Can we explore the galaxy with self-replicating probes?

00:00:07
Speaker 1: It's a vast, cold, and mostly empty universe, and as lucky as we've been to live in a time when we are discovering planets around other stars, we still have no evidence that there is anyone out there living on any of them. The prospect that we are the only life in the universe, or the only intelligent life, still haunts us. Of course, people say the aliens could be out there even if we haven't found them. So let's flip the script and think about it from the aliens point of view. Why haven't they found us? Could they find us? There are lots of planets in the galaxy, after all, and we're only on one of them. Today in the podcast, we're going to explore the prospects for a technology that promises to unleash vast exploratory power, one that might allow us to visit any planet in the galaxy and to look for aliens in a reasonable amount of time, or to allow aliens to have come and visited our planet. That technology is self replicating probes spacecraft that build more of themselves, growing our fleet exponentially. Because if we can do it, we probably will do it, and if we can do it, aliens likely could also welcome to Daniel and Kelly's extraordinary and so far barren universe.

00:01:38
Speaker 2: Hello.

00:01:38
Speaker 3: I'm Kelly Reader Smith.

00:01:39
Speaker 4: I study parasites and space, and I had so much fun in today's conversation.

00:01:43
Speaker 3: But also I'm not going to sleep tonight.

00:01:46
Speaker 1: Hi. I'm Daniel. I'm a particle physicist, and though I'm technically self replicating, I haven't made any exact copies of myself.

00:01:54
Speaker 4: Oh, I guess no human has made an exact copy of themselves.

00:01:58
Speaker 1: Yet, although you know so my two children are both blonde, and my wife is blonde and I'm not, and so people joked occasionally, since she is a biochemist, that maybe she had just cloned herself in the lab.

00:02:12
Speaker 3: Hazel does look very Katrina issue.

00:02:15
Speaker 1: Yeah, yeah, they both. Do you know those dominant Viking genes.

00:02:18
Speaker 3: I guess they wiped out your genes.

00:02:22
Speaker 1: I'm not unhappy about that. They both look great.

00:02:25
Speaker 4: So my question for you today, you are super excited about aliens.

00:02:31
Speaker 3: Yes, no news to anyone there.

00:02:34
Speaker 4: If you could make self replicating probes and release them into the universe so that you could communicate with aliens, would you do that?

00:02:41
Speaker 1: M Oh my gosh, Well, you know, I'm on the record for being willing to invite aliens to come visit Earth as long as they share the secrets of the universe, even if they send us to the hydrogen minds and enslave us. That's how badly I want to meet aliens and learn how the universe works.

00:02:58
Speaker 3: I'm so glad academics have like zero hour over anything.

00:03:02
Speaker 1: Well, that's what makes me free to say these ridiculous stus, because I'll never ever be around that table where we make these decisions. So I mean, if you're asking me, like Daniel, would you launch self replicating probes that could put us in touch with aliens that might be able to tell us the secrets of the universe, even if it risks like wiping out all matter in the galaxy because the probes go crazy and convert everything into paper clips essentially, Then yes, I think I still would because the other alternative is too disappointing, Like we are stuck on this planet and the aliens are stuck on that planet, and we can't talk to each other because we're afraid that our technology is going to run amok. I don't know. That just seems too dark.

00:03:46
Speaker 4: That's more dark than all of us getting to still exist but not talking to each other.

00:03:51
Speaker 3: You're warped, man.

00:03:53
Speaker 1: Yes, yes, I'm warped. I will totally admit I'm warped. But I want to know who's out there. I want to know if our other civilizations, and the idea that there could be out there and we never find them, that's to me just too difficult to accept. So any technology that's going to help us make contact with other civilizations. Yes, I'm a booster. I'm an investor, all right, So.

00:04:15
Speaker 4: If you are anything like me, you might be thinking, one, I hope Daniel never runs for public office, and if so, he does not have my vote. And two, I want to know how likely is it that we could build these self replicating robots so that if some Daniel Ooyd individual gets a position in Congress, do they have a chance at wiping this all out with these self replicating robots?

00:04:36
Speaker 1: And by self replicating robots, Kelly is referring to not just probes that NASA builds here on Earth, bespoke things that take ten years, and we send out one of them and we cross our fingers and hope that it survives, or even two or ten of those. We're talking about probes that can make more probes, Probes that are not birthed here on Earth, but out somewhere in the galaxy, maybe five ten generations down, that let us tap into the power of exponential growth so that we can effectively explore the entire galaxy without ever leaving our rock.

00:05:08
Speaker 4: Amazing, and so we asked our extraordinaries, what do they think can we even build self replicating probes to explore the galaxy?

00:05:18
Speaker 1: Is this actual technology that's around the corner, maybe a few hundred years from now, or is this just Daniel's fantasy? Thanks very much to everybody who answered these questions. Think about it for a moment. Do you think we are around the corner to building self replicating probes? Here's what our listeners had to.

00:05:34
Speaker 2: Say, Yes, we can.

00:05:36
Speaker 1: It will take a while until we can, and they might eat us afterwards.

00:05:40
Speaker 3: It's plausible that we are somebody else's self replicating probe in.

00:05:45
Speaker 2: Terms of practicality, I mean, you know, the short answer is no. The long answer is well, no, I.

00:05:54
Speaker 5: Don't think our current engineering precision is up to the tech of making probes that would in turn make probes with equally good precision. And so on all the way down.

00:06:10
Speaker 6: You know what would happen if this w a ry and we had bands of world destroying robots roaming the galaxy. That seems like, if it's not already the plot of a sci fi novel, it should be.

00:06:23
Speaker 2: You know what, we don't know. We really just don't know yet.

00:06:26
Speaker 5: So I feel like, yeah, maybe in the next twenty thirty fifty hundred years.

00:06:31
Speaker 2: We can't write now, but we won't be able to later.

00:06:33
Speaker 1: I mean, we can do anything if we.

00:06:35
Speaker 6: Try once we have sufficiently advanced AI.

00:06:39
Speaker 7: I can't see what the point is unless we also invent near to speed of life travel. Once they got to another solar system, we'd be long gone. So who would they actually be exploring for.

00:06:54
Speaker 2: We can't even build self replicating robots that'll do our dishes and our laundry, Yet in a near future dish might very well be pushable.

