WEBVTT 1 00:00:04.280 --> 00:00:07.799 With laurn segle And from London and Gerard Read from Berlin. 2 00:00:08.320 --> 00:00:11.119 This is Redefining Energy Today. 3 00:00:11.439 --> 00:00:16.199 On Refining Energy job, we're going to talk about space. 4 00:00:16.600 --> 00:00:25.800 The final frontier, Space a final frontier. 5 00:00:31.920 --> 00:00:36.759 And there's been so much debate conversation around data centers 6 00:00:36.759 --> 00:00:40.960 in space, and especially all the excitement around the SpaceX 7 00:00:41.039 --> 00:00:43.759 IPO that we really wanted to know if it was 8 00:00:43.799 --> 00:00:45.439 science fiction or if it was for. 9 00:00:45.479 --> 00:00:48.840 Real absolutely charty, right, Yeah. 10 00:00:48.439 --> 00:00:51.520 Before we introduce our guests, I have a one minute 11 00:00:51.560 --> 00:00:57.359 clip from Elon Musk explaining why half of the centers 12 00:00:57.479 --> 00:00:59.640 will be in space in a few years. Let's listen 13 00:00:59.679 --> 00:01:01.679 to the. 14 00:01:00.719 --> 00:01:03.880 The availability of energy is the issue. If you look 15 00:01:03.880 --> 00:01:07.560 at electrical output outside of China, everywhere outside of China 16 00:01:07.560 --> 00:01:09.680 it's more or less flat. But if you're putting data 17 00:01:09.680 --> 00:01:12.280 centers anywhere except China, where you're gonna get you electricity, 18 00:01:12.400 --> 00:01:14.840 especially as you scale, the output of chips is growing 19 00:01:14.879 --> 00:01:18.159 pretty much exponentially, but the output of electricity is flat. 20 00:01:18.519 --> 00:01:20.719 So how you're going to tell them chips aren't It's 21 00:01:20.760 --> 00:01:23.200 harder to scale on the ground than it is to 22 00:01:23.200 --> 00:01:26.000 scale in space. But also you're going to get about 23 00:01:26.040 --> 00:01:29.599 five times the effectiveness of solid pals in space versus 24 00:01:29.599 --> 00:01:31.840 the ground. And you don't need batteries because you don't 25 00:01:31.840 --> 00:01:36.159 have a day, night cycle, or seasonality, clouds or an 26 00:01:36.159 --> 00:01:39.640 atmosphere in space. So it's actually much cheaper to do 27 00:01:39.719 --> 00:01:42.480 in space. And my prediction is that it will be 28 00:01:42.719 --> 00:01:45.519 by far the cheapest place to put AI will be space, 29 00:01:45.680 --> 00:01:48.359 and thirty six months less, you can mark my words, 30 00:01:48.920 --> 00:01:52.040 in thirty six months, the most economically compelling place to 31 00:01:52.040 --> 00:01:53.400 put AI will be space. 32 00:01:53.959 --> 00:01:57.359 Lauren, I'm not going to bet against Ela Musco on that. 33 00:01:57.359 --> 00:01:58.359 That's what I would say too. 34 00:01:58.959 --> 00:02:01.359 Now is timing my be out a little bit as usual. 35 00:02:01.480 --> 00:02:03.480 I definitely think we're going to get there. I definitely 36 00:02:03.519 --> 00:02:05.159 think we're going to get there now. 37 00:02:07.120 --> 00:02:09.479 Not the only thing in space. There's a lot of 38 00:02:09.520 --> 00:02:14.719 satellites already and more coming. So our guest had an idea, 39 00:02:15.280 --> 00:02:18.639 and this guy has already built two space companies, so 40 00:02:18.879 --> 00:02:21.479 you know what he's talking about is at the Innovation 41 00:02:22.039 --> 00:02:26.639 and Engineering Committee at NASAU. His idea is to help 42 00:02:26.800 --> 00:02:28.599 refuel the satellites. 43 00:02:29.360 --> 00:02:32.240 Yes, Laura, and we were really lucky to get one 44 00:02:32.280 --> 00:02:34.759 of those pioneers that come on the show, and this 45 00:02:34.840 --> 00:02:38.479 is Andrew Rush is a CEO of Star Catcher. As 46 00:02:38.479 --> 00:02:40.560 you said, what they're doing is they're building the world's 47 00:02:40.599 --> 00:02:43.560 first orbital electrical. 48 00:02:42.879 --> 00:02:48.159 Grid, and he's signing PPAs his satellite company, which is 49 00:02:48.199 --> 00:02:52.319 absolutely extraordinary. So let's bring you on the show and 50 00:02:52.400 --> 00:02:56.159 try to understand what energy as a service platform in 51 00:02:56.240 --> 00:02:57.520 space is all about. 52 00:02:57.879 --> 00:03:15.319 Absolutely, Andrew, welcome to the show. Thank you for having me, Hundrew. 