AGI, Immortality, & Visions of the Future with Adam Becker

0:00Love me some future talk, but they spook me though with where they take the end of civilization. But we don't know whose hands are on the steering wheel. We don't know who's shaping this future, and that's why there's concern. Well, I know who's shaping it, and I'm scared to death. It's you. Let's watch Chuck be scared to death as we discuss all the ways tech will be shaping our future. Coming right up. StarTalk, special edition. Welcome to StarTalk. Your place in the universe where science and pop culture collide. StarTalk begins right now. This is StarTalk Special Edition. Neil deGrasse Tyson, your personal astrophysicist.

0:47And when I say special edition, it means I turn to my right. And Gary O 'Reilly is sitting there. I know. Hi. Where'd you get your British accent? Stole it. just a thief chuck hey you're otherwise known as lord nice but we can call you lord of comedy lord of comedy can we do that okay let's do that so today we're going to explore a vision of the future and everybody's got their take on the future yes everybody's got everybody and but they all have different takes because they come in from a different place so you got to hear it all if you're going to assimilate it into something that you're going to take action on.

1:26True, yeah. So either make something happen or prevent something else from happening. Yeah. So set us up, Gary. All right, so what does the future hold for us? That'll include scientists, science fiction authors, tech CEOs, and the so-called futurists. Everyone has their own idea for the future technologies. Vision of AGI, nuclear fusion, the singularity, transhumanism, living on Mars. We've got to get to the moon first. No, you don't. You go straight to Mars. Do we? Don't get me started. Get your ass to Mars. And there you have it. Stuff we talk about all the time on StarTil. And in the face of new technological developments, we're quickly going from science fiction to science reality.

2:11But are we headed towards the utopian or are we headed towards dystopia? That we'll get into. are these technologies as close as they claim? Is science fiction always a guiding light or can it be a blueprint for those in power? So on that note - So who do we have today? Adam Becker. Adam, welcome to StarTalk. Thanks for having me. It's good to be here. All right. You got a PhD in computational cosmology. Yes. Love it. That was back in 2012. Yep. And you wrote a book in 2018 called What is Real? Yep. That's audacious. You know, the unfinished quest for the meaning of quantum physics. Sound like there's a little bit of philosophy in there.

2:57Yeah, yeah. Or a lot of philosophy. Oh, there's some. Yeah. And you just came out with a new book because you've been a writing science popularizer maniac ever since your PhD. Pretty much. Yeah, yeah. So here's the title. I love this. More Everything Forever. AI overlords, space empires, and Silicon Valley's crusade to control the fate of humanity. Oh, I got a better title. We're effed. You know, you should have had that title. Yeah, you know, we considered it. We just didn't think that it would really, you know, sell. So my researchers told me that we've corresponded before. Because I've only just met you now.

3:43Yes, that's true. So what? So they told you we corresponded, but they didn't tell you. Wow. Okay. So what happened? So they set me up. That's what happened. That's so unlike us. What? Oh, God. Well, what happened was. What happened was. Exactly. I was a snot-nosed kid in grad school and came to visit the museum and noticed what I thought was a mistake on one of the plaques. and so I emailed just the general astronomy department email here at the Rose Center. And then two weeks later, you wrote back. Oh. And what did he say? Well, he - I'm sure I would have been polite. Was there a mistake? And did Neil school you if there wasn't?

4:36Whether there was a mistake or not was a matter of some debate. Oh, really? Yeah, so the question was the size of the universe. Oh, okay. There it is. Yeah. Well, that's still a debate today. Well, no, that's not in the way that he's described. No, not like this. Yeah, yeah, yeah. So I deal in observables, okay? Right. As do many practical scientists. Right. So when you speak of what the universe is doing, you speak of what you see it's doing. Right. And we can see galaxies whose light has been traveling for 14 billion years, 13.8. Right, right. And so we will loosely say, well, it's definitely the age of the universe.

5:16But we speak of the size. We can be a little sloppy and say it's 13.8 billion light years to the edge of the universe. Right. But that's not strictly accurate. What you have to do is, since then, the universe has been expanding. Yeah. So where's that galaxy now? it's like 45 billion light years away but you can't see it so you have to stick it into a model of the expansion rate of the universe and come out with a number that you cannot observe so you're being snot nosed about that but that's fine tell me I was polite because I think I'm polite you're polite you're polite we had a little back and forth and eventually eventually I think you were getting a little impatient oh really and you said why don't you you know make a presentation of this to a wide audience in a way that you think is suitable.