00:07:04
Speaker 3: Okay. I absolutely loved these answers.

00:07:06
Speaker 4: There was a lot of diversity in these answers, and as always, there was some like hilarity in there too. I liked the short answer is no, long answer is.

00:07:16
Speaker 3: Correct me up.

00:07:17
Speaker 4: But yeah, lots of people thinking we could do it, some people saying why should we do it? Some people say we could do anything. I loved it.

00:07:24
Speaker 1: Some people on the Kelly whipb blanket side saying they might eat us afterwards. You know, that's fair, that's fair.

00:07:30
Speaker 4: I think it's important to point out the possible pitfall so we can make decisions with clear eyes.

00:07:35
Speaker 3: Maybe that's just me.

00:07:37
Speaker 1: That's good for when we have that Oppenheimer moment. You know, are we going to push signs forward at risk of destroying everybody? Or are we going to cower in the darkness? Those are really the only two options available.

00:07:48
Speaker 3: Okay, do not run for Senate.

00:07:51
Speaker 4: So so today we got super lucky because we have an amazing guest on the show we do.

00:07:57
Speaker 1: We got to speak to Phil Metzger, who actually has a patent for space concrete. He knows what he's talking about. He's really thought about building industry off the Earth, and he disagrees with Kelly about stuff in space. So we thought who better to come on the podcast.

00:08:13
Speaker 4: I think our disagreements aren't about like the facts. I think our disagreements are about optimism. And he is a more optimistic human being that I am, and he's probably happier for it.

00:08:23
Speaker 1: Yeah, and I think disagrees with Kelly describes like a vast swath of the space industry community.

00:08:31
Speaker 3: Yep, yep. And that's a okay, all.

00:08:34
Speaker 1: Right, So let's jump into our interview with Phil. It's my pleasure to introduce to the podcast Professor Phil Metzger. He's a planetary physicist with the Florida Space Institute. He has the distinction that he has designed spacecraft. He also has a three mile wide asteroid named after himself, and he has strong opinions on the definition of a planet. Plus he has studied the issue of building industry in space, so he actually knows what he's talking about. Phil, Welcome to the podcast.

00:09:02
Speaker 2: Hi glad to be here.

00:09:04
Speaker 3: So can I dig into one of those things?

00:09:05
Speaker 4: So I know Phil pretty well for his work on like you know, regolith on the Moon and stuff like that, but I don't know about Phil's strong opinions about the definition of a planet. So is Pluto a planet or not a planet? What am I missing?

00:09:18
Speaker 2: I'm going to say it's a planet. And that's because the definition of a planet, going back to the Copernican Revolution was not based on orbits. It was based on the geophysical nature of the objects. And that was really a crucial part of the Copernican argument.

00:09:34
Speaker 1: What do you mean the geophysical nature? You mean, like, is it mostly spherical?

00:09:37
Speaker 2: Well before the Copernican Revolution, they thought that the majority of view is that planets were made out of unchanging ether and they were perfect spheres. They followed heavenly physics, not earthly physics, and so the Copernican Revolution said, no, Earth is in the heavens, and these objects are geological bodies, just like the Earth is. And the primary example they had was the Moon because they could see it with telescopes. Galileo saw mountains and those the existence of mountains and the existence of earth shine reflecting off the Moon allowed him to create arguments about this category of objects called planets. And the category he was arguing for was all the geological bodies in our Solar system, including the moons of Jupiter which he called planets, and our moon, and it was all based on the fact that they are geological bodies like the Earth. It was not broken into what they orbit. Now. Kepler introduced the category of secondary planets, meaning a planet that orbits another planet, and that was the primary term, the technical term we had for that subcategory of planets all the way until well as early in the nineteen hundreds that this taxonomy got lost, and it was for non scientific reasons. We ended up with the terminology that's most commonly used today.

00:11:00
Speaker 4: If it were up to you, we would have like hundreds of planets then, because every moon would be a planet.

00:11:04
Speaker 3: Is that right?

00:11:05
Speaker 2: Well if there, we've refined upon Galileo's definition since then, and we now understand that there are small bodies that wasn't known at the time, and we need to have a lower size limit because they become dissimilar and the category is not useful if we include everything down to a dustpec. And so it was Kuiper in the nineteen fifties who proposed the lower limit based on gravitational rounding. He didn't understand the planet formation exactly. And since then we've refined our understanding of planet formation, and so Alan Stern and well Alan I think first proposed a refinement to Kuiper's definition where he said it didn't matter the formation process, if it ended up large enough to become gravitationally rounded, then it should be a planet. So that's the history in a nutshell.

00:11:57
Speaker 1: Well, maybe the solution is buried in your previous comment, and you know, maybe this category is just not useful. It's sort of historical and archaic and reflects our feelings about the importance of the Earth, and that now we're doing all this like layers upon layers upon layers to try to preserve it as a thing. Maybe we should just give up on it and accept the fact that the Solar System is filled with all sorts of stuff, from tiny specks to huge blobs.

00:12:20
Speaker 2: Well, you bring up a great point, and I've heard Neil de grass Tyson say the same thing, that maybe we should say that planet is not a useful category. But you know that's really the outcome of the way it's currently defined. It's not useful. But if you went back to the Galileo and you know, the refinements to the Galileo definition, then it actually is useful again. And the idea is that planets are unique in the cosmos because those are the locations where chemical complexity develops and geological complexity and biology emerges, and civilizations emerge. And that's an important concept in understanding our place in the cosmos, And in fact, I would argue that might even be the most important concept I think planets not only is it a useful concept if you go back to Galileo's definition, but it might be the most important concept in physics and understanding why we're here in the cosmos.

00:13:17
Speaker 1: And incredibly that actually provides a transition to the topic of the episode. So this whole thing wasn't just a digression, because imagine that we wanted to explore the galaxy and to look for other civilizations. Where should we look, right, should we look in the hearts of stars, should we look in stellar atmospheres, or should we look on planetary surfaces? So from that point of view, it's helpful to define like our target locations in the galaxy. So, Phil, if you could look anywhere, is that where you would look? You'd look for things we currently call planets or the field definition of planets, and look on their surfaces for civilizations.

