Forecast 2050 - Episode IV - Transcript - Philip Johnston on Space

Recorded December 13th, 2025 in Redmond, WA

Starcloud Founder & CEO Philip Johnston believes data centers in space are the first step to becoming a Kardashev level civilization. And today, Starcloud announces a $170,000,000 Series A on a $1.1 billion valuation, making it one of the fastest startups out of YC to reach unicorn status.

I caught up with Philip at Starcloud’s HQ in December to talk about his vision for the next 25 years. The conversation covers everything from luxury space tourism, whether Tesla’s Optimus robot will become a self-replicating galactic probe, the hard problem of consciousness, whether the world is a simulation, Mars becoming the next America, and Philip’s greatest concern for our long-term future.

Watch the full conversation:

Transcript:

Starcloud and the Case for Space GPUs

Finn Murphy: Very happy to have a friend and fortunately a founder I got to invest in from Nebular: Philip Johnston, the CEO and Founder of Starcloud. And before we get into talking about the future, it’d be awesome to know a little bit more about what you’re doing, and what we’re doing here. Where is Starcloud today?

Philip Johnston: Starcloud is building data centers in space initially to provide edge and cloud services for other spacecraft and then to compete with almost all data centers on energy costs. So in a ten year time frame, we’re expecting most new data centers to be built in space purely because of the constraints that we’re facing on energy.

And that’s what we’re doing here. We’ve just launched our first spacecraft, Starcloud-1 has the first Nvidia H100 onboard. We’ve trained the first model in space. We’ve run the first version of Gemini in space, and at this facility, we’re building the second spacecraft for Starcloud-2.

Murphy: When we think about this problem of energy, there’s conservative forecasts that say, in the next five years we’re going to 2x the demand for energy from AI compute. You know if we go out five years, people are saying it’s going to 4 or 5x. People like Leopold Aschenbrenner think that that’s going to be on a much steeper exponential. Much of the conversation comes back to this energy bottleneck. What’s your perspective on how this energy bottleneck plays out?

Johnston: Yeah, I mean you’re right. We’re on this insane exponential. The forecasts I’ve seen are along the lines of another 50 to 100GW of new energy in the US just in the next three years for AI rollout and then essentially doubling every 3 to 4 years beyond that. So we’re talking about replacing the entire US power grid every year on a ten year time frame.

So another 400GW a year every year to be deployed, um, on a ten year timeframe. And that is just completely impossible to do if you’re going to try and build 400 new nuclear power stations in the US every year. Partly the reason it’s so difficult to do is the permitting constraints. It’s a 5 or 10 year long lead time on permitting to build that, to build one new nuclear power station in space, you don’t have that constraint, and you can scale extremely rapidly on the 5 to 10 year timeframe.

Elon Will Bring Launch Costs Down with Starship

Murphy: What are the components that you think need to be in place for that scale to be delivered on? Is there going to be a Starcloud factory producing compute units? They’re going to be producing solar units. What has to happen for this scale to be unlocked?

Johnston: So to start on the launch side of things, you certainly do need the launch cost to come down and the launch capacity to go up very dramatically. And that’s certainly what we’re expecting with the Starship launch vehicle coming online at the end of 2027 for the first commercial payloads we’re expecting.

So Starship, for people who don’t know, is designed to lower the cost of launch by between 50 and 100x. What’s more interesting is that it’s going to increase the capacity by maybe 500 to 1000. And the reason for that is it’s the first ever fully reusable rocket. And so, for example, with the Falcon 9, if you build a new one every day for a year, at the end of the year, you still only have one Falcon 9.

Whereas with Starship, if you build a new one every day for a year - and by the way, they’re planning on building three every day - but at the end of the year, you have 365 starships because they’re reusable, because this upper stage is reusable. And so that means the payload capacity is five times that of the Falcon 9.

You can fly the booster ten times as frequently as a Falcon 9 booster. So you’re really talking about a five year timeframe. Yeah, thousands of times more tonnage per year that we can get from SpaceX than we currently can. So that’s one big thing. The second thing you mentioned will be producing large quantities of these orbital compute modules that will fit in the Starship will require large manufacturing capability in the US. And so that’s just factories basically.

Murphy: What do you think are the key enabling technologies beyond Starship, will these like data centers be networked? Will they be assembled in orbit? What are the other important building blocks that you expect to see for the space economy to emerge over the next couple of decades?