6:05I remember that correspondence now. Okay, okay, I get it. So you said that. Okay. And then I went up and did a podcast about it and sent it to you. Okay. All right, so in your book, Overlords, Space Empires, you just go all out. Yeah. And you're coming to it from a physicist with a philosophy flavor. Yeah. So you're going to see this in ways pure tech people wouldn't. Yeah. Or politicians or just regular everyday folk walking up and down the street. So how did you prepare for this book? I read a lot of really bad writing by, you know, tech CEOs and people defending tech CEOs online, you know, writing long essays and books about why the future is inevitably going to be all about super intelligent AI.

6:54Why the future is going to inevitably be about space organization. These are tech people writing all this. Yeah, exactly. Some of these things, you know, were ideas that I had the expertise to say, okay, no, that's not true and here's why. But some of them were ideas about, like, you know, biology or areas of physics that I don't have expertise in or, you know, other things. And so then I went and interviewed a bunch of people who have expertise in those areas. Now you're playing journalist in that capacity. Yeah, yeah, yeah, yeah. And, like, read books on those subjects and, you know, pulled out what I needed and stuff like that.

7:25And then I tried to interview, you know, the tech CEOs themselves. And almost all of them said no. No, right. Yeah, because they're looking at you as somebody who is intellectually honest with some integrity. And they're like, we can't talk to you. Okay, because they know they're full of crap. Yeah, well, I think that they just didn't see any reason to, right? You know, like I was very honest. I said, you know, like this is a book that's going to take a critical look at your ideas. That was your problem. Well, you can't do anything else. If you had gone in and said, I'm enamored of the fact that AI is going to be such an integral part of the next chapter in human history and that you guys are the progenitors of this amazing tech.

8:11They would have been like, come on in, let's talk for a second. No, you don't meet an appointment. Stop by any time. That was your first mistake. Yeah, man. Yeah. Well, but you know, I got journalistic integrity. All right. Well, good. Well, look, let's explore some of the scenarios that are going to be potentially the reality of us as a human race. Just go right on down the list. Yeah, the laundry list. Okay, well, let's look at Mars in 2050. Oh, yeah. Are we saying maybe, maybe not? You're kidding me? Oh, that's definitely going to happen. Yeah, Elon Musk. He has said he wants to put a million people on Mars by 2050 to have a self-sustaining civilization that will survive there even if the rockets from Earth stop coming because there's been an asteroid strike or nuclear war or something here.

8:56That's definitely not happening. There are a lot of reasons why that's not happening. Getting anyone to Mars by 2050 and bringing them back alive or just having them live there for a while, that would be incredibly difficult. The challenges just to put boots on Mars, the way that we did on the moon, are enormous, right? Just learning how to keep someone alive in deep space that far away from Earth for as long as it takes to get to Mars stay on Mars, come back. We do not know how to do that yet. Chuck, that's the problem. They wanna put boots on Mars instead of sneakers on Mars. If you get a sneaker contract, you'll pay the whole way.

9:32Nike would've been there by now. They just do it. Yeah, yeah, yeah. Exactly. Excellent. They just do it. Well played. It ain't about boots, it's about sneakers. I mean, so the - What are the biggest challenges of going that far into space? Is it radiation or - Yeah, there's radiation. And that's not just when you're in space, that's also when you're on Mars, right? You know, the two things that primarily protect us from radiation here on Earth are our, you know, the Earth's magnetic field and the thick atmosphere that Earth has. Mars doesn't have either of those things. So when you're on the surface of Mars, you're getting pretty much the same radiation dose that you do like out in space.

10:09And that's not good, right? You know, like the thing that I tell people is the movie The Martian is science fiction. One of the things that's science fiction about it is if Mark Watney really, you know, had to do all the stuff that he did in that movie, he'd come home and he'd be dead of cancer in a couple of years because he had too much radiation exposure hanging out in Mark.

10:38I'm Brian Futterman, and I support StarTalk on Patreon. This is StarTalk with Neil deGrasse Tyson.

