who he meets up with that's what you have when you have global uh maritime awareness is what you have with satellite imagery so it's good for my monitoring assets like fishing vessels they can't hide yeah right they're slow moving vessels they go maybe 20 knots okay and so that way we now have an awareness issue we have to marry that with an enforcement forcement and then we will make a significant dent using existing laws and regulations we don't have to make any new laws we don't have to make any new regulations we just need to take the knowledge and awareness we have and get it to an enforcement nature and then we're going to have excellent benefits for the planet from an environmental perspective.

20:23Also, when these bad actors in these nation states are forced to buy their protein sources from other nations, we now increase trade. And so I have a belief that goods and services cross borders are bombs and bullets too. And I'd rather see goods and services cross borders, so I would like to see ag and agriculture and protein and grains to go to these nations in exchange for capital and have more trade as opposed to war. Yeah. So I'd always like to end each episode with some rapid fire predictions on the topic in hand. So we've just seen Starship Flight 6. It didn't go quite as well as Flight 5 clearly, but it's still going right?

21:04Yeah. So based upon that, probability that Starship lands on the moon by 2029. I'm gonna go I'm originally gonna say 95 % I'm gonna put 90 % because there's global conflict in the world and that could interfere with all other priorities but assuming it's a priority I think it's a certainty okay probability that Starship lands on Mars by 2030 I'm gonna put that lower at 60 % I and I'm just gonna say there's not as many there's not as many winners to go to Mars where you can go to the moon almost any other Tuesday. I think between now and then at 30, I think there's only three windows anyway. So I'm going to have to put that much lower.

21:49Yeah, much lower. Okay. Probability that there's a human on the moon by 2029. Wow. By any means. I get it. A human says that - I got 50-50. 50-50. All right.

22:08Probability that Skyfi is the world's number one resource for satellite imagery by 2028. I wonder if we already are. I'm going to put that very, very high. We have an excellent team. Obviously, you know, I'm the protector of the vision. I founded it, et cetera. but we have great experts in making it usable and accessible for everyone. Our partners are on that. We have also, we're working on white labels for other providers. So in some way, shape or form, we are the dominant provider of satellite imagery. All right, last one. Probability that Bill Perkins goes to space in his lifetime. I think I'm going to put that close to zero unless in my lifetime.

22:56I have to be alive. They can't put my ashes. No, ashes don't count. You know, as a kid, I really, really wanted to be an astronaut. And when I first moved to Houston, I drove by NASA and I cried. I cried because it was so emotional. It was so powerful. But as I've gotten older, I've gotten very scared of rockets and going into space. I like being tethered to Earth now, so I'm going to put that at 0.1%. Wow. All right. I feel like I'm emphasizing above. It's almost a risk to myself, you know, the risk of the mission going wrong is below 3%. I'm getting my ass on that. Actually, let me revise that because there might be a shuttle service to Japan in half an hour.

23:37And how can I not take the shuttle service to Japan in half an hour? Exactly. Like you're like a man. You're the man of efficiency. How can I not? I'm just going to have to take a Xanax and just check out. I mean, if it gets a... It goes in orbit, right? Yes. Right. Well, yeah. It depends. It's like low-eth orbit. The stars... The heat off orbit where you have to actually... Right. If you're doing a parabolic flight, it's a little bit different than having to go the full 17 ,000 miles an hour. See myself running to the spaceport to get to Japan on time. I'm going to miss my rocket. Yeah. Awesome.

24:06Thank you, Bill. Thank you. Thanks for having me again. Amazing. Thank you, guys.