00:13:53
Speaker 2: Sure, and that is what we've been doing. We've been looking for exoplanets, looking for biosignatures. There's also a lot of interest in the large moons of our own solar system, like is there life under the ice on Europa? Even people talk about on Pluto Deep under the surface of Pluto. It's believed that there may be a liquid ocean still liquid, surprisingly, and maybe there's life because there's organic material on Pluto, and there's energy that it's kept at liquid this long energy from nuclear decay apparently. And the literature uses the word planet including those types of objects. So the people that are actually looking for life and looking at geological complexity do use the Galilean definition of a planet just by default.

00:14:40
Speaker 1: All right, So let's say we want to explore all the quote unquote planets around all the stars in the Milky Way and look for civilizations we can chat with. Why can't we just scale up what NASA is doing and do a lot more of it. Why do we need to consider self replicating probes.

00:14:58
Speaker 2: Yeah, that's a great question, I think. And the reason why I would argue is because the people or the civilizations out there that we might detect that might be communicating may not be biological. It may be that they have transitioned so that there's now machine life in the cosmos, and machines are can be designed to be more inherently capable of long distance travel within the galaxy. They could be designed to withstand the environment. They could also be immortal, live for a very long time. Not get bored. Just program yourself to not be bored. It's hard, and.

00:15:39
Speaker 1: So I've tried that for myself. It just doesn't work.

00:15:43
Speaker 2: Yeah, And so it gets back to when Kardishov Alexei Kardashev was looking for signs of life and the cosmos. He defined type one, type two, and type three civilizations because he was pointing out that what we're looking for might not be similar to what we have here, and self replicating probes is one of the common channels people have discussed of how civilization might go at a larger scale and end up colonizing the galaxy.

00:16:12
Speaker 1: I think you're saying that what we should be looking for is aliens self replicating probes. Is that the comment you're making. Are you saying that it's important for us to send us self replicating probes because they're more likely to have a fun conversation with alien self replicating probes.

00:16:27
Speaker 2: All of the above, I think that it's important for us to get beyond the limits of our biosphere here on this planet and take life, not just human life, but take other species with us beyond Earth. And I think the only economically viable way to do that as well, self replication, industrial self replication off the planet. And you also raise another interesting question, would would advance civilizations even bother to talk to us if we saw that we're primitive biologicals when they are much more advanced machine intelligence, And you know, maybe that's a factor too.

00:17:04
Speaker 4: And just to make sure I'm understanding, are we saying that the life in the universe is no longer squishy it's actually machines, or that they're still squishy life somewhere and they're sending machines out to do the exploring and that's what we would be communicating with or both of those options.

00:17:20
Speaker 2: Yeah, both of them. I had in mind the idea that eventually machine intelligence may replace biological intelligence. Out there in the cosmos. People have talked about dice and minds, where you build a dice sphere and use all the energy of a star to support one mind, one gigantic compute. And so maybe there are dice in minds scattered across the cosmos, and they're so far above us that they were not their peers, so they don't bother talking to us. But maybe they're out there, maybe they're watching and they're aware of us.

00:17:53
Speaker 4: I feel like I could personally benefit from a more broader look at self replicating probes, and so, like you know, we've talked about how they can have this exponential growth, but I'm not quite sure what they're growing from or how they're growing, And so can you give me a bigger picture look at what these probes do and what they are and why we want these probes?

00:18:13
Speaker 2: Sure? So the first person I know who talked about this concept was I think as Robert frietas Free to Us, writing in the nineteen eighties. He was associated with a NASA Ames Research Center study in nineteen eighty he talked about where you send a probe to a star. That probe will then mine the gas giants. Maybe it'll set up factories on the moons of the gas giants. But he looked at all the elements that you could get from a belief to be nominal star system, and could you create a complete industry using those resources? And you can, and you could have this factory start small. He called it a seed factor, and the seed would be planted on this icy moon at a giant planet, and it would start to build larger factories and it would all be robotic with autonomous labor and eventually it would start to build other seed spacecraft, and then those seed spacecraft would be launched from there and go to other star systems. And so he tried to do some scaling of the economics of autonomous labor on outer icy moon planets and colonize in the entire galaxy. Since then, we've started to develop some of these technologies in order to support NASA. We've been working on mining the soil on the Moon or Mars, getting resources making metal, and as we've started to do this, it got us excited. We started thinking, wow, you know, maybe this idea from FreeDOS is possible. And so we've done a little bit more recent work trying to bring Freetus's ideas into a more concrete instantiation where we talk about what exact types of robots would be at these factories, what would be their metabolic throughput, so how fast can they self replicate and start to build other spacecraft. So that's the general idea.

00:20:18
Speaker 1: Doesn't it go back a little bit further. Wasn't it von Neuman who introduced this concept of a von Neuman probe and the universal constructor or something which can build itself.

00:20:27
Speaker 2: You're absolutely right. Yeah, I forgot about von Neuman. So he was before freed us.

00:20:32
Speaker 1: He's always before everybody and everything. He's got his fingers and every time he's like Euler.

00:20:36
Speaker 2: You know.

00:20:38
Speaker 1: Right, Let's go through the exercise of thinking about the exponential factor of self replicating probes, because I think a lot of people are like, why can't we just get Elon to do his SpaceX multiplication on our current thing? You know, why is it really necessary to have the probes build more probes?

00:20:57
Speaker 2: Well, it's a matter of scale scaling up if you want to explore the entire galaxy, and maybe you don't, you know, maybe you don't care about that, But if there's.

00:21:07
Speaker 1: A civili I do, if there's a want to overlook something.

00:21:12
Speaker 2: So if there are any civilizations out there that have had the same motive that you have, then it's an economic question. How do you how do you explore ten to the twenty star systems, and if your labor force is only ten to the nine biological creatures you know, in the order of billions, how do you have an industry that can explore on such a vast scale. And so you need to have more autonomy and you need to have a lot of industry with that autonomy in order to build all the assets, all the capital necessary to go out into that gigantic cosmos. So it's a scaling question.

00:21:52
Speaker 1: Yeah. So if I do a simple calculation, you know, if you start out with like five self replicating ships and each one can make fine more than it's only twelve generations before you have a billion ships out there in the galaxy exploring for you, the power of exponential functions is just really amazing.