53 00:03:15.599 --> 00:03:18.120 Great to have you, and I'd like to really kick 54 00:03:18.159 --> 00:03:22.000 off with a simple question, which is if I look 55 00:03:22.039 --> 00:03:24.439 at what you're doing, you're trying to build the world's 56 00:03:24.520 --> 00:03:28.199 first orbital energy grid. Just please explain to us what 57 00:03:28.240 --> 00:03:28.639 that is. 58 00:03:29.240 --> 00:03:32.120 Yeah, we live in an amazing time. We live in 59 00:03:32.159 --> 00:03:35.159 basically the second Golden Age, and we've seen the number 60 00:03:35.159 --> 00:03:38.479 of satellites on orbit go from one thousand to now 61 00:03:38.520 --> 00:03:41.599 ten thousand and soon to be one hundred thousand. Every 62 00:03:41.639 --> 00:03:45.199 satellite in space has its own power generator. They're all 63 00:03:45.240 --> 00:03:47.560 going on basically camping trips to space. They have their 64 00:03:47.560 --> 00:03:49.759 own solar rays. When they see the sun, they charge up. 65 00:03:49.759 --> 00:03:52.120 When they don't, they live off the batteries. This is 66 00:03:52.159 --> 00:03:56.080 really constraining, and in fact, the average satellite in space 67 00:03:56.400 --> 00:04:00.439 has something like fifteen hundred blatts of power availability on 68 00:04:00.479 --> 00:04:03.840 board about as much as my kid's gaming computer uses. 69 00:04:04.479 --> 00:04:07.759 But we want our satellites to act like cell phone 70 00:04:07.800 --> 00:04:11.080 towers in space. We want to do orbital compute in space. 71 00:04:11.120 --> 00:04:12.360 We want to do manufacturing. 72 00:04:12.360 --> 00:04:12.759 One day. 73 00:04:12.840 --> 00:04:15.800 We take amazing high resolution images and process those and 74 00:04:15.840 --> 00:04:18.560 send them down to Earth. All of these things are 75 00:04:18.600 --> 00:04:22.319 really power intense, so there's a disconnect between the amount 76 00:04:22.319 --> 00:04:24.279 of power available and what we want to do in space. 77 00:04:24.759 --> 00:04:27.920 At Starcatcher, the company that I co founded, we are 78 00:04:27.959 --> 00:04:31.560 solving this problem by building a power grid in space 79 00:04:31.680 --> 00:04:34.920 that collects energy from the Sun and concentrates it and 80 00:04:34.959 --> 00:04:39.040 then distributes that energy to our client satellites by sending 81 00:04:39.120 --> 00:04:42.920 them basically the kinds of light that solarrays are already 82 00:04:42.959 --> 00:04:46.000 really good at turning into power, which basically visible and 83 00:04:46.120 --> 00:04:49.439 near infrared light. So we're building a grid in space 84 00:04:49.839 --> 00:04:52.519 that will beam energy from our satellite to our client 85 00:04:52.519 --> 00:04:55.279 satellites so that they don't have just fifteen hundred blots 86 00:04:55.279 --> 00:04:58.279 of power availability, they have like fifteen thousand months or 87 00:04:58.319 --> 00:05:02.240 one hundred thousand months, begin to have enough power to 88 00:05:02.279 --> 00:05:05.720 answer the call of what we are building in the 89 00:05:05.720 --> 00:05:06.600 future in space. 90 00:05:06.879 --> 00:05:09.240 And Andrew just let me cut across. Say when you 91 00:05:09.279 --> 00:05:11.160 say when you When I heard the words be me, 92 00:05:11.480 --> 00:05:16.000 I thought of beam me up, Scotty, right and sorry. 93 00:05:16.040 --> 00:05:19.439 Explain to me how you're transmitting the power. 94 00:05:19.920 --> 00:05:21.800 Talk a little bit about that, Jerald. 95 00:05:21.800 --> 00:05:24.279 I appreciate that because I grew up on this steady 96 00:05:24.319 --> 00:05:28.160 diet of science fiction and sites fact and specifically Star Trek, 97 00:05:28.199 --> 00:05:31.600 so appreciate the reference. So we have on board our 98 00:05:31.600 --> 00:05:35.800 satellites we have multiple wavelengths of lasers that we use 99 00:05:35.879 --> 00:05:39.000 to generate essentially of flashlight, and you think of it 100 00:05:39.000 --> 00:05:41.720 as a flashlight in space, like a big column of 101 00:05:41.800 --> 00:05:44.319 light that we can steer in space. It's about a 102 00:05:44.399 --> 00:05:46.680 meter and a half across, so it starts with lasers, 103 00:05:46.680 --> 00:05:48.959 but it really is more like a flashlight. And that's 104 00:05:49.079 --> 00:05:51.759 enough to bathe the average solar ray that's on a 105 00:05:51.759 --> 00:05:56.000 spacecraft on orbit in energy so that that can supercharge. 