Johnston: The second biggest one is most spacecraft right now do not have great connectivity. They have to wait for ground stations. And so being able to plug into the constellations like Starlink and potentially LEO from Amazon will definitely be a huge enabling technology. I think low cost launches enable things like manufacturing in space, asteroid mining, lunar resource mining, certainly space hotels and lunar hotels. Those are the big ones that are enabled by low cost launches.

Murphy: Excluding compute, when we go forward 25 years, which one of those do you think will be the biggest driver of the space economy?

Johnston: Compute will be the biggest driver by a country mile by maybe two orders of magnitude more than any other one. The things that support that infrastructure will be the second, so communications will be one. Although arguably once you’ve got more than 100 compute modules in or satellites in orbit, you can do your own backhaul basically.

So not necessarily an outside business, but the other one on a longer term more like a 20 to 30 year time frame will be utilizing the resources of first the moon and then the asteroid belt, particularly, I think. One thing Elon mentioned, which is a very smart, and logical thing to do, is manufacturing solar and chips on the lunar surface and then launching them from there.

Because the moon doesn’t have the gravity well of the Earth, you need about 20 times less energy to get a spacecraft from the lunar surface to Leo to low-Earth orbit than you do from Earth to Leo. And so it makes a lot more sense to if you can manufacture these on the lunar surface and then yeet them to LEO than it does to to manufacture them on Earth.

The 2050 Space Economy

Murphy: And do you think scaled manufacturing and production on the moon is something that’s going to happen like five years, ten years time? Like NASA’s most recent mission is to put people back on the moon in the next couple of years. There’ll be a permanent outpost within the next decade. Do you see those timelines delivering?

Johnston: Within five years, Optimus will be being mass manufactured. And Optimus in a spacesuit requires basically no modification to operate on the lunar surface. You might need to retrain it slightly, but in terms of thermals and radiation, you don’t need to change the Optimus hardware. So putting one human on the lunar surface in five years is going to be much harder than putting a thousand Optimus on the lunar surface in five years.

And I would imagine we will do that first. Before that, if you lose 1000 Optimus, nobody’s going to cry about it. And you can iterate and figure out what we’re wrong and launch quickly and keep iterating. If you lose a human, you’re never going to be able to get funding for any other program after that. And so it’s just a slower, super annoying time to deliver stuff. And also there’s not really much point in having a manned base on the moon. I mean, for tourism maybe, but beyond that…

Murphy: Moon hotels!

Johnston: I’m all in favor of moon hotels for tourism. Great. But once Optimus is up and running, humans are just going to get in the way of a manufacturing process. Like you don’t need humans up there to do any kind of manufacturing. So I think within ten years, I think we will see the beginnings of large scale manufacturing on the moon.

Murphy: And has Optimus been designed to be functional, be it in a lunar or Martian environment already?

Johnston: No. At the moment they’re only designed to work on Earth. So they’re designed with an atmosphere and no radiation. But so are humans. And the way that you get around that is you put in a human spacesuit because a human spacesuit takes care of thermals and radiation. So you don’t need to modify the hardware.

You just need to get this human spacesuit, which sounds a little bit wacky in sci-fi, but, I mean, it’s a very simple solution. And the reason you do it, rather than building a special Lunar Optimus is because this Earth based version will be manufactured at such high volume that it’s much cheaper. I would imagine to just put it in a human spacesuit rather than to set up a manufacturing process for a special space based version.

Murphy: Spacesuits for robots. Big business opportunity?

Johnston: Huge business opportunity. They’re not going to be too different from spacesuits for humans. What you’re solving for is roughly the same, which is thermal and radiation. So and by the way, that will be done by SpaceX because they already have their own spacesuit.

Murphy: This is a conversation that probably comes up a lot for you…if SpaceX will do it.

Johnston: Yeah.

Murphy: So in science fiction series like Alien, there’s Weyland-Yutani, solar system wide corporations. Do you think that that’s the universe we’re heading for with SpaceX?

Johnston: It certainly looks to be heading that way. Another probably reasonable analogy is the British East India Company, where they had just the most insane monopoly for at least a century. But it doesn’t mean because the East India Company was there, that there were no other profitable businesses operating in and around that infrastructure, but they did have an insane monopoly.

Murphy: Taking, say, looking at something like Blue Origin and New Glenn, looking at Stokes Space. Can anyone maybe not catch up, but like even get to where we’re at today compared to SpaceX? Or how do you see that competitive environment evolving?

Johnston: It’s very tough. And the reason is nobody right now is even trying to make it except for Stoke Space actually. I would give them credit, but Blue Origin is certainly not even trying to compete with SpaceX on a reusable upper stage. And so let’s say today they were to say, okay, we want to make a reusable upper stage.