00:22:11
Speaker 2: That's correct. Yeah, And we do see exponential growth like that. If you put bacteria into sugar water, their population will double, double, double until it uses up all the sugar and then of course you get population collapse at that point. But we do see the exponential scaling occur in some systems. We also see it in technology Moore's law, for example. There's been some discussion why does Moore's law exist. Some people have argued that it's a self fulfilling prophecy that companies try to meet that metric, But for it to persist over so many orders of magnitude, I have to believe it's there's something fundamental that's more than just a self fulfilling prophecy because all of industry has to scale up so that each piece of equipment can meet that exponential growth rate. And so I think that technology does have an inherent exponentiality to it, where technology builds technology and because of that feedback loop, it scales up exponentially. And so extrapolating that idea, you eventually fill up your planet, you end up ruining your planet and you have population collapse, just like the bacteria and the sugar water. So I think it's important to get life outside of the planet so that we don't ruin this for biology, and then we could do greater things as well. Of course, that also raises questions about the ethics of self replicating probes unleashing them in the cosmos, which you hinted at at the beginning of.

00:23:40
Speaker 1: This podcast, letting them tap into that galactic Sure. Yeah, and there's the second element of the industrial aspect, which is not just the exponential growth, but also starting from space. Right, Like, we don't necessarily want to build everything on the service and then have to lift it up out of our gravity. Well, if you can have industry in space, then you never have to overcome that, right. Isn't that a big factor.

00:24:01
Speaker 2: Yeah, that is. And there are ways we can benefit Earth by putting industry in space. They're not always obvious, Like Jeff Bezos talks about moving all of heavy industry off the planet and only keeping light industry on the Earth. But the problem you get into is how do you transport all the mass of manufactured goods down through the atmosphere to the surface, because re entry physics does damage the atmosphere, and you know, ablation of materials puts tiny particles in the atmosphere which contribute to the greenhouse effect and driving chemistry, and the heating of the atmosphere drives chemistry.

00:24:36
Speaker 1: And if I buy dog chew toys on Amazon, I don't want them melted from re entry in the atmosphere, even if they were manufactured on the Moon.

00:24:43
Speaker 2: Right, yeah, right, But despite these problems, there are ways we can move industry, at least parts of industry into space to do a great benefit to our planet. I think by the end of the century we could have fifty percent of our industrial footprint in space.

00:24:57
Speaker 3: That's a lot, all right.

00:24:59
Speaker 1: So I'm fascinated the technical questions you raised about whether we could actually put this thing together and make it happen, build a factory that can make factories to make factories. But let's take a break and come back and then dive into those technical details. All right, we're back, and we're talking to Phil Metzger about building self replicating probes that go out and explore the galaxy and maybe get the attention of those crazy dice in minds so we can learn what they know about the universe.

00:25:44
Speaker 3: That would be pretty awesome.

00:25:46
Speaker 1: So this sounds like a pretty daunting task to build a machine that could build machines to make more machines. Let's talk about the first piece of it. How you get the materials, how you mine it? Because if our machine is like landing on some alien moon or orbiting some gas giant, it's got to find the bits to make more of itself, right, so it needs whatever it's made out of. It's got to find all of those bits locally. How does that work? How do you build a machine which is capable of like mining pieces for itself. Can we build autonomous mining devices?

00:26:18
Speaker 2: Yeah, there's no new physics required, but the technologies are very immature. Some of them are only conceptual as we have conceived of these. They are all very doable. It's just going to take some time and some industrial engineering to develop them.

00:26:33
Speaker 1: I love your optimism.

00:26:34
Speaker 2: Fil Yeah, it's just going to take a few trillion dollars and you know exactly whatever.

00:26:43
Speaker 1: It's just we know how to do it. It's just an engineering problem. We just got to get it done right.

00:26:48
Speaker 2: Yeah. So here on the Earth we do it using human intelligence, human labor, and we've scoured this planet for thousands of years looking for all the best resources, and we've discovered there are special metals. We've discovered certain types of rock, certain type of ore that we can extract these metals out of. And so we don't just grab any material off the ground and start trying to build robots out of it. We have this gigantic logistical network on the Earth, transportation hubs and giant container ships, and we have mining is distributed all over the planet bringing together the materials we need to build this industry. So if you wanted to set up an industry on the Moon, the first problem we have is that we don't have that logistical We don't have thousands of years of developing that logistics, nor do we have all the deep understanding of where the resources are in the Moon. The second problem we have is that the Moon lacks a lot of the geological processes that the Earth has had. So, going back to what we said at the start of this conversation, Galileo argued that it was a planet because it has the same geological processes. But now we know that it doesn't have all the same geological processes and Earth is pretty special. So if we want to build industry on the Moon or other simpler objects, we're going to have to develop tech to extract the resources out of minerals that we would normally pass over. It can be done, but it's not as efficient. It takes a lot more energy, and the chemical processes to do that haven't been developed yet. So people have conceived of how to use sodium hydroxide to break down rock to get all the different atoms out of the rock, or how to use fluorine to do that processing, but we've never had anybody get funding to go build a fluorine metal extraction device, which would be very dangerous. Working with fluorine is hazardous, and so it's going to take a lot of money and there's not really a market for it. Nobody wants to go build it because you're not going to make any money off doing it. So this is the problem we get into that the technologies that we need to live and operate beyond Earth are pre economic. We think that eventually they will have a very important role in our civilization, but not yet.

00:29:11
Speaker 4: We had a question from a listener where they said they really wanted to hear about how engineering on the Moon would differ from engineering here on Earth. Tell me if this is too far afield, but could we talk a little bit about how the Moon environment differs from the Earth environment in ways that would make engineering interesting but also perhaps more complicated.