10:51What about the ISS? If Scott Kelly could stay up there for a year One of the twins One of the twins One stand on Earth And one alone Right Why couldn't you just extend that For whatever time necessary To go to Mars? Even if it's not to live there If it's just to go there And dig a hole And come back Right So there's a couple of things First of all On the ISS They're still in the Earth's magnetic field They still have a bunch of the shielding. Oh, wait, and what's that called, Neil? Wait, the field that goes all the way out like that? Oh, yeah. Oh, it's called the... Magnetic field. No, it's not the magnetic.

11:23It's the magnetosphere. Yeah, think of X-Men. Yes, the magnetosphere. Go ahead. Yeah, exactly. It is like the X-Men. Yeah, they've still got that protection. Also, if something goes wrong on the ISS, they'll be back on the surface of the Earth in a matter of hours. Like, they can just abort and come back home, right? At most. I mean, you come out, you're down within a half hour. Exactly. The hours is you want to line up so you don't land in the middle of sharks. Right, exactly. Yeah, yeah, yeah. So you can get out easy. Easy. And you can also have a real-time conversation with people on the ground because they're not that high up.

12:00And so the speed of light delay with the conversation doesn't matter. on Mars, you know, it's a minimum of something like, I think, eight minutes each way and a maximum of something like 15 or 20 one way. And so if you send out a message, you are waiting at least 15, 20 minutes to get a message back. Maybe more like 45. Better be a good message. Yeah. It better not be like, so how's it going? Over. Yeah. Yeah. Put some content in there. Or watch out for the cliff. Yeah. Exactly. Yeah. And the other thing is like, if you have a problem on the surface, to Mars and you want to come back, that's going to take you at least nine months, maybe more if you happen to be near a launch window where, you know, the Earth and Mars are like in the right positions.

12:46If you're not near a launch window, it could be well over a year before you can come home. And a full up round trip mission to Mars with ideal launch and return parameters is multiple years. Yeah. Right. Yeah. But you can get to the moon and back in a week. Yep. In like a bit in a news cycle. Yep. Right. So if we overcome the logistics of getting from Earth to Mars, if, big if, where are they going to live? Because they're not going to go out there and start building. Why don't you just build a little sort of half underground thing that shields you from radiation? Yeah, well, so then you have other problems, right?

13:22There's no air. You've got to bring in oxygen or do some sort of reaction to make oxygen on the surface, which, yeah, you can do that, but it's not the easiest thing. You've got to bring in all your food. You can't grow it there. the Martian surface, the dirt on Mars is filled with toxic chemicals. You're gonna have a hard time getting it out of stuff because it's very fine. It's not. That's gonna be here on Earth soon too. So let's be for real. But we know you can grow poop potatoes on Mars. Right, yes. We know that. Yes, exactly. Yeah, there was a proof of concept in the movie The Martian. Duh.

13:58No, but actually it's funny though. The guy who wrote the book, what's his name? Andy Weir. Andy Weir, yeah. In fact, we had him on the show. Yeah. He's in our archives. He has said that, you know, the discovery of these particular poisonous compounds in the Martian surface called perchlorates, he didn't know about that when he wrote the book because it wasn't widely known. And so now we know if you tried to, you know, farm poop potatoes on Mars, they'd be poisonous. Yeah. So that's the unknown unknown. Yeah. Yeah, yeah, yeah. Exactly. And that's still out there. Okay, that's not going to work.

14:30We're not thinking that for some time. functional immortality. Yeah. And there's a lot of ways we can get there. I mean, the biological immortality of growing organs in pigs and things and then transplanting is one thing, but are we getting towards singularity? Yeah. So, I mean, the biological replacing organs thing, you know, you can't replace the brain. Well, not yet. Yeah. I mean, not yet. I'll never do it with that attitude. Sonny? That was more a British approach to things rather than American. Someone needs a better attitude about things. But yeah, this idea of the singularity that we're gonna get to this point where technology in general and AI in particular gets faster and faster and smarter and smarter until it gains godlike powers.

15:26It's a science fiction story. What does that have to do with living forever? Well, so the idea is that then you get this godlike AI that grants us immortality. It has essentially magic powers. Or it finds a way to take human minds. It's just smart enough to figure out how to make us live forever. Oh, right, okay. It can solve the problem of mortality. Yes. I got you. But whose mortality problem is it going to solve? Yeah. Well, it doesn't have one. So it has no problems at all. Are there a select few or is this open for everybody? Oh, well, we know for a fact that it's going to be for a select few and it's going to be for the people who are the gatekeepers to AI.