Now you’ve got maybe a 5 to 10 year development cycle just in that. SpaceX have been going on this for at least a decade now, and they still haven’t made it work. So if Blue Origin were to start today to do that, you’ve still got another decade before they’re potentially even have their first reentry over an upper stage.

And then by that point, SpaceX will be in a huge mass manufacturing stage. So I think it’s extremely hard for anybody to catch up now.

Murphy: How much? Just so people can understand how much the cost of the vehicle is the upper stage.

Johnston: For a Falcon 9? You’re looking at an order of magnitude about $10 million that you’re throwing away every time you fly a Falcon 9 in that upper stage. And for a Starship, I would imagine it is quite a lot more expensive than that, because for sure, they will also offer a dependable version. I think it’s instead of 100 tons to LEO with a reusable upper stage, it’s something like 400 tons to LEO.

If you throw away the Starship, I would imagine that you’re probably talking about $20 million for the Starship upper stage.

Murphy: So we’re sitting here in 2050. SpaceX controls the routes, controls the communications. You’re building a business that exists within this ecosystem. What do you think of people looking at the space economy and seeing that future? What are the smart strategies for building businesses? Is it being a vendor and supplier to space? Is it picking particular technology trees? What do you think are the smart strategies here?

Johnston: The way I would look at it is I would say what businesses today do not work, but would work if the launch cost was either 50 times cheaper than it is and or we had 1000 times the launch capability in terms of mass per year. Anything like that is worth looking at. So certainly what we’re doing is one of those space based solar probably will be a thing at some point.

They’re breakeven. Launch cost is way lower than ours, as in, it will be a longer time before we hit that breakeven launch cost. Things like manufacturing and space. But that requires reentry, which is going to be expensive for a while. Space hotels. Asteroid mining. These are the types of businesses that if the launch cost was much lower.

That’s the type of thing I would be going after. And then maybe supporting infrastructure. So robotic spacecraft maintenance is a big one that people are looking at other things like that.

Murphy: Asteroid mining is something that’s been talked about for like a very long time, going back to like orbital materials. There’s been companies that have had Sergey Brin, everyone around the table putting hundreds of millions of dollars in, and you see this YouTube video of X asteroid in the asteroid belt has 14 quintillion tons of gold.

And that obviously, if you were able to bring that gold to earth, the value of gold would collapse, ending the economics of it. But what have you seen that gives you a belief that, on this timeline, we’re going to see effective asteroid mining?

Johnston: I really am a big believer that within 10 to 20 year time frames, asteroid mining will be a big business. And it’s purely a launch cost calculation. So the reason that mining is so expensive right now is that having a failure on one launch is so expensive, because flinging stuff out to the asteroid belt is very expensive.

Right now that your iteration cycles are extremely slow, you can’t take risks, and in order to make massive efficiency gains, you need to take risks. And I think we can see this in the Starship program. To me, it’s ever so slightly worrying that Blue Origin has not had a failure on a New Glenn yet, because what it means is they’ve massively over provisioned everything.

If you see what I mean. Like the reason SpaceX allows failures in the test campaign is they want to know at what point it’s going to fail. And if you don’t test a failure, you don’t know at what point it’s going to fail. And what that means is you’ve just got a huge amount more mass probably than you would need in order to have redundancy and make sure it’s not going to fail in any condition, which you don’t necessarily want.

And the same is true in the asteroid mining business. It’s so expensive to get stuff out there that the iteration cycles you’re launching may be one every year and a half. You need to be launching like one every month in order to have this quick iteration cycle. And that’s coming with Starship.

Murphy: And so this would be Earth launch, Mars assist orbital insertion? Or do you think we’re capturing asteroids and bringing them into closer orbits?

Johnston: I’m not sure you need a Mars assist. I think you can just go straight to Earth to the asteroid belt, and then we don’t bring them back. I think we mined 20 tons of gold, and we just brought 20 tons of gold back. If you’re doing platinum, you can do it. I think each mission, I think they’re looking at bringing back something like $50 million worth of platinum. If you can get out there and back for less than $50 million, that’s a good business.

Murphy: So what are the technologies that need to be developed? It’s again Optimus. So you have a bunch of robots that you’re sending out on their robot mission. They’re landing. You’re doing this orbital insertion. You’re landing on the asteroid. You’re preparing whatever or happens to be there. You have enough fuel to do the return mission when you…

Johnston: Yeah. When you get Optimus, everything becomes a lot easier. Right now, what they’re doing is so asteroids are they don’t have a huge amount of gravity themselves. So you’re not really landing on the asteroid, more like docking on it. So you’re both moving through space like this, and you’re kind of docking like this, which means that regolith is like spewing everywhere.