00:29:29
Speaker 2: Yeah. So it's extremely challenging to try to build hardware to operate outside of Planet Earth because the environments are so radically different. On the Moon, you're dealing with a temperature swing of I forget the number, but it's like four hundred degrees difference between day and night. We're dealing with hard vacuum materials like plastics will outgas and lose their flexibility, and therefore washers and O rings will start to fail. We're dealing with this tremendously abrasive dust, which comprises between twenty and up to fifty percent of the mass of the soil in some locations, because the Moon lacks a water cycle to wash the dust out of the soil and to turn it into mud and then mudstone, and so the dust just builds up over geological time scales, and working in that extremely abrasive dust is maybe the biggest challenge. That You've also got low gravity, and then you've got the radiation environment. We don't have Earth's atmosphere to shield us from these high energy particles coming down from space. It doesn't have a magnetic field to also deflect particles away. The ultraviolet light ruins materials, the space plasma effects. So we can go on and on listing the challenges of working in space, and we don't even understand all the physics of some of that. We don't understand the space plasma environment and how it interacts with the lunar surface. So it's a really interesting field to be in. I've always worked in groups that typically have a ratio of one third physicists two thirds engineers, and it's a really cool working environment because the scientists are trying to understand the basic physics, and then the engineers are taking that knowledge and creating the technology and then we need the technology to go learn the physics, so it's a feedback. They're both supporting each other, which makes it a really interesting field. It's also a really hard field to work in because you can't do the tests that you want to do on your hardware. You just cannot replicate the lunar environment or even the Martian environment well enough. Here on Earth, even in the Giant Chamber. You can't get the gravity right and you know, etc. So we have to rely on simulations. But we can't write computer simulations that are good enough because we don't understand the physics yet. So we really have to get data from those objects. We got to do more mission through the Moon, more missions to Mars to learn the science.

00:32:03
Speaker 4: But it also sounds fun, like a fun challenge to have all of those pieces.

00:32:07
Speaker 3: I like a good challenge.

00:32:08
Speaker 2: Oh, it's tremendous fun. It is tremendous fun. And when I speak to students, undergraduates or high school students, I'll show them pictures of the amazing things that humans have done already, like these fabulous skyscrapers or these unbelievable bridges, and you know, when I drew Gopher Bridge, I'll look at it and like, I think, how did we get all this mass up here in the sky before there was a bridge? You know? And these are really daunting problems, but we've managed to solve them by doing straightforward engineering. Break it down into smaller problems, get the funding, do the engineering. But it needs to be done for space still. We need to have young people working on these problems. And there's so much work to be done, so much discovery still ahead of us, that I think is a great for young people to be getting into these fields.

00:33:03
Speaker 1: I have that same feeling when I see like the Golden gate Bridge, like look upon my works, E mighty right, it is awesome, And I like how you describe the scope of this challenge. I mean, here on Earth we're not capable of building robots that can do very much yet certainly not capable of building robots that can make more robots. And it's supported by this incredibly vast mining industry, which requires a lot of human work, you know, many cases like terrible labor conditions. Right, so we're so far from being able to do this. Give us a little bit of that fill optimism. What are we capable of doing or what do you think is the first thing. I mean, you've studied like actual lunar industry, you know, processing regolith etc. What do you think is going to be the first thing we accomplish down the road towards being able to do this well.

00:33:49
Speaker 2: We we're currently seeing a lot of progress in robotics and in automation. And there's one company, for example, that has robots. They typically will post on social media pictures of their robots folding the laundry and able to pick up these cloth pieces and fold them very carefully. So the dexterity and the machine vision, the autonomy to be able to do task like that is making tremendous progress. And again it comes down to an economic question. Is there a consumer need for these technologies because there's not a lot of funding going into them unless they can make a profit. People aren't going to put their retirement money into something unless it's going to help them retire. So we're seeing a lot of advancement. I think the big killer app is going to end up being AI. I truly believe AI servers are going to have to go to space because the environmental costs are greatly increasing. The pushback to building servers is growing for good reason, And already servers could be profitable if they went to space, just not as profitable if they build them on the ground. But I think that the tipping point's going to come where they start going into space. There are already people like Eric Schmidt and Sam Altman in the AI world talking about how there is inevitable we're going to build servers in space.

00:35:10
Speaker 1: Do you mean service in space to support space industry or do you mean service in space to support like people who want help organizing their day through JATGBT on Earth.

00:35:20
Speaker 2: Yeah, I think that that all the AI servers that are supporting people on the Earth are eventually going to be in orbit around the Earth, maybe distant orbit, because the latency doesn't matter that much for most compute and so Eric Schmidt and Sam Altman that's what they're talking about. They're talking about putting the AI servers that we would have built on the Earth putting them in space instead because of the environmental impact costs of excessively building data servers on the Earth.

00:35:48
Speaker 1: But how do you balance that against the issues of like cooling, right, because in space you have to cool everything radiatively, and you know, technical support, how do you go reboot those servers if they're in distant orbit, is that really going to be economically feasible.

00:36:02
Speaker 3: Harden them against radiation?

00:36:03
Speaker 2: Yeah? Yeah, those are good questions. And as far as the radiative cooling, the energy in has to equal the energy out, and that's the same for every spacecraft. So the scale of solar panels and radiators, that ratio will be the same on servers that it is on any other spacecraft. It's just a matter of scaling it up to an unbelievably gigantic scale, which is, you know, super ambitious, but nonetheless that's where people are talking about going. And as far as being able to have radiation hardening, yes, that's going to require additional mass around the servers to harden them against radiation. And as far as being able to repair it, yes, that's going to require better robotics and more autonomy. But this is I think is going to be an economic driver that will push those technologies forward because I don't believe there's ever going to be an upper limit of demand on intelligence. I think the intelligence will become the customer for more intelligence, and it'll create that feedback loop which will have exponential growth which would destroy our planet if we don't push it off into space.

00:37:14
Speaker 1: Well, I want to invest in Fill's Space Optimism company at this point.

00:37:20
Speaker 4: All right, so we've talked a little bit about the resources that you would need to find in space and then extract and why that might be difficult. So say you have those resources, what's the next step?

00:37:32
Speaker 2: Yes, so the next step is well, beneficiation. That's where you improve the quality of your resources before actually doing the extraction process. I have some patents in beneficiation. My university owns them, but they were inventions that I had, and I like to tell people. If you go to the patent search and look on patents on concrete, there's literally over a million patents on concrete. Look at concrete for off the planet, there's only two and I have one of those too, And so there's still room for nine and ninety nine thousand more patents on concrete. And that's why it's such a great field to go into. We've only just begun developing these technologies.

00:38:20
Speaker 1: Would you call that exo concrete or astro concrete or something.