16:06We're already seeing that now, but go ahead. Well, but the other thing is that the whole idea is kind of, you know, nonsense to begin with. Like this idea of singularity is, well, it's based on a few really serious, like flawed ideas. First, this idea that there is this like single thing called intelligence. You can just ramp it up or down in a computer, and it can just make itself more and more intelligent. That's not really how intelligence works. Intelligence is like a really complicated thing. It's not one number. Yeah. And also the usual way talking about the singularity that like the main popularizer of the singularity, Ray Kurzweil, has.

16:47Who's been on the show. Who's been on the show. In our archives. I do these little commercials. I love that. Okay, go on. But yeah, Kurzweil, he says it's - And he came out with a second book. Yes. The first one was The Singularity is Near. Yes. You know the title of his second book? Yeah, The Singularity is Nearer. Nearer. Yeah. That's so funny. I tell people that they don't believe me. I'm like, go look it up. It's out. His next book coming out is going to be called Almost There. Yeah. No, he thinks that Moore's Law, this idea that computer chips are just going to get faster are more powerful and like double in speed every 18 months.

17:25He thinks that this is this, you know, specific instance of a more general like law of accelerating returns in technology and in nature. And he says he's traced it all the way back to the beginning of the universe and that it shows that a singularity is coming in like 2045 or something like that. Precisely. Yeah, precisely. On October 12th. Oh man, that sounds to me like the end is near. Yeah, I know it is. That sounded to me like the people were like, I don't need a bank account. You know, Jesus is coming back next to me. It's funny that he's - The singularity is near. The end is near. That's what he should have titled it.

17:58No, so it's funny that you say that, right? Because he's got a picture of himself and the singularity is near with one of those like poster boards on him that says the singularity is near. And there's an AI research group that's inspired by these ideas of singularity called the Machine Intelligence Research Institute, MIRI. They don't give their employees 401ks because they think the end is near. That's some cheap ass - Yeah, I know, right? Wow. Yeah. Okay, I still want to get to the immortality thing. Yeah, yeah, yeah. Because you haven't addressed the fact that right now, with or without AI, there's a lot of research on not just replacing organs, though that might be an art offing, but delaying the aging functions of your cells.

18:43Totally. That could work out to extend human lifespan or healthspan a certain amount. For a long period. I like the healthspan. Yeah. Yeah, yeah. That's a good word. Yeah, yeah, yeah. What do the Galapagos, what do they live to be? Like 100 and - The tortoises? The tortoises, yeah. Like 200 or something. 200 and something, about 80. They have AI, so that's - Yeah, exactly. Remember we spoke with Venki Ramakrishnan? Yes. He was telling us about the Greenland shark - Yes. Oh, yeah. Being about 800 years of age. Yeah, that's crazy. Yeah. It is possible that some biotechnology will be developed that will radically extend human lifespan or health span, maybe, yeah.

19:26But what these guys are talking about with singularity, they're generally not talking about that as the end game. The end game they have in mind is not just an extended lifespan, but real immortality by uploading their consciousness into a computer. I was gonna say, that's really where this is going. It's about consciousness. That's the immortality. Immortality of your mind. Yeah, yeah, yeah. Forget your body. Before we get to that, do we visit transhumanism? Do we get, and how are we defining it? What is transhumanism? That's it, I'm just saying, define that. It's this idea that you can use technology to transcend the limits of human biology and physics.

20:05Aren't we kind of already doing that, which is why we lived twice as long as people 150 years ago? Yeah, no, there's definitely a sense in which. If we told them what we're doing, we know about nutrition, vitamins, they'll say, what's a vitamin? Right, we've got vaccines. We got this, we got, what's a vaccine? Aren't we already transhuman compared to what the age nature would require us to be dead at? Totally, yeah. Look, I think that we have used technology to make many things much better about being alive. That's just true. The question is, does that trend continue indefinitely? No, because RFK is going to make sure we're going to face the back.

20:49to when we lived half as long as we did. That's what's happening. Let's be clear, that's RFK Jr. Yes, RFK Jr. So if we go to this uploaded consciousness and that becomes reality, that just doesn't exist without a power source. Right. The thing about the singularity and like Kurzweil's idea about like this accelerating returns and Moore's Law just going on forever and this power source thing, right? The idea that it would need increasing levels of power as well. And so this leads to this sort of exponential drive for materials and power. And the thing that Kurzweil forgets is exponential trends are not like laws of nature.