106 00:05:56.480 --> 00:05:59.759 So the Saturday light will keep their batteries, but they 107 00:05:59.759 --> 00:06:02.839 would receive your beam. So how far away are you 108 00:06:02.920 --> 00:06:04.279 going to be and how long is it going to 109 00:06:04.319 --> 00:06:06.560 take to chalk the battery. 110 00:06:06.639 --> 00:06:10.439 We put a lot of emphasis on being backward compatible. 111 00:06:10.720 --> 00:06:12.839 People don't have to buy a custom receiver to work 112 00:06:12.879 --> 00:06:15.480 with us, They don't have to change the orientation of 113 00:06:15.480 --> 00:06:17.680 their spacecraft on moor a bit, and we just beam 114 00:06:17.759 --> 00:06:21.120 energy to their arrays whether they're in sunlight themselves and 115 00:06:21.120 --> 00:06:23.519 they can actually get more energy by bias putting more 116 00:06:23.600 --> 00:06:25.839 light on them, or when they're an eclipse, and we 117 00:06:25.879 --> 00:06:28.839 can keep them powered. How long it takes to charge 118 00:06:29.000 --> 00:06:33.079 is really mission dependent. If somebody says, hey, my satellite 119 00:06:33.319 --> 00:06:37.000 has a thousand lots of power native on board, really 120 00:06:37.040 --> 00:06:39.199 like another two thousand watts, well, we're just going to 121 00:06:39.279 --> 00:06:43.680 put a couple of suns of energy equivalent on them consistently. 122 00:06:44.199 --> 00:06:47.519 Now contrast that with somebody that says, hey, I'm going 123 00:06:47.560 --> 00:06:49.360 to go do this maneuver. I'm going to use my 124 00:06:49.480 --> 00:06:52.600 electric propulsion thruster to boost my orbit up a little bit. 125 00:06:52.959 --> 00:06:55.079 So I just need additional power for like this two 126 00:06:55.079 --> 00:06:57.439 hour period, and so then we would just give them 127 00:06:57.480 --> 00:06:59.959 a dish that power for about that two hour period. 128 00:07:00.399 --> 00:07:02.399 And then in terms of how far we can throw 129 00:07:02.439 --> 00:07:05.560 the light. Now, we're a slave to physics like everyone is. 130 00:07:05.959 --> 00:07:08.639 So the configuration that we have, the design we have, 131 00:07:09.120 --> 00:07:12.040 we can beam energy up to about two thousand kilometers 132 00:07:12.199 --> 00:07:15.879 away from our core power node satellites and so we 133 00:07:16.040 --> 00:07:18.800 cover everywhere in LEO by putting up a constellation of 134 00:07:18.800 --> 00:07:20.160 about two hundred satellites. 135 00:07:20.720 --> 00:07:23.319 Now I've done a bit of research, and of course 136 00:07:23.360 --> 00:07:28.399 what's been amazing is the decrease in the cost of 137 00:07:28.480 --> 00:07:32.920 sending some mass into space. So if I look twenty 138 00:07:32.959 --> 00:07:35.680 five years ago, you had the legacy, the low key, 139 00:07:35.839 --> 00:07:40.160 the high end, they would charge you ten thousand dollars pokidogram. 140 00:07:41.120 --> 00:07:44.319 Now Falcon nine is going to charge you one thousand, 141 00:07:44.399 --> 00:07:47.839 five hundred, so that's what eight times cheaper. And of 142 00:07:47.839 --> 00:07:52.399 course Starship announced two hundred dollar pakido Are you banking 143 00:07:52.439 --> 00:07:53.560 on those advancements? 144 00:07:54.480 --> 00:07:57.879 The reduction in launch cost over the last like ten 145 00:07:57.959 --> 00:08:01.639 or fifteen years has just been an enormous boon to 146 00:08:02.240 --> 00:08:06.560 the commercial space industry. Really, it birthed this modern commercial 147 00:08:06.560 --> 00:08:10.959 space industry, like having launch cost way down and expanding availability, 148 00:08:11.879 --> 00:08:16.480 and that's what's enabled next generation remote sensing Earth observation 149 00:08:16.720 --> 00:08:20.120 companies like Planet and that's what's enabled the direct to 150 00:08:20.199 --> 00:08:24.240 sell connectivity constellations we're seeing being put up from AST Mobile, 151 00:08:24.680 --> 00:08:29.639 Amazon and SpaceX themselves, and that's what is enabling the 152 00:08:29.680 --> 00:08:34.039 future of space. But you really need additional infrastructure. Can 153 00:08:34.120 --> 00:08:37.720 think of launch as basically transportation infrastructure the roads of space, 154 00:08:38.120 --> 00:08:41.879 and we really need additional infrastructure to unlock the true 155 00:08:42.399 --> 00:08:45.440 next generation applications of space, and that's where building power 156 00:08:45.440 --> 00:08:49.799 grid like we're doing Starcatcher comes into play. We're excited 157 00:08:50.000 --> 00:08:53.360 for launch costs to continue to come down, but we're 158 00:08:53.399 --> 00:08:57.840 not reliant upon our business, our technology. Our constellation doesn't 159 00:08:57.879 --> 00:09:01.639 require starship for exams sample to be up and flying 160 00:09:01.679 --> 00:09:05.120 and available at two hundred dollars a kilogram, because we've 161 00:09:05.120 --> 00:09:08.200 actually seen a little bit of a retraction in the 162 00:09:08.279 --> 00:09:11.679 last two or three years in that trajectory of launch cost. 163 00:09:11.960 --> 00:09:14.639 We got down to about that thirteen hundred dollars kilogram number, 164 00:09:14.679 --> 00:09:16.840 and now it's actually gone, it's ticked up a little 165 00:09:16.879 --> 00:09:19.080 bit to maybe more like two thousand Andrew. 166 00:09:19.120 --> 00:09:22.320 Could you just explain to me, if I am a 167 00:09:22.360 --> 00:09:27.440 satellite company today, what's the economic benefit for me actually 168 00:09:27.480 --> 00:09:30.440 going and using your services right now? 169 00:09:30.759 --> 00:09:34.039 If you're a satellite owner operator, you buy everything up front, 170 00:09:34.200 --> 00:09:37.240 the capex, to design your system, to build your satellites, 171 00:09:37.279 --> 00:09:39.720 and the field those is the biggest hurdle for you 172 00:09:39.799 --> 00:09:42.639 to clear and in fact, a lot of space companies 173 00:09:42.759 --> 00:09:46.679 die in that capital formation capex part of the story. 174 00:09:47.159 --> 00:09:49.759 So one of the huge benefits that we provide just 175 00:09:49.799 --> 00:09:52.919 from a business perspective is that the power system is 176 00:09:52.919 --> 00:09:56.120 fifteen to twenty percent of the cost of a satellite. 177 00:09:56.200 --> 00:09:58.480 But what we do is we're selling energy to you 178 00:09:58.559 --> 00:10:00.480 not as a big upfront call it is a big 179 00:10:00.480 --> 00:10:03.799 capex cost, but actually just as an opex cost. So 180 00:10:03.879 --> 00:10:07.240 we sell energy like on a WAT per year delivered basis, 181 00:10:07.320 --> 00:10:10.440 as an operating expense, and so that lowers that initial 182 00:10:10.519 --> 00:10:14.159 cost barrier and enables folks to get up faster and 183 00:10:14.200 --> 00:10:17.600 more cheaply, and also stay in a smaller form factor 184 00:10:18.120 --> 00:10:20.840 because although launch costs have come down, it's still not 185 00:10:20.960 --> 00:10:24.480 free to put things in space. We enable satellites that 186 00:10:24.559 --> 00:10:27.879 are a few hundred kilograms that costs maybe single digit 187 00:10:27.960 --> 00:10:32.720 millions to have the same power availability as satellites that 188 00:10:33.000 --> 00:10:36.120 are thousands of kilogram and cost tens of millions of dollars. 189 00:10:36.279 --> 00:10:39.320 There's just enormous savings and changes in the way that 190 00:10:39.360 --> 00:10:42.240 we design systems. Now, of course, there are other benefits 191 00:10:42.320 --> 00:10:45.320 to life extension of assets on orbit. That's a really 192 00:10:45.360 --> 00:10:48.039 big one of being able to letting people increase the 193 00:10:48.120 --> 00:10:51.159 uptime that they have available on their spacecraft. All of 194 00:10:51.200 --> 00:10:54.360 these things are additional ways that we provide a lot 195 00:10:54.360 --> 00:10:56.720 of value just satellite designers today. 