00:38:24
Speaker 2: Yeah, I don't. We didn't come up with a name for it, but but our idea was that if you're going to be making concrete by absorbing microwaves, microwaving the lunar soil until melts. Some minerals are better at absorbing microwaves than others, and so using magnetic fields, we can sort the minerals out and improve the microwave absorption by something like seventy percent, which results in a dramatic reduction in the energy and a much greater efficiency. So that's an example of beneficiation. Let me just mention the reason we need beneficiation is because we're not going to be able to go all over the moon and find these native ore bodies of each mineral. Instead, we're going to be scooping up the dirt off the ground, which is a mixture of minerals, and so sorting the grains is an early step. After that, then you have chemical processing, and there has been some work on this. One of the processes is called molten regular electrolysis. That's where you melt the soil. They have an anode and a cathode, and you run an electrical current through the molten well basically lava, and that electric field breaks down some of the minerals so that the oxygen is released, and then the metals will sink to the bottom and you get two melted materials. One is the oxides on the top, which you can use to make ceramic, and the other one are the pure metals. Now we call that a mongrel alloy because it's going to be a mixture of iron, magnesium, aluminum, calcium, and even some silicon, and so it's an iron silicon mongrel alloy. It's very heavy. It's weaker than steel, but it's pretty good. You know, it's stronger than iron, and so there's a very rudimentary building material. But if you want to do better than that, now you need to have metallurgy. You need to further refine the metals to separate them from each other using the standard processes we here on the Earth, but adapted for lower gravity. And so it'll just be hardcore industrial engineering doing electrochemical processes to break down the atoms and then separating the different material streams into making feedstock, and then it's just standard industry. After that, it's casting forging parts. You could do three D printing, although the throughput may not be as high on three D printing. Three D printing is very automatable, so making parts. Then you have to have a set robots they can put all the parts together. Typically we envision these being humanoid robots so that they have they have flexibility analogous to a human, but they needn't be humanoid. They could be any kind of robots that can build things. Now, one of the challenges we get into is that here on Earth, our industrial supply chain includes something like twenty thousand different types of screws, and you don't want to have enough machines on the Moon to make twenty thousand types of screws if you can get by with three types of screws, and so we need to do a lot of industrial ecology to figure out how to create a self replicating or a closed ecosystem of machines using fewer parts and fewer machines. So there's a gigantic field of work that hasn't even started yet for industrial engineers, architects, computer programmers, mathematicians. The math on writing an industrial economy is really complex and fascinating math and has really fabulous theoretical approaches, but it hasn't been applied far enough yet to look at doing this on the Moon. So there's a lot of work ahead.

00:42:14
Speaker 4: Still, would we need to get to the point where we're making like computer chips for our humanoid robots in space? Is that like how far we need to get before they can replicate?

00:42:23
Speaker 2: Not at first, but I think eventually you will need that. In the modeling that I've done, we assumed that you would start making simple things like metal, and then you would go through a series of generations of hardware to and it's all a material science question. It's what material can we make next, and then what material after that? And the goal is to make an increasing fraction of the parts for your industry. And during that interim time you're continuing to bring things from the Earth, and then the assembly robots are putting your parts together with the ones that were brought from Earth, and over time you wean yourself off of the earth made parts. Now, the very last thing that we assumed is that you're making computer chips or I've done some modeling for Mars where humans would be on Mars, and so in that model, the very last thing you would make would be the pharmaceuticals. And it's a question of the mass of product divided by the mass of capital. You want that ratio to be sequenced. You want to make the industries that have the highest ratio first, and then work your way down through all the sectors of the economy and do the ones that produce the least mass last, which would be pharmaceuticals and computer chips.

00:43:39
Speaker 1: So the things that are the least mass you want to do last, because so they're cheapest to bring from Earth because they're low mass.

00:43:45
Speaker 2: Exactly nice yep, but yet very expensive to stand up those industries on the new planetary body.

00:43:51
Speaker 1: I mean even here on Earth, Like to make computer chips is like one company that can make them, and they rely on several single source manufacturers of like devices and lenses and stuff like that. So we're talking about replicating that entire supply chain in a robot that can replicate that entire supply chain, It just it seems sort of fantastical.

00:44:11
Speaker 2: Yeah, So I think we should get away from the idea of self replicating robots and talk about self replicating industry or self replicating factories because it will be a whole family of robots at these factories on this icy moon around a Jupiter like object. Yeah, so trying to do this in one robot, I just don't see that happening. I mean, we have self replicating biology. We call it self replicating, Like raccoons can make other raccoons, but they're not independent. They are part of a biosphere and they depend on other species.

00:44:50
Speaker 1: Are you suggesting we send raccoons to space to explore the galaxy eventually?

00:44:55
Speaker 2: Yeah, I would love to see that.

00:44:56
Speaker 1: I think I've seen that movie.

00:44:58
Speaker 2: Yeah, well, I mean Guardians of the Galaxy. Yeah, But even self replicating biology is not really standalone. Maybe some simple bacteria can go live off of rocks and self replicate, but if you want to produce anything economically useful for civilization, then I think it has to be EcoSpheres of robots and EcoSpheres of factories.

00:45:21
Speaker 1: But you still have to have the initial thing which lands for the first time on that planet and begins replicating. I think I'm getting that you're saying, it's not just a robot which makes other robots that look like it's going to make like a foundry, and it's going to make like helper robots that it's going to make the whole industry. But you still need the thing which lands and starts everything off, like the seed. As you were saying, earlier.

00:45:43
Speaker 2: Right. One way you could think about it is you're going to have to relive the entire industrial revolution that we went through on Earth on this new planet. And so you're going to have a box, and that box is going to contain computers that know the whole process. You know, it knows where it's going. But in the first generation, it's not going to make everything. It's going to have some supplies. Even a seed, you know, a seed has food in the seed so that it can live off of what it has stored until it can make its own food. And so you're going to need some supplies in that box to live off of until it creates the ability to you know, to replicate everything.

00:46:26
Speaker 1: So we're going to have like a coal powered steampunk era on every planet we land on.

00:46:32
Speaker 2: Yeah, I used to say something like that, and somebody once pointed out that that's probably not how it's going to happen, because maybe we'll invent nanotech, and maybe nanotech can be smaller scale and support the self replication as a smoother process. But you know, that's something we're just speculating about at this point.