And it’s actually a pretty tricky environment to navigate. And then trying to drill into that refine with like a robotic arm, which is kind of essentially what they’re doing right now is all unbelievably complicated. But if you could stick a few Optimus tethered on the outside to try and pull themselves around, I think things will get a lot easier, a lot quicker.

Murphy: Well, I suppose we had that sample return mission recently...

Johnston: Incredible, incredible mission. But if it brought back like this much sample and the engineering that went in to get that much sample back, just mind blowing. So yeah, I think we do need the iteration cycles to speed up a lot.

The 2050 Martian Colony: A Stepping Stone for Exploring Space

Murphy: Up until space compute. Obviously the big space future business model was Mars, and now we’ve maybe space computers going to fund the future Martian colony. Let’s go to 2050: Do you believe there is a martian colony?

Johnston: Absolutely.

Murphy: What are we doing on Mars that makes this the right place to go?

Johnston: We can send 100,000 Optimuses to go pimp it out first. We’ll send a few pathfinder missions with humans just so we can say we landed a human there, but they’re not going to do anything, really. They’ll just hang out. We’ll send shitloads of optimists to go beef it up and make it a nice place. All of the domes with the waterfalls and the trees inside and all that crap, that’s coming for sure.

Murphy: Even in a low gravity, high radiation, you’ll just build the ship out.

Johnston: We’ll figure that out.

Murphy: One of my pet projects is that I think people are wrong about Mars. And actually, we should be looking at Venus. When you think about the why of Mars, do you think it’s the long term terraforming prospect? How do you think humans operate there? You’re in an environment where you are just subject to lower gravity and therefore like biological degradation.

Johnston: As I understand, there’s ways you can deal with some of those gravity effects. I mean, like you can do exercises every day that kind of make it more manageable. I like Mars because it’s a stepping stone towards the moons of the gas giants. I think Mars is quite strategically placed. If you go inwards towards Venus, you ain’t going anywhere from there. You’re just going back to Earth, probably. Whereas Mars is a decent stepping stone to Ganymede and Europa and Titan.

Murphy: You’ve got volcanism: Europa. You’ve got subsea oceans: Ganymede. You have a similar kind of environment on Titan. You have these methane oceans. What inspires you to go there? Do you think it’s just exploring the final frontier?

Johnston: I think it’s as the final frontier it’s worth doing. For that reason, I think we can make decent, habitable only rings, though. You probably wouldn’t put them that far out. You’d probably put them in like the lunar elves.

Murphy: Can you talk a little bit more about this? I think most people don’t know about O’Neill Rings. I suppose these are world ships.

Johnston: Yeah. So you can imagine a long cylinder that’s maybe a kilometer, a few kilometers wide that rotates at a speed that if you were inside on the inside of this cylinder, it would simulate Earth’s gravity. So you can have maybe 20km long and a few kilometers wide. So if you were to want to build a large colony in space, and it doesn’t mean you could have it many kilometers wide as well, it doesn’t need to be only a few. If you wanted to build a very large colony in space, I think that would be a good way to do it.

Murphy: Is that your vision of how humanity moves out from beyond Earth?

Johnston: I’m very keen on trying to get to the nearest star to Alpha Centauri. For the next few thousand years, we’re fairly confined to the solar system. It doesn’t take very long, actually, to get to Alpha Centauri, so probably you could do it. If you were to minimize the amount of fuel you would need, it would take about 57,000 years.

And you can do that with today’s technology. You could literally launch that in like in a year or two if you wanted. I think that’s worth doing.

I quite like the idea that it would definitely be the last Earth probe to arrive in Alpha Centauri, but it would be the first to leave in the last to arrive. Presumably there’ll be better technologies. So we could have been there for at least 10,000 years.

So I imagine they’ll upload it across the finish line. And the colony of Alpha Centauri will capture a probe that was launched now and put it in the Grand museum of the Alpha Centauri colony, and then it would stay there for another few million years, or however long any future civilization that wants to leave from our solar system to Alpha Centauri that is optimizing for mass and not for speed, will arrive after us.

If you’re optimizing for kilos per unit of fuel, you’re going to take the exact same trajectory, except if you leave in a year, you’ll arrive a year after, so you’ll want to. You’re aiming for the tangential intercept. So Alpha Centauri is coming towards us about 20km a second. It will pass by us. It will be the closest point in about 17,000 years.