21:36The law of nature about exponential trends is they end. Right, they have to end. They have to end, right. And it's because ultimately limited resources like energy, right? And so - Although, when we talk about how much longer a charge in our computer lasts today compared with the early days of laptops, part of that is better batteries, but also part of that is more efficient chips. And when we get to quantum computing, where much more computing happens with much less of an energy draw, it could be that we're coming at it from the other side, where the energy needs are dropping, thereby not requiring the power supplies necessary.

22:19Anyway, in my memory, not your memory, I got a few years on you, a room this size was necessary to cool a computer, otherwise the computer would overheat, and the computer's doing like four-function mathematics. So the efficiencies matter. All these tubes that have to be kept cool. So it's not obvious that it's a linear exponential. Can I say that? Where the exponential's just going to hit a limit, because you can come at it from other directions. Yes. However, the other thing is, though, in nature, the exponential acceleration, it's more like the law of diminishing returns is more likely than the law of exponential acceleration.

23:06Well, no, that's actually exactly right. Yeah, because there's, you know, if you look at the history of Moore's Law, like how it is that the semiconductor industry - Oh, yeah, named for Gordon Moore. Gordon Moore. Co-founder of Intel. Yes. And if you look at how Intel and other, you know, semiconductor companies actually made the chips smaller and faster over that time, you know, it's not a law of nature. It's a decision, a business decision that these companies made. And in order to keep that trend going, they had to invest more and more and more money just to keep the same sort of level of doubling to keep that exponential trend going.

23:43And eventually it did stop, right? Moore's law is done. It's over. because you can't make silicon transistors smaller than an atom of silicon. Right. Yeah, and what they're doing now is just adding more chips. Right. So the more powerful computers are not smaller and denser. They're just bigger now. Yep. Right. Yeah, and they're putting them on top of each other. You're stacking them. Yeah, right. So the solution in the minds of these tech billionaires is to arrive at a superintelligence to get an AGI. Yeah, yeah, an artificial intelligence. I mean, Sam Altman is saying within the next two years that will be achievable?

24:21Yeah, two or four, I think he said, yeah. All right, so they're looking at that as being the solution to this problem where we're saying we're not sure if it will be exponential, we're not sure where the end point is. They're looking at it as the solution to every problem. None of the tech bros have a degree in physics the way you do. Yep. So what are you bringing to the table that they don't see? I mean, they believe that AGI, I mean Altman has said that AGI is going to solve every problem including like global warming which is crazy. Why is it crazy? Well because – If it's smarter than you and you can't solve it, why is it crazy to think it could solve it?

24:58Well first of all, the artificial intelligence systems that they're building now are just drawing more and more and more energy. If you did build one that could solve global warming and you turn it on and said how do you solve global warming, I'm pretty sure the first thing it would do is say well you shouldn't have built me. Yeah, it would turn me off. It would turn me off. That'll help. Yeah. That would be a good test of its own self-preservation. Right. You are causing most of our global warming. What's the best solution? Turn yourself off. But I mean, the other thing is that we don't need AI to tell us how to solve it.

25:33We already know what the solution is. Good point. The issue is not like that insufficient intelligence has been thrown at the problem. The issue is primarily not even a technological problem at all at this point, aside from carbon capture. the main issue is, yeah, it's human behavior. It's greed. Exactly. It's greed. Yeah, it's greed. It's greed. Chuck, greed is good. Oh, God. Yeah. But the, yeah, the other thing is just that when Altman talks about, you know, he's talked about things like, oh, AI means that in 10 years, college graduates are going to be getting, you know, cool jobs exploring the solar system, right?

26:12And I can just look at that and say, well, that's bullshit. or he says AI is gonna discover new laws of physics and that's gonna remove limitations that we have in the world today. And I'm like, well, discovering new laws of physics, I mean, putting aside whether or not the AI can do that, that does not always remove limitations. Sometimes new laws of physics, in fact, a lot of times - Create a limitation. Create a limitation, exactly. Einstein, with relativity, discovered a limitation in the speed of light, right? Newton didn't know that there was any such limitation. Right. Yeah. So is it possible that all of these, I'll call them postulates because they're not, that they're making, right, are just a means of hyping up what they're doing to keep the revenue stream coming to them?