00:46:53
Speaker 1: Well, do you know anything about this argument between Eric Drexler, who wrote Engines of Creation, who is a big proponent of nanotech. And Richard Smalley, the guy who won the Nobel Prize for buckminster Fullerene, who argues essentially that you can't have nanotech self replication because the pieces need to be nanotech and they can't they just like force the chemistry together. Have you followed that conversation, No, I have not. Well. Smally essentially says you don't make a girl and a boy fall in love by pushing them together. He's essentially saying that, you know, you can't just manage chemistry by squishing things together at a nanotech level. They had some like two years of open letters where they were arguing with each other about whether this is ever going to be possible.

00:47:36
Speaker 4: But what you're not just smooshing them together, right like? You know, you know, we know you can smoosh things together and expect certain chemical reactions depending on what you're smoohing together.

00:47:45
Speaker 3: I'm going to weigh in an argument I don't know anything about.

00:47:47
Speaker 4: But it seems like like probably people had more complicated opinions than just we're going to smoosh things together, right.

00:47:53
Speaker 1: Well, I think this series of letters which people should go out and check out is not an example of good faith arguing as we often see online.

00:48:02
Speaker 3: So, yeah, got it?

00:48:04
Speaker 4: All right, Well, let's take a break and when we get back we'll talk about energy sources and autonomy.

00:48:29
Speaker 3: All right, we're back.

00:48:30
Speaker 4: So we're talking about self replicating robots and all of the steps that would go into making them, and I feel like maybe we want to have an energy discussion at two different scales here. Phil was talking about like smelting and things that would require very high temperatures, and so I'm wondering what would be the source of power for that? And then I also, on a tinier scale, want to know what would be our source of power for the robots that are building everything.

00:48:55
Speaker 2: Yeah. So in the paper I wrote on this topic, we assumed everything would be solar powered, and so the question came down to can a self replicating set of robots create solar panels that will create enough energy for that process? In other words, does the metabolism close? And there's a lot of hand waving in that paper. In fact, I didn't expect the paper to get as much attention as it did. In the opening of the paper, we said, this is a preliminary study, which is just designed to get more people interested so that then later we can do a proper study. But everybody got real interested and it took off, and we never did the proper study. There's never been funding for it. So in our hand waving arguments, we used the evidence we had available on how much energy will it take to make metal? And you know, a lot of handwaving. And then for a safety factor, I said, well, let's assume that in every generation we create thirty times more solar panels than we think we're going to need in the next generation. So I had a factor of thirty uncertainty. And even with a factor of thirty uncertainty, the metabolism closed. So solar alone should be enough to do it. But you will be making a lot of solar panels. And we know that you can make solar panels out of lunar soil. They're already there are already two companies doing it. Blue Origin has a technology for making solar panels out of lunar soil. And there's another company called Mana Electric MAA and A. They're in Europe and they also have technology to do this, and they claim they can make solar panels using something like ninety nine point eight percent lunar soil and only zero point two percent brought from Earth.

00:50:45
Speaker 3: Wow, Okay, so how are they doing that?

00:50:47
Speaker 4: So are they are they doing it from like they're extracting the resources, they're manufacturing them in space. They're doing all of that stuff just using like equipment that they shift from Earth.

00:50:57
Speaker 2: So I haven't seen the details from either of the to companies, but the press announcements tell us that they have made solar panels out of simulated lunar soil. And the way you do it is use a process like molten regular electrolysis, or fluorine or sodium hydroxide, or you know, some method to break apart the molecular bonds in these minerals. So we're dealing with minerals like basalt and ilminite, and a NORTHO site. You know types of rock and mineral that are in the lunar soil, and we know the composition of these minerals, so we know there's iron and aluminum, and we know there's calcium. Typically you're gonna have a hard time finding hydrogen on the Moon unless you go to the polar regions. But I don't know if these companies process requires hydrogen. Carbon is another one that's hard to get On the Moon. We know there's some carbon in the ice of the poles of the Moon, but not much and it's only the poles. But maybe you don't need carbon in these processes. So anyways, apparently they're doing the chemical reactions, they're producing these materials, and they're laying them down in a wafer so that you have p N junctions so that they are electronic devices and are photosensitive so that they can convert photons of energy into voltage. And they claim that they've made it work.

00:52:25
Speaker 1: But why try to do solar power? I mean, if you're landing on some random surface, you don't know how far away that planet is from its star, how bright that star is. Isn't nuclear power something that's going to be more robust. We already know how to build those things fairly miniaturized for stuff here on Earth. Why not nuclear powered probes?

00:52:44
Speaker 2: Yeah, so for the actual self replicating probes, I think Free just did talk about nuclear and that's part of the reason why he wanted it at a gas giant planet, so that you would have a lot of hydrogen and you would have helium so that you can do fusion for example. I don't remember if he was using fusion or fission in his analysis, but yeah, that is the goal to eventually have nuclear power so that you're not bound to being too close to a star. It's a crawl walk run type situation. So the technologies we're actually developing right now are ones that we think will be useful for NASA and useful for commercial companies in the near term. And it'll be a while before you can get your whole supply chain up to making nuclear reactors.

00:53:31
Speaker 4: All right, So let's imagine we've got we've got the power figured out, we make we're replicating these robots we're scaling up. Let's start talking about like the ethics and some of the other bigger problems we might encounter. So, first of all, how do you make sure you don't get like bad copies? This is like humanities ambassadors that we're sending out into the solar system. How do we make sure that they remain good ambassadors.

00:53:58
Speaker 2: Yeah, that's a big problem. So I'm not actually working on that problem because it's still pretty far down the road, but it is something that we need to consider. We need to have ethesis and philosophers thinking about these things, and they're surprisingly there are people working on these problems. One of the reasons that we're thinking about it is because we're trying to detect is there already life in the cosmos outside of Earth? And we're asking the question, why don't we see radio signals coming from all the other stars? You know, why is it not a star wars galaxy? So this is the question of the Fermi paradox or the great silence? Why is it so silent out there? And there's a number of theories. One of the theories is the dark forest hypothesis, where in game theory you have to consider the possibility that there are bad actors that if they discover your presence, they're going to come and wipe you out because they know that you might develop self replicating probes, and the probes you develop could take over the galaxy and wipe them out. And so in the game theory it becomes a part of the puzzle like how does self replicating probes fit into the dark forest hypothesis? Also, if self replicating probes are possible, why are they not already here? Because we think we can get there in you know, a few hundred years or less. I honestly think that we could get there by the end of the century or maybe within one hundred years. And so I think there are people alive today that can see this happening.