196 00:10:57.080 --> 00:10:59.639 So you can to announce two hundreds of them, one 197 00:10:59.679 --> 00:11:00.960 of them, how much does it cost. 198 00:11:01.799 --> 00:11:04.320 The general approach that we take at Starcatcher is a 199 00:11:04.399 --> 00:11:08.320 crawl walk run approach to technology maturation and then fielding 200 00:11:08.399 --> 00:11:12.759 system we're building. We've actually built our first satellite. It's 201 00:11:12.759 --> 00:11:16.200 a subscale system that will fly in the next year. 202 00:11:16.559 --> 00:11:18.240 We are closer to that launch date. We're going to 203 00:11:18.279 --> 00:11:20.639 talk more about that, and then we'll field in a 204 00:11:20.679 --> 00:11:24.360 subscale operational prototype pilot power plant a few years after that, 205 00:11:24.840 --> 00:11:27.720 and then get to full up operational capacity. By the 206 00:11:27.840 --> 00:11:31.159 end of the decade. Each satellite will have about one 207 00:11:31.240 --> 00:11:34.679 hundred kilo loots of energy delivery capacity, and the total 208 00:11:34.759 --> 00:11:38.240 constellational be able to service about twenty percent of the 209 00:11:38.320 --> 00:11:42.000 current power generation demand on orbit today. Now, of course, 210 00:11:42.039 --> 00:11:44.960 that demands a moving target ten x number satellite space. 211 00:11:45.000 --> 00:11:46.480 That demand is going to go up by ten x 212 00:11:46.799 --> 00:11:49.639 and so the constellation size actually will grow as the 213 00:11:49.759 --> 00:11:51.879 number of customers grow in space. 214 00:11:52.360 --> 00:11:57.120 I'm making comparison which don't make any sense, but for me, 215 00:11:57.679 --> 00:11:59.879 that makes sense. You know, it's a bit EV charging 216 00:12:00.600 --> 00:12:04.039 and EV's but a good thing is now the EV's 217 00:12:04.080 --> 00:12:07.399 already in space, so you arrived with your EV charging 218 00:12:07.799 --> 00:12:12.960 satellite charging. So the economics, let's say one of your boxes, 219 00:12:13.039 --> 00:12:15.879 because I don't know ten million for the sake of 220 00:12:16.039 --> 00:12:20.080 the argument, maybe it's one. Maybe I don't know, how 221 00:12:20.120 --> 00:12:22.919 confident are you that you're going to say, say a 222 00:12:23.000 --> 00:12:26.039 million a year? How does the economics of one of 223 00:12:26.039 --> 00:12:27.039 your box works? 224 00:12:27.480 --> 00:12:30.039 I love the comparison to the EV chargic network. 225 00:12:30.679 --> 00:12:31.279 One of the. 226 00:12:31.200 --> 00:12:33.679 Things that I'm proud of that in the way that 227 00:12:33.720 --> 00:12:36.480 we've designed this is that the way we send the 228 00:12:36.600 --> 00:12:40.159 energy is compatible with sort of the existing gas cap 229 00:12:40.440 --> 00:12:43.480 that'sllites have, like with their existing solo right, we don't 230 00:12:43.519 --> 00:12:47.039 have this Tesla plug design versus the next plug design 231 00:12:47.200 --> 00:12:50.000 kind of war going on there. But to your question 232 00:12:50.039 --> 00:12:52.519 of you know, how confident are we in the economics 233 00:12:52.519 --> 00:12:55.639 of it? One of the first things I did when 234 00:12:55.679 --> 00:12:59.320 we started this company was just call a whole bunch 235 00:12:59.320 --> 00:13:01.879 of my friend the space industry. This is actually the 236 00:13:01.919 --> 00:13:05.320 third space company that I have fled or scaled. Then 237 00:13:05.360 --> 00:13:07.639 the first two had quite good outcome. I was the 238 00:13:07.679 --> 00:13:09.960 CEO of a company called Made in Space, which was 239 00:13:10.000 --> 00:13:11.879 the first company to three print off the face of 240 00:13:11.919 --> 00:13:13.919 the planet, and built that up and sold that quite 241 00:13:13.960 --> 00:13:17.320 successfully in twenty twenty. And then I was the chief 242 00:13:17.320 --> 00:13:20.519 operating officer of red Wire through taking that company public 243 00:13:20.559 --> 00:13:23.159 and now it's a couple billion dollar firm. And so 244 00:13:23.600 --> 00:13:27.600 I called CEOs of space station companies, coos, remote sensing 245 00:13:27.639 --> 00:13:31.600 companies and tele communications companies and described what we wanted 246 00:13:31.600 --> 00:13:34.879 to build, described the ambition, and every single one of 247 00:13:34.879 --> 00:13:38.120 those prepol said, hey, this is really cool. I would 248 00:13:38.200 --> 00:13:40.159 use it this way like I would use it like 249 00:13:40.200 --> 00:13:42.600 in the way that you suggested, and hear other use cases, 250 00:13:42.720 --> 00:13:44.360 and so that gave us a lot of confidence. And 251 00:13:44.399 --> 00:13:48.200 then from there we've gone and actually gotten forty letters 252 00:13:48.200 --> 00:13:51.480 of intent signed by basically every flavor of space company. 