But if you aim for that point, you need more fuel than if you aim for where it’ll be just after that. They call it the tangent to intercept. That’s where you intersect at 57,000 years.

Murphy: But what if you actually want to enter orbit?

Johnston: If you want to enter the Alpha Centauri orbit? Yeah. You need to figure out a way to slow yourself down at that point, which is a lot harder.

Murphy: Yeah, I feel like the other sci-fi version of this is Avatar, where you see the ships burning fuel [to slow down].

Johnston: Yeah, but actually it’s not that difficult. Like, so you need 20km a second of ∆v as in change in velocity from Earth. You could fling yourself straight to Alpha Centauri from here with that. And that’s basically if you filled a 600kg spacecraft with xenon and put a beefy ion thruster on the back of that, you can easily get that, and it will cost less than $10 million.

In order to get there and then slow yourself down, you need 40km a second rail to be because you need 20km a second. Speed up, and then you turn yourself around. And as you approach another 20km second to slow stuff down, which is not like that’s definitely within the realm of sci fi.

Murphy: What do you think are the key technologies that we haven’t talked about that enabled that phase of building out human civilization beyond Earth?

Johnston: Yeah, I think certainly long duration power supplies are one. And people have talked about either fusion or fission in space. That’s definitely a very key enabling technology, because once you get past Jupiter, if you’re getting 100W/m² here, you’re getting one square watts per square meter here.

So you can’t really use solar when you get past there. The Voyager probes have a decent solution. They have this radioactive decaying material that produces a temperature gradient for, for electricity. But things like that very long duration energy supplies are definitely one of them sleeping pods.

Certainly not within our own solar system. You don’t really need anything like that. Like freezing pods. To get humans to Alpha Centauri, you probably need some kind of freezing and unfreezing. Or you could build a massive spacecraft that people can have many generations of children on, but that seems a bit pointless and impractical.

Murphy: What do you think are the aspects of building in space and the technologies? I think Apollo is famous for the amount of spin off technologies, like the Biro pen was the reason we have that is because of space exploration, whether it’s wireless TV transmissions…

What do you think, if you take this 2050 viewpoint, are going to be the technologies that have been built for space that will end up having the biggest impact here on Earth?

Johnston: There’s one technology that we’re building that I think has very wide applications, but they’re mainly other space applications, and that is building a very large, low cost, and low mass deployable radiator to dissipate the heat from the chips that were where they were running in space. It’s useful for data centers in space, but it’s also useful for basically every other application of needing to get rid of heat in a vacuum.

And so that’s definitely things like space hotels, astro mining, refining, materials manufacturing and space. But to come back to your question, I certainly think all of the communications technology is going to be very valuable for people just living on Earth. You can already see that with Starlink.

Murphy: Do you think all telecommunications will be delivered via satellite?

Johnston: Elon recently gave an interview where he said very dense cities will never be delivered by satellite. If you were to want to deliver, say, Manhattan via satellite, today, you probably could if you just had an enormously dense constellation with a very high bandwidth. I think what he’s saying is by the time we have that constellation up, everybody’s going to be doing video streaming almost continuously.

And the data rate needed is going to also increase. And by that point it’s just impossible to deliver. So you’d still want a local fiber network.

Murphy: So some things will change. Some things will stay the same. And just because you mentioned it like the space hotels. When do you want to go to space?

Johnston: I think I’d be happy being the thousandth person in a space hotel.

Murphy: Why do you think Elon hasn’t gone to space?

Johnston: He’s been pretty explicit about it in some interviews. And that is, he thinks if he were to die, it would massively slow down the space program, essentially. And he’s more concerned that the space program exists than that he goes to Mars. And he said he wants to die on Mars, but not on impact.

Murphy: That’s a very reasonable proposition. And so when the thousandth person goes up, realistically, the place for humans is going to be in leisure activities. We talked a little bit about space hotels. What are the right ways to build a business like that? Is SpaceX going to handle getting you there. What do you think people are going to want to do in space?

Johnston: So there’s a few things I’ve been thinking about. Luxury luxury hotels on the moon definitely make sense to me. You would want to land on a new moon as it is from the Earth. And so the Earth is. The sun is directly behind the moon, and so the Earth will be a full Earth as you’re viewing it from the moon on one horizon behind you with the sun.