00:55:31
Speaker 1: That's my biggest question, right, Like, if this really is possible, if we're close to it, then surely aliens have been close to it. And if it doesn't take more than fifty thousand or on hundred thousand years to explore the whole galaxy with these probes, then why haven't we been visited? So what's your personal answer to that? Film.

00:55:47
Speaker 2: Yeah, so there's I think there's three or four really interesting hypotheses. One is the dark forest one. Another one is that life is just incredibly improbable and so Earth might be alone within the visible universe.

00:56:02
Speaker 1: And just to underscore that this is such a powerful technology that it would allow any alien civilization in the entire galaxy to visit us in a fairly small amount of time. And we're talking one hundred thousand years or so, So you're suggesting that the lack of visiting self replicating probe suggests that we might be alone in the milky and not just like rare, but like literally alone.

00:56:23
Speaker 2: Yeah, in terms of advanced intelligent life technological species. In fact, it's worse than that. There was a paper done by a philosopher at Oxford a few years ago. His name is slipping my mind. It might have been Stuart Anderson, where he showed that one civilization in another galaxy could set up a linear accelerator and dismantle one planet the size of Mercury, turning all that mass into self replicating probes. And if they did that a billion years ago, then ever, every single galaxy in the entire visible universe would already have every single star colonized. And so it's not just the galaxy, it's all the galaxies that are involved in this question. Okay, So yeah, it's a great and very important question. There's also the theory that civilizations always go extinct and they don't get that far. I don't think that's very plausible anymore, because we're already close to that point. You know. That's the great filter hypothesis. The other hypothesis is that it's the transcendence hypothesis, and I like this one a lot. And the idea is that civilizations go so intelligent that they actually figure out that self replicating probes are dangerous, so then they don't set them loose, and they don't really need the material of the rest of the galaxy, and they're more interested in just watching and seeing how other star systems develop rather than colonizing. And so they're not our peers. They're far above us. If that's possible, maybe that window of danger where you unleash self replicating probes is a very narrow window, and maybe they're watching out for that, you know, I mean, if we're going to get contacted by aliens, I think we're close to the point where it would happen, because we're just about to transition to having superintelligence and self replicating probes, and we're just about to become a danger to our part of the cosmos.

00:58:30
Speaker 3: And you're trying to hasten that.

00:58:34
Speaker 7: We all know.

00:58:35
Speaker 4: I'm a what blanked, right, And so you said this is your favorite hypothesis. That feels to me like you wouldn't want to be working as self replicating probes. Then what am I missing?

00:58:45
Speaker 1: He wants to get us to the place where we're responsible with our self replicating probes. Is that the idea?

00:58:50
Speaker 2: Yeah, one of the problems I have in life is I always try to take a very nuanced approach to everything, and it's really hard to describe a nuanced position. And so the nuance in this one is that I think we need to have industry outside of planet Earth in order to save the Earth. But there is a danger. In fact, there's multiple dangers. So as we're going forward towards this right future, we're going to have to solve major ethical problems along the way, and so preventing runaway destroyer probes from setting out from our planet destroying our planet and then all the other ones, you know, that's one of the one of the big concerns. There's other concerns even before then, like if you've got self replicating industry in space, whoever owns that industry is not going to need to dilute their equity. They won't need any labor, they won't need anybody's property on planet Earth. They can just go out there and replicate, and within twenty years they can have more industry than our entire planet. And therefore, even if the whole planet pooled all of our resource together, we would not be able to buy a signif figant share of that industry. Even if the owner wanted to sell, And so there's the potential for more wealth concentration once we've removed labor from the equation, and once we've removed the planetary scale limits from the resource equation, And so that major sociological and ethical problem has to be solved, and we only have about forty years to solve it in my opinion, So yeah, there are major issues we got to solve. But still, nonetheless, I think that we're not going to be able to slow down industrial growth on our planet. We're not going to be able to slow down demand for intelligence because it's geopolitical, and if we want to save our planet, we're going to have to start developing these technologies off of the planet.

01:00:42
Speaker 1: So then the phil's optimistic view of the future is that over the next few decades or centuries, we develop these off planet resources as a way to salvage Earth and stop putting such a great environmental burden on it and to expand out to the rest of the Solar System. But that once we develop self replicating industry, we are wise about it and we don't release it out into the universe to run like a crazy virus and take over the rest of the universe and we're still here because aliens have also been responsible with their technology.

01:01:14
Speaker 2: Well, either they don't exist, or yes, they became smart along the way they underwent this crisis. This point of crisis where your technology becomes truly dangerous, you know you have AI that exceeds the sum of human capability, and you know that at that level of danger, at that crisis time, either you come through it or you don't. I think inevitably we're going to get there. Honestly, I'm not. I don't feel like I need to push for industry to happen off the planet. I think it's going to happen no matter what. And so what I'm trying to push for is to democratize it so that people all over the world are involved in the process and owners developing equity as we go to try to improve the odds that we will solve those societal problems.

01:02:05
Speaker 4: Well, this conversation has given me a lot of things to be optimistic about and a lot of things to panic about.

01:02:09
Speaker 3: Tonight when I'm trying to fall asleep.

01:02:12
Speaker 1: Yeah, I just hope that if there are aliens listening to this podcast, that they take Phil's comments to Heart and that they are wise and responsible with the use of their self replicating technology.

01:02:22
Speaker 3: Thanks so much for being on the show.

01:02:23
Speaker 2: Phil, my pleasure, Thanks for having me.

01:02:32
Speaker 4: Daniel and Kelly's Extraordinary Universe is produced by iHeartRadio.

01:02:36
Speaker 3: We would love to hear from you, We really would.

01:02:39
Speaker 1: We want to know what questions you have about this Extraordinary Universe.

01:02:43
Speaker 4: We want to know your thoughts on recent shows, suggestions for future shows.

01:02:47
Speaker 3: If you contact us, we will get back to you.

01:02:50
Speaker 1: We really mean it. We answer every message. Email us at Questions at Danielankelly.

01:02:56
Speaker 4: Dot org, or you can find us on social media. We have accounts on at Instagram, Blue Sky and on all of those platforms.

01:03:03
Speaker 3: You can find us at D and K Universe.

01:03:06
Speaker 1: Don't be shy write to us.