253 00:13:51.840 --> 00:13:54.919 It's several billion dollars potential contract value associated with that. 254 00:13:55.159 --> 00:13:57.879 And then we even converted seven of those two power 255 00:13:57.919 --> 00:14:01.120 purchase agreements where folks that pay us a deposit to 256 00:14:01.279 --> 00:14:03.720 lock in the power rates contemplated in their letters of 257 00:14:03.759 --> 00:14:07.120 intent for a period of years. So that plus the 258 00:14:07.159 --> 00:14:08.759 work that we've been able to do with the US 259 00:14:08.879 --> 00:14:12.679 government gives us really high confidence that the Economics Pencil 260 00:14:12.720 --> 00:14:15.480 out not only on a per spacecraft basis, but for 261 00:14:15.559 --> 00:14:18.320 the overall TAM and the overall constellation. 262 00:14:18.879 --> 00:14:21.120 Andrew can I also just to ask a little bit 263 00:14:21.120 --> 00:14:23.960 about data centers in space, and I know they're not 264 00:14:24.039 --> 00:14:27.080 necessarily doing that, but I can imagine if they do 265 00:14:27.240 --> 00:14:29.960 take off, this is just a huge market, a huge 266 00:14:29.960 --> 00:14:32.399 opportunity for you, right, So what's your views on that 267 00:14:32.600 --> 00:14:34.879 and how do you fit into it? And I just 268 00:14:34.879 --> 00:14:37.799 actually just look to hear generally how you see the 269 00:14:38.000 --> 00:14:39.519 success possibilities. 270 00:14:40.080 --> 00:14:44.759 I'll lead off by saying, essentially every orbital data center 271 00:14:45.360 --> 00:14:48.639 startup has signed either a letter of intent or a 272 00:14:48.639 --> 00:14:52.039 power prise agreement with us because it's this interesting like 273 00:14:52.120 --> 00:14:57.799 symbiotic relationship there that they know power is the key thing, 274 00:14:58.200 --> 00:15:02.519 and for us it represents huge upside potential. We recently 275 00:15:02.600 --> 00:15:06.240 closed a series a round of sixty five million dollars, 276 00:15:06.279 --> 00:15:08.720 which is one of the largest series as for space 277 00:15:08.759 --> 00:15:14.000 technology company ever and that was before SpaceX went public 278 00:15:14.120 --> 00:15:17.960 on a thesis of building data centers in space, and 279 00:15:18.039 --> 00:15:21.559 so there's even more kind of upside that we see there. 280 00:15:22.200 --> 00:15:25.679 In general, the thing that's interesting about data centers in 281 00:15:25.720 --> 00:15:27.840 space is if you kind of break it down into 282 00:15:28.279 --> 00:15:31.159 use cases. The most near term use case for compute 283 00:15:31.159 --> 00:15:34.799 in space is just GPUs to do high speed processing 284 00:15:34.919 --> 00:15:37.919 and inference and AI analysis of things and then providing 285 00:15:37.960 --> 00:15:41.879 that data and providing insights to folks, not just raw data. 286 00:15:42.039 --> 00:15:46.279 That's something that's happening now on satellites. What I believe 287 00:15:46.600 --> 00:15:50.480 the near slash medium term use case for data centers 288 00:15:50.480 --> 00:15:53.480 in space like that actually underpins the thesis of them 289 00:15:53.840 --> 00:15:55.120 is the fact that. 290 00:15:55.039 --> 00:15:57.279 We are moving a lot of. 291 00:15:57.200 --> 00:16:01.080 Telecommunications into the space layer. Your cell phone, if it's 292 00:16:01.159 --> 00:16:04.639 less than like two generations old, probably has the capacity 293 00:16:04.799 --> 00:16:07.559 to send text messages or other data through a satellite, 294 00:16:07.759 --> 00:16:11.399 like directly to a satellite, and that brings satellites kind 295 00:16:11.399 --> 00:16:14.519 of into the telecommunications tapestry of the planet, and so 296 00:16:15.039 --> 00:16:20.480 services that benefit from being geographically close to that telecommunications tapestry. 