The other horizon will be a full Earth. And then over a three day period, the sun will slowly rise. There’s a bunch of craters there. I would want an infinity pool lunar hotel on this crater, looking towards this full earth glass of Veuve Clicquot.

Murphy: And like the champagne, what does a pool feel like in 0.1 G?

Johnston: So pooling 0.1 G. It’s actually fascinating. We don’t know because we haven’t ever watched it. But you can infer. I mean, one of the cool things you could do is you can swim like a dolphin. So you can propel yourself out of the water, and then it’s going to be a messy affair because the water is going to go everywhere.

If you build a big enough dome, then all is well.

The Kardashev Scale of Civilization

Murphy: Could you maybe talk a little bit about the Karsashev civilization scale and like where we’re at? Talk us through a little bit about what that scale means.

Johnston: So Kardashev, a Russian physicist or rocketeer, proposed this scale of civilization, which is where you judge a civilization’s progress based on how much energy it’s consuming or turning into useful work. So you have three different levels, although I find it extremely frustrating, there aren’t four levels because to me that would be complete.

But the three different levels are Kardashev level one, which is where you’re using the incident energy on your planet from the sun or from your star. So we’re about 1% of the way or less than that, to being a Kardashev civilization level one. Kardashev level one would be if you cover the entire surface of the planet with solar panels, you would generate Kardashev one.

Kardashev level two is where you surround a star with solar panels. That’s kind of like a Dyson sphere. And then Kardashev level three is where you’re consuming or turning into useful work. The incident energy of all stars in a galaxy. So you’re talking about a few 400 billion stars in our own galaxy.

If you had built Dyson spheres around all of them, that would be Kardashev level three. I think the reason it’s stopped there is well, there’s two reasons to stop there. One: people thought Andromeda was a nebula when Kardashev proposed the scale, so he didn’t realize there were other galaxies besides the Orion galaxy.

And number two: the thing that’s become more apparent recently is no matter how fast we go, even if we’re growing at the speed of light, we can only get to the local cluster of galaxies because of the expansion of the universe. So we actually, in theory, couldn’t become a Kardashev Level four civilization.

Murphy: So you don’t believe in faster than light transportation?

Johnston: It’s above my pay grade. I don’t think my brain is big enough to comprehend something like that!

Murphy: And with this Kardashev scale, most people would be like, we only use 1% of the energy from the sun today on Earth. What do you imagine we’re doing with that energy? If civilizations were to be able to capture it?

Johnston: It almost certainly is data processing for simulations. If you were to want to have a bunch of consciousnesses running in a simulated reality, and you could potentially have much more complex consciousnesses than us running in much more complex simulations than the world we see around us. That would require almost unlimited compute.

We Live in a Simulation: The Fermi Paradox & The Great Filter

Murphy: Why do you think someone would want to do that?

Johnston: For the same reason that we create simulations, computer games for modeling because people enjoy living in simulations.

Murphy: So do you think we’re living in a simulation?

Johnston: I think it’s almost certain. Yeah. Okay, there’s two possibilities, and I’m open to both. One is that we’re heading to calamity within the next, like, pretty soon, like next few hundred years. And the second is we’re in a simulation. I don’t think there’s any other possibility besides that.

Murphy: Is this because of the Fermi Paradox?

Johnston: Yeah. The Fermi paradox says that we should see life everywhere in our galaxy, and we don’t for some reason, assuming that we are not staggeringly rare, which is also a possibility. But we’ve been going now for 14 billion years, and there’s about 400 billion stars in our galaxy. Each one has ten planets.

Let’s say if any meaningful proportion of those had had life on them at any point in the 14 billion year history, it really would from there only be a couple of million years until you colonize the whole galaxy. And we would see life teeming throughout our galaxy, it would not be difficult to find life in our galaxy, and we don’t see life anywhere.

There’s a bunch of bogus theories about what they could be hiding and all this crap. No evidence for that. And it’s a bad theory in my opinion. The two solutions are intelligent. Life is pretty short lived for some reason, as it doesn’t survive more than a few thousand years needed for it to develop technology to expand to other stars.

That’s number one. And that’s honestly what I think is the most likely intelligent. Life was pretty short lived. The other is that we’re staggeringly rare, given that life only took 200 million years to get started on planet Earth. I think that’s an unlikely one.

Murphy: I know what you’re referring to is the great filter of there’s some point at which intelligent life doesn’t come past. Yeah. Some people thought that creating nuclear weapons was the moment that most intelligent life wipes itself out. Some people believe that it’s the creation of AGI, which I feel doesn’t really make sense because then we would see machine civilizations.

