You look good here. You think you look good? I'm happy. But they look at me. You got something in your teeth. All right. You got it? Yeah, that's good. All right. Are you excited? Yeah, I'm excited. Are you nervous? Uh, a little bit.
I mean, I have you to thank. She gave me material, but I also have to do research, man. Oh, okay. Okay. And I also have to try to learn what, you know, rockets do and like solar arrays and panels and...
Philip, you're officially a rocket scientist now.
Uh, yes, as much as non-technical, yeah. From a non-technical standpoint, sure. Yeah, we probably both know about the same.
No, I'm pretty sure you know quite a bit more. All right, cool. Good. Okay. You ready?
Y Combinator, McKinsey & Company, Harvard Business School, Stanford Business School. Philip, the list of organizations that have rejected you over the years is pretty impressive.
It is a very extensive list, I'll give you that. Yeah.
You went to Harvard Kennedy School, where you were classmates with some of the smartest students in the world that were rejected from Harvard Business School.
That's not such a bad thing.
What was it like being surrounded by people who are just as bitter as you were?
If you want bitter people, you should go to Wharton.
That's true, actually. At least half of the Kennedy School class didn't apply to Harvard Business School, whereas 100% of the Wharton class applied to Business School. Fine. But you're nothing if not persistent. You know, despite getting rejected from Harvard and Stanford's Business Schools three times each, you finally broke through when your girlfriend got into Stanford and you rode her coattails onto campus.
Is it exciting to be an SO at Stanford?
It's definitely an experience. I was having this conversation with my girlfriend yesterday. Like SOs have a certain persona. SOs are like background noise. And if you put your head above that parapet, then you get rejected by the SO colony.
That's so true. It's like every time, by the way, your girlfriend said you're like the most popular guy at GSB anyways. Oh, well, that's... Everybody knows you. Do you hang out with students to like poach them for employment? Or is it like looking for investors? Or...
We have an engineers-only recruitment policy.
Oh, is that right? So, zero MBAs probably for the next two years. You're the only exception.
Okay. So, that gets to be special. Well, you do deserve some perks as a CEO, no?
Shall we get into Starcloud?
Yeah, let's do it. Once this fire truck gets out of the way.
Okay. All right. Now we get into the meat and potatoes of things. Describe what Starcloud is trying to do in like one sentence.
Do you know that meat and potatoes is a euphemism in British slang?
You might have to guess. 'Cause you know what it means in America, right?
It means like the main...
Exactly, yeah. Kicks him right in the meat and potatoes.
All right, that's interesting. What was the question? Describe what Starcloud is trying to do in one sentence.
We're building very large data centers in space in order to be able to take advantage of the abundant energy in space and the ability to scale almost indefinitely.
Got it. So, basically Starcloud is trying to put data centers in space. Yes. But isn't there already enough garbage up there?
Uh, if that's a serious question, there is a fair amount of orbital debris between the altitudes of about 400 to 800 km where Starlink and other satellites fly. If that was you calling my satellites garbage, I would say our satellites will be among the most useful satellites there are.
So, speaking of garbage, I recently read Starcloud's white paper on training AI in space. How did you take such a fascinating topic and put people to sleep with it?
I would take that as an insult were it not for the fact that that was written by my co-founder Ezra, who has a PhD. Oh. And we've received much praise that it's actually incredibly easy to read.
Oh, is that right? If you wanted to, you could make that much more boring than it is.
Oh, really? That's shocking.
No, I'm just kidding. I actually really enjoyed reading Starcloud's white paper on training AI in space. I'm just wondering, will you guys be publishing any more works of fiction?
Yeah, as I mentioned earlier, we'd rather not be science fiction for too long. Sometimes people claim that this is a moonshot, and a moonshot we are hoping it will not remain.
So, no more works of fiction. I mean, you guys are only going what, 300 km up in space, so... Not quite the moon yet.
350 km, yeah. 300,000 would be the moon, but we might have some data centers orbiting the moon at some point as well.
Oh, okay. Is that your Series B target?
Um, probably like Series F.
We've got quite enough to chew on right now, but... Back when you were a teenager, I understand that your grandmother once walked in on you polishing your own rocket. Have you always been into space tech?
Getting caught using your Harvard ID to pick up Alabama girls is more embarrassing than getting caught masturbating by your grand... Yes, I would know, actually.
I haven't always been in space tech, but I will just say something about that anecdote. My grandmothers died before I was born, and I deliberately put that joke into my stand-up comedy so that if anybody ever thought that the entire thing was 100% true, I had some verifiable fact. So, it's not true. They were both dead before I was born.