297 00:16:20.679 --> 00:16:25.919 So think like inference nodes, think content distribution networks. Those 298 00:16:26.000 --> 00:16:29.320 things make sense to put in space where they're geographically 299 00:16:29.600 --> 00:16:35.559 close to Starling satellites, Amazon satellites, ast mobile satellites. Now 300 00:16:35.679 --> 00:16:38.200 the longer term, it will data centers be as cheap 301 00:16:38.240 --> 00:16:40.000 to put in space as they are to build on 302 00:16:40.039 --> 00:16:43.679 the ground. That is a maybe it's not a today 303 00:16:43.679 --> 00:16:46.879 thing it's a good marketing ploy for the SpaceX S one. 304 00:16:47.279 --> 00:16:49.480 I don't think it's going to be realistic that it's 305 00:16:49.519 --> 00:16:51.519 going to be as cheap to put a data center 306 00:16:51.559 --> 00:16:53.120 in space in two or three years as it is 307 00:16:53.159 --> 00:16:53.919 de bulod plant. 308 00:16:54.440 --> 00:16:57.480 So I've done a bit of a back and on 309 00:16:57.639 --> 00:17:00.600 the loop calculation. You get it if I'm or wrong. 310 00:17:01.039 --> 00:17:04.319 If you put a one gigat orbital data center, it 311 00:17:04.480 --> 00:17:08.200 requires one hundred thousand tons of hardware, and even if 312 00:17:08.200 --> 00:17:12.279 we use the starship level pricing, which is two hundred pakido, 313 00:17:13.480 --> 00:17:15.519 the cost of that data center is going to be 314 00:17:15.799 --> 00:17:19.799 forty trillion dollar, which is the GDP of the US, 315 00:17:20.440 --> 00:17:23.799 knowing that on Earth one gigaba data center costs two 316 00:17:23.880 --> 00:17:26.119 hundred times less. Am I wrong? 317 00:17:27.240 --> 00:17:30.960 The reporting that I have seen for a one gigawatt 318 00:17:31.200 --> 00:17:35.119 data center is if you make some fairly aggressive assumptions 319 00:17:35.279 --> 00:17:38.920 about what the spacecraft would cost and what SpaceX is 320 00:17:39.079 --> 00:17:42.799 cost for launches, not the price but their cost, you 321 00:17:42.920 --> 00:17:46.920 get more like one hundred billion. And maybe you get 322 00:17:47.079 --> 00:17:50.400 into like the sixty or eighty billion, but it's more 323 00:17:50.440 --> 00:17:53.000 like one hundred. And I've also seen folks say that 324 00:17:53.680 --> 00:17:56.680 a one gigawatt all up no data center is the 325 00:17:56.759 --> 00:17:59.319 tens of billion of cost depending on where you're building it, 326 00:17:59.440 --> 00:18:00.519 regulatory kind of things. 327 00:18:00.559 --> 00:18:02.559 I think I've seen. That's what I've seen. 328 00:18:03.200 --> 00:18:06.559 So what you're saying is that not my back on 329 00:18:06.759 --> 00:18:10.359 the envelope calculation, but some real experts think that the 330 00:18:10.599 --> 00:18:14.599 price in space is in the current zone of what 331 00:18:14.640 --> 00:18:16.359 we're seeing on the land. 332 00:18:17.119 --> 00:18:20.519 To modify that slightly, what I've seen from folks that 333 00:18:20.920 --> 00:18:23.559 are close to the metal here is that they see 334 00:18:23.559 --> 00:18:27.000 a pathway where there might be price parity if you 335 00:18:27.079 --> 00:18:32.440 make assumptions about launch cadence for Starship and real optimization 336 00:18:32.519 --> 00:18:36.160 on launch cost for SpaceX to put their own hardware up, 337 00:18:36.240 --> 00:18:38.599 and that they can get the satellites to a very 338 00:18:38.640 --> 00:18:42.759 aggressive cost point. Obviously, Starship is not flying operationally yet 339 00:18:42.880 --> 00:18:46.279 and therefore it's not cost optimized for launch, but it 340 00:18:46.319 --> 00:18:50.000 paints a picture to I think in the five or 341 00:18:50.039 --> 00:18:54.079 ten year timescale that you could achieve a reasonable cost 342 00:18:54.079 --> 00:18:56.039 parody if you're totally vertically ingrated. 343 00:18:56.559 --> 00:18:59.960 Andrew, that's fascinating for me to actually hear that, because 344 00:19:00.200 --> 00:19:03.200 maybe I just come back a little bit too, what 345 00:19:03.240 --> 00:19:05.720 are the problems that we see in space? Because I 346 00:19:05.759 --> 00:19:07.759 get the whole idea of a data center in space, 347 00:19:08.000 --> 00:19:09.960 But what are the issues? What are we missing? 