Johnston: The machines themselves are pretty self-destructive. I mean, a swarm.

Murphy: So you think that AGI would destroy itself as well?

Johnston: Yeah. Pretty much. I mean, a swarm of attack killer AI drones would make mincemeat of both themselves and the planet pretty quickly. And it looks like the two major superpowers are building swarms of AI killer drones as fast as they can. So yeah, it would not surprise me if machine intelligence was pretty self-destructive as well.

Murphy: How do we get past this great filter?

Johnston: Well, first, people need to recognize it. And that’s one of the reasons I’m quite passionate about the idea of sending a probe to the nearest star. Because people say, if it’s so easy to get to the nearest star, why haven’t we sent a probe? And then the answer they’re expecting is because we don’t know if we can get past the Oort cloud, or because we haven’t got the propulsion technology and none of that’s true.

We actually literally can. We just haven’t yet. So we should do that as soon as possible. And so people can be like, oh shit, we’ve just sent a probe to the nearest star. Why do we why don’t we see probes coming in from every star? Like we should see them? Like they should be everywhere. So that’s number one.

So we need people to recognize that the Fermi paradox is a thing. And this, I think, is a way to draw attention to that. I have a pretty fatalistic, nihilistic perspective, and I would want to do everything I can, but it’s not like other civilizations wouldn’t have thought the same thing. Other civilizations are going to be like, hang on a minute.

There’s this Fermi Paradox, like where the fuck is everybody in our galaxy? We should see, like, Civilization and then. Oh, fuck, we just killed ourselves. Like they would have had the exact same experience happen. It’s not like we’re particularly unique that we would have realized that there’s this phenomenon.

So it leads me to be very pessimistic, to be honest my p(doom) is extremely high.

Murphy: Is it high because of the progress in AGI right now?

Johnston: I don’t even want to hypothesize about what the cause is. It’s high because of the Fermi Paradox. It’s high because we don’t see intelligent life in our galaxy. Now, AGI probably increases it because what it does is it makes me very confident that we could get to Alpha Centauri very easily, like seeing the rate of progress in robotics and propulsion and intelligent design and everything else makes me like, well, it’s like blindingly obvious that if you run this for 200 years, we are sending stuff out of Centurion.

It’s going to get that fast. So what happens in the next 200 years that stops that? Because why hasn’t it happened on any other planet? I don’t want to say it’s definitely an AGI. I don’t want to say it’s nuclear. It could be many different things and it also could be one of the more benign filters. It could be things like, we can’t get past the Oort cloud, but I don’t think it’s that.

The reason I don’t think so is that Dyson spheres would be extremely obvious. They have a very clear signature, which is an infrared signal with no light signal, basically. Actually, I have heard some other interesting theories about this, which are both bad theories, but I’ll tell you them anyway.

The first is that 80% of the mass in our galaxy is dark matter, and it’s not evenly distributed. In fact, it’s quite patchy. Dark matter has basically a similar signature then you would expect from a Dyson sphere, which is a gravity signature with no infrared or light signature. A matrioshka brain wouldn’t have an infrared signature.

And so there’s an argument to be made that 80% of the stars in our galaxy are already matrioshka brains. Some other people have argued that there’s billions of black holes in our galaxy alone. There could be a black hole in our solar system and we wouldn’t know it.

Murphy: This isn’t the theory for the Planet Nine theory, it is that actually the orbits within the solar system are explained by a like basketball sized primordial black hole somewhere out between Neptune and the Oort cloud. Interesting. But do you think a lot of this just begs the question that there’s still so much we just don’t know about the universe?

Is this the most important question that we need to find out? The answer is like, why is there not life in the universe?

Johnston: It’s the most pressing in terms of our own existence and survival. There are a few other questions I’m very interested in. I’m interested in understanding the hard problem of consciousness, for example, and a few others.

Cracking the Hard Problem of Consciousness

Murphy: To go into the hard problem of consciousness, do you think consciousness is something that can be created in the digital space?

Johnston: Very much so. There’s a company called Cortical Labs. It’s a bunch of PhDs in either Singapore or Australia. They are basically isolating individual neurons, and they’re simplifying them down to really the most basic components of a neuron. And they’re building up computer systems out of neurons.

You know exactly what’s happening in each neuron and is very deterministic. And there’s nothing magic about it. And you’re just stacking them together. Now, that thing will be conscious because there’s no difference between that and your brain. So you definitely can create consciousness in a lab and in a petri dish.

Murphy: But what is your definition of consciousness? Some neurons firing?