Philip, how can you expect us to believe you're the man to grow data centers in space when you can't even grow a mustache right here on Earth?
That's a beautiful mustache. What are you on about? That is a first-class mustache.
That is a first-class piece of crap mustache. Putting data centers in space is quite a massive undertaking. But at least if you fail, you can always go back to working at Abercrombie.
I'm very proud of this picture, so you're all more than welcome to share this.
I mean, this is a picture of you sending this to your girlfriend, which she says she's never personally experienced.
So, she's missing out on the real Phil experience. Maybe one day when I'm... What I said the other day, well, I've lost in abs, I've made up for in potential billionaire probabilities.
So, what I'm hearing is you do think you would someday become a billionaire. And that's worth the six-pack abs.
My probability of becoming a billionaire has increased slightly.
Has tripled from 0.1 to 0.3%?
I'll take 0.3, yeah. I will take 0.3. I would quite like to start going to the gym again.
Yeah, I know what you mean. This was when you were 20?
Yeah, I was in peak physical form at 28. Right here on campus. 'Cause I used to... You know the Dish? Yeah, I used to run the Dish like five days a week. And lift weights. And I was like, you know, trying to eat healthy and all that good stuff. And then my first year in Dubai back in McKinsey, I was still pretty good, but then, you know, like every year you're like incrementally worse. And then all of a sudden you're like, how the hell did I get to this point? Yeah, yeah.
Yeah, I think I was peak immediately after grad school. I spent like... I had like a three-month period where I got in extremely good shape, and then...
Uh, immediately after, like when I was 31, I think, or 32, maybe.
Okay, okay. How old are you now?
Yeah, same. Oh, yeah, yeah. It's like... Someone once told me, by the way, like your late 30s are when you're in peak physical form for marathons. Like Iron Man and all that, yeah. Yeah, we'll find out.
So, let me get this straight. Starcloud is going to set up a solar array that's five times the size of Central Park up in space to generate five times the power of a nuclear reactor just so that ChatGPT can write us better breakup texts?
Pretty much. I mean, I think there'll be other more practical use cases or useful outputs from this than breakup texts, but yeah, for example, AI-generated AI is going to use an enormous amount of inference compute. So, and we will require massive new data centers, and powering them terrestrially is going to consume an enormous amount of resources. So, it makes sense to do it in space.
I love that 40-second visual you guys have. Which is like this huge, is like this massive star. And the music was so eerie, by the way.
Oh, yeah. Yeah, we had this production company make it, and they're like, we went through like five different versions of the music, and in the end we were just like, fine, that'll do.
But it was good, though. No, it was good. I like that. And I like the modularity of these data centers coming out of the, I guess it's a SpaceX Starship.
Yeah, yeah. So, actually, fun fact, we actually had that thing looking like a Starship, and then SpaceX were like, uh, we haven't got this approved by our comms people, and we think there's people going to be pissed off with this huge data center. So, can you make it not look like a Starship? So, even though we are launching with them, I thought that'd be pretty exciting for them, yeah. Yeah, yeah. No, I mean, maybe in future they'll change their mind. SpaceX.
No, but actually it was a good visualization of what you're doing, which helped me understand, actually. Oh, yeah. Nice. Yeah, and that makes sense. Like, okay, so when you're up in space, there's more space to set up these data centers. You can do it in a modular way so that you can keep adding capacity. The sun's right there. You have a huge solar array. Yeah. Uh, you know, for me the question is like, how do you beam it back down to Earth?
Well, yeah, beaming power from space to Earth is a huge challenge, and people were trying to do that using microwaves for a long time. The concept of space-based solar is not a new one. Since the 70s people have talked about it. But the problem is you have to use microwave, like RF, to beam that down, and that's like 95% efficiency loss getting power from space to Earth using that method. Yeah. So if you can use that power in space by crunching, by basically shipping all of your data center up to space, that makes a whole lot more sense, and then you're just transferring data down, and you can do that using, you know, any kind of radio or optical, like lasers, space lasers essentially.
Yeah, it's completely inevitable actually that we will be building large data centers in space. It's really just a matter of when the launch cost gets to the point that makes it economical, and with Starship we think that's going to be in a couple of years. It could be that it's in 10 years, but it is going to happen at some point. It's just really a matter of when the launch cost gets to...
Did you think through all these things while modeling for Abercrombie? Are you going to show the picture again?
I will show the picture again.
That's pretty amazing. I mean, you're obviously very passionate about this.