348 00:19:10.640 --> 00:19:13.119 There's a lot of engineering that needs to happen. And look, 349 00:19:13.359 --> 00:19:16.880 I mean, I just I'll just like identify myself as 350 00:19:16.960 --> 00:19:20.880 a space optimist. I believe deeply in the potential that 351 00:19:20.920 --> 00:19:24.119 space has for humanity and for enriching people's lives and 352 00:19:24.200 --> 00:19:28.480 the planet. It's exciting to be talking about new uses 353 00:19:28.519 --> 00:19:31.720 of space. That is what data processing in space is. 354 00:19:31.279 --> 00:19:34.319 It is a new use in space. All that being said, 355 00:19:34.319 --> 00:19:36.119 if you think about a data center on the Earth, 356 00:19:36.519 --> 00:19:41.440 you are constantly maintaining it. You're constantly rebuilding it, putting 357 00:19:41.480 --> 00:19:44.839 the latest and greatest GPUs in GPUs do have a 358 00:19:45.079 --> 00:19:49.119 useful operational life, a useful economic life of several years. 359 00:19:49.680 --> 00:19:53.119 But ten year old GPUs is not commercially viable, so 360 00:19:53.200 --> 00:19:56.839 you have to maintain and upgrade those systems. Obviously, thermal 361 00:19:57.119 --> 00:19:59.680 like heat is a big issue. Now, that's an issue 362 00:19:59.680 --> 00:20:01.920 that we have that on the one hundred kill a 363 00:20:01.920 --> 00:20:05.359 watt scale we have solved in space already, Like the 364 00:20:05.400 --> 00:20:08.440 International Space Station has a thermal control system that's one 365 00:20:08.480 --> 00:20:11.119 hundred hundred feet kill lots. So the size of spacecraft 366 00:20:11.200 --> 00:20:14.039 that folks like SpaceX and others are talking about thermal 367 00:20:14.119 --> 00:20:16.920 has been solved at that class of heat rejection. Now 368 00:20:16.960 --> 00:20:19.000 gig a lot. If you're going to have a thousand 369 00:20:19.079 --> 00:20:21.240 satellites to do that, then you just got to build 370 00:20:21.240 --> 00:20:23.119 a thousand of the systems. If you want to create 371 00:20:23.160 --> 00:20:25.359 one big monel ethic one, there's a whole lot of 372 00:20:25.400 --> 00:20:28.079 other building structures in space, things that you've not done 373 00:20:28.119 --> 00:20:29.519 yet that would need to be solved. 374 00:20:29.839 --> 00:20:32.240 So then my prediction at the beginning of the year, 375 00:20:32.319 --> 00:20:34.400 which was we would have the first data center in 376 00:20:34.440 --> 00:20:37.359 space by the end of the year is probably are 377 00:20:37.359 --> 00:20:37.680 you were going. 378 00:20:37.640 --> 00:20:39.279 To put odds in it? What would you give me? 379 00:20:40.319 --> 00:20:40.559 No? 380 00:20:40.559 --> 00:20:42.799 No, Gerald, you won your bet. You won your bet. 381 00:20:42.799 --> 00:20:45.519 That Depending on how you define a data center, it's 382 00:20:45.559 --> 00:20:48.400 already happened. Like Phil Johnson at star Cloud has already 383 00:20:48.400 --> 00:20:51.079 put GPUs in space, and I think we did that 384 00:20:51.079 --> 00:20:53.160 in October of last year. If on the other hand, 385 00:20:53.160 --> 00:20:54.440 you're saying, hey, I want to see a gig a 386 00:20:54.480 --> 00:20:58.160 lot of power in space or gigle lass like, yeah, 387 00:20:58.200 --> 00:20:59.240 you're you're not going to win that. 388 00:20:59.200 --> 00:21:01.359 Bet, all right, Andrew you I've won the bet. So 389 00:21:01.400 --> 00:21:03.160 this is making difference, Thank you very much. 390 00:21:03.240 --> 00:21:04.359 If you just answered I won't. 391 00:21:04.400 --> 00:21:05.799 That's all right, that's all accounts. 392 00:21:06.000 --> 00:21:08.599 Yeah, so let me explain you how it works with Jah. 393 00:21:08.680 --> 00:21:10.759 So we throw a dots on a wall and then 394 00:21:10.880 --> 00:21:13.599 he paints the round around. 395 00:21:13.240 --> 00:21:17.880 Date, so it's bullseye. It's always want time under you've 396 00:21:17.920 --> 00:21:20.880 made my day. That'd be a certaince. 397 00:21:21.880 --> 00:21:24.279 And just maybe to finish off there, maybe talk a 398 00:21:24.319 --> 00:21:26.880 little bit about where you're at now as a company 399 00:21:27.000 --> 00:21:30.240 in terms of real commercial readiness, and then you know 400 00:21:30.359 --> 00:21:33.000 what the next steps are, and then really how do 401 00:21:33.039 --> 00:21:34.960