Johnston: As far as we know, that’s the only evidence of consciousness. We or that’s anything we know if consciousness, because that is the thing we are when a human is talking about consciousness, when I say, for example, I see the blackness or the whiteness of this thing or the yellow ness of this thing, and I have this yellowness sensation, and it has this kind of like magical quality, because inherent within a deterministic atom system, there shouldn’t necessarily be or doesn’t need to be a yellowness quality.

If you were to analyze the software going on, as I’m describing slowness, you could. If you were to have a very clear understanding of that software, as I’m describing the yellowness, it would all be deterministic. I think that’s where we’re going to understand. That’s where we’re going to find the end state of understanding consciousness is to understand what it is that makes the software describe: “I have yellow-ness experience.” That is not an impossible problem to solve from a physics perspective.

So I think we will crack consciousness. I think the hard problem is crackable. To me it’s not magic.

Murphy: But do you think humanity’s long term mechanism of exploring the galaxy will end up being the like beings and consciousnesses that we create?

Johnston: Definitely. Elon even tweeted about it. He said, Optimus is the von Neumann probe, where the von Neumann probe being a self-replicating probe that you send to one star, it replicates and then it sends ten to the next nearest star. When it gets there replicates, then it sends ten to the next. Definitely. I think Optimus is a good candidate for that. You send a 100,000 Optimus, they build an Optimus factory on Alpha Centauri, and it sends out 100,000 Optimuses…

Murphy: It’s kind of like bringing all of this back to Earth. Obviously, we want to make it past the Great Filter, wherever that may be. But if sentient robots and AI are exploring the galaxy on our behalf, we’re mining asteroids and bringing back gold. Platinum resources become abundant. What do you think that will change on Earth?

There is a Calamitous Event Waiting for Us

Johnston: There’s what I hope and what I think I hope for this world of abundance. And it’s certainly possible there’s nothing against the laws of physics to stop that happening. But as I say, what I think is we have a calamitous event waiting for us and probably manmade because of the Fermi Paradox.

Murphy: Between now and then, what can I do, what can you do, what can everyone do to prevent that happening?

Johnston: I’m imagining the millions of times this question has been being asked throughout the galaxy over the last 14 billion years. And the answers they would have given each time would have not worked. Certainly we need to avoid going to war. That is like the key thing we can do to avoid this is to avoid going to war.

Now, a simplistic answer to that would be a one world government type situation, but that also doesn’t seem to have been very effective in stopping wars. Although arguably the UN has been pretty effective at stopping wars and the EU. My best guess is you need to try as hard as possible not to go to war. And probably my best guess at that is more collaboration at an international level, through bodies like the UN and others.

Murphy: Do you think exploration of space provides a medium like the ISS is still, even today, one of the rare overlaps of like US-Russia collaboration?

Johnston: Definitely.

Mars is the Next America

Murphy: And it’s just because we’ve got here and I feel like this is probably a rabbit hole. We could go down for some time, but the last new country on Earth is South Sudan, created in 2014. When you think about this ability to move to the moon, you think about the ability to go to Mars. What will the governance of those places be? Will it continue to be the same structures we have on Earth? In sci-fi there’s typically always been like The Martian. Yeah, the Martians rebel…

Johnston: That’s definitely gonna happen for sure. There will be a Tea Party rebellion. They don’t want to pay tax to Earth anymore because it makes sense. Like that type of over taxation. Rebellion is very common throughout history because in order for us to pay for a US colony, us being the British in that instance were required an enormous amount of resources, and then they want to recoup the resources.

But over time, the people that live there are like, in fact, this we ain’t going to keep paying you. And that’s probably going to happen again. Like we’re going to make your calling. It was unbelievably expensive. We’re going to want to suck resources back to Earth because we paid for it. And then the Martians are going to get pissed off with that.

It’s a story that plays out again.

Murphy: So you think the next America, the country on the frontier, will be Mars or Venus? That’s where the cutting frontier of liberal values. Frontier-ism. Definitely exploration. That’s where it’s going to be.

Johnston: Without a doubt. You already see it on the West Coast of the US. The West Coast of the US is the furthest you can go without being in the Pacific Ocean. And that’s why all of the space industry is on the West Coast. Because they got air and they’re like: I guess we’re going up there boys!

Murphy: I feel that’s a pretty good place to end it. So thank you so much. This was such a good conversation. We’re going to have to eventually, in 25 years, do this in the cloud cities of Venus. We’ll lock it in. Thanks again. This is great!