I'm very passionate about it. Got some drama here. Yeah, and what's been nice is seeing since that we put the white paper out in September, and to be honest we got completely pilloried on X for putting the white paper out by like armchair physicists and some real physicists who had lots of objections on the thermal management side of things and all the rest of it. In the last like six months it seems like the industry has really shifted around to the way that we're thinking about it. So yeah. People that were really against what we're doing are now saying it's like completely inevitable, and like I think what people don't understand is the cost of launch is going to come down by an enormous amount. The capacity to get stuff in space is going to go up by even more, and the demand for energy and the demand for compute is going to go through the roof.
So yeah, like I posted this thing this morning that Eric Schmidt was in Congress last week saying that electricity consumption in data centers will go from 3% of total to 99% of total, and he's saying by 2030 that we're going to need an extra 70 gigawatts worth of electricity for data centers. Right now all of data centers combined are like 50 gigawatts. So in five years we're going to be doubling that. More than doubling that. I think it's just insane the amount of energy these things are going to require.
So at some point what you're working on isn't just going to be a luxury but actually a need.
Complete necessity because even if nuclear fusion works, you can't dissipate all of that heat from a nuclear fusion running a data center terrestrially. Yeah. The way you get rid of that heat is by essentially turning water into steam and pumping it out into the atmosphere. Yeah. And just that process alone is going to warm up the atmosphere by way more than anything that fossil fuels are causing. So you need to dissipate the heat directly into the vacuum of space, and the only way you can do that is by having huge data centers in space and huge radiators. I mean, our entire engineering team is focused on building these radiators. That is really the core of technology we're developing.
What do the radiators do?
They dissipate that heat into vacuum.
Yeah, so 'cause we don't have an atmosphere, there's no convection or conduction. So you have to just... you dissipate it with infrared. So you have these huge panels that you keep warm and they emit infrared out into space.
Got it. We came here to talk comedy and instead you're giving me a science lesson.
I feel wonderful. I feel like Barack Obama. You know, he would only go on like talk shows if he was allowed to give his message.
Well, it was pretty funny though. For the listeners, I made him know what he is.
That's so true. I owe you your career to me. I do. I owe you your entire life and future prospects to me.
I put on a stand-up comedy show that inspired the rest of your life.
And not only that, by the way, for listeners, Phil almost didn't let me into the show because I was a last-second addition.
And then he was like, hey, let's get on a call, let's hear your set, and then we'll make a decision. And I read it out to you, and you were like, yeah, it's pretty good.
Completely away... You can't resist. You can't resist. Then I was like, everyone has to keep it under seven minutes. And then you went on for like 21 minutes. And at the end of it I was like so pissed off 'cause I was sitting there in my chair like grinding my teeth. I was like, when the hell is he going to shut the hell up.
So I get on stage, I had a whole post-Joel-set roast ready for you, but I couldn't be bothered at the time. So I got on and I was like, thank you so much. And I didn't even say thank you so much. I literally just got on and went, and then the next one, and then I went into my next guy bit.
That's so funny. It was hilarious. You were definitely the best, the funniest of the night.
Thanks, man. And you know what, I actually... So the funny thing is everyone went back to like, you know, being, living their lives, and then I was sort of like, ah, you know, I can't really go back to who I was before.
Your girlfriend tells me that you dress so poorly that you've had people give you money before because they thought you were homeless. Is that how you raised $21 million for your seed round?
I literally have had people give me money because... or often... Actually no, I had somebody ask me if I was homeless and I think was going to sell me fentanyl if I was. No, that's the state of my dress. So now I'm permanently decked out in beautiful Starcloud merch.
Over 200 VCs were chasing after you to get into Starcloud's seed round. I guess even crappy ideas sound good with a British accent.
Marc Andreessen has an amazing tweet. It says something like, overheard in Silicon Valley: if somebody's pitching you with a British accent, deduct 50 IQ points. If somebody's pitching you with an American South accent, add 50 IQ points.
Oh wow, that's... So it's actually a crush against you.
Uh, well, if people adjust, it would be a crush against me.
I mean, you're there pitching them solar-powered data centers up in space and all they're hearing is Downton Abbey. It's kind of weird. I mean, I did listen to your stand-up at Word and where you were complaining about your British accent wasn't getting you girls. Yeah. But it seems not to be the issue with VCs.
Yeah, depends on the VC again. Depends on the VC.
Your co-founders couldn't join us today because they actually had work to do. But they will be watching this video, so why don't you go ahead and introduce yourself and tell them what you do at the company.
I mean, number one, I very much doubt they will be watching this video. Number two, it's a Sunday afternoon. I'm allowed to go on a ride on a Sunday afternoon.
Startups never sleep, my friends.
Startups never sleep, true. And in theory this is building visibility of the Starcloud brand.
There we go. There we go. I love it. I love it. You don't think they'll watch this video?
