Neal Stephenson: Sci-Fi, Space, Aliens, AI, VR & the Future of Humanity | Lex Fridman Podcast #240

1. 0:00 – 0:43

   Introduction

   1. LFLex Fridman0:00

      The following is a conversation with Neal Stephenson, a legendary science fiction writer exploring ideas in mathematics, science, cryptography, money, linguistics, philosophy, and virtual reality, from his early book, Snow Crash, to his new one called Termination Shock. He doesn't just write novels. He worked at the space company Blue Origin for many years, including, technically, being Blue Origin's first employee. He also was the chief futurist at the virtual reality company Magic Leap. This is the Lex Fridman Podcast. To support it, please check out our sponsors in the description. And now, here's my conversation with Neal Stephenson.

2. 0:43 – 9:28

   WWII and human nature

   1. LFLex Fridman0:43

      You write both historical fiction, like World War II in Cryptonomicon, and science fiction-

   2. NSNeal Stephenson0:50

      Mm-hmm.

   3. LFLex Fridman0:50

      ... looking both into the past and the future. So let me ask, does history repeat itself? In which way does it repeat itself? In which way does it not?

   4. NSNeal Stephenson0:59

      I'm afraid it repeats itself a lot. Um, so I, I think human nature kind of is what it is, and so we tend to see similar behavior patterns emerging again and again. And so, uh, it's, it's kind of the, uh, exception rather than the rule when something new happens.

   5. LFLex Fridman1:20

      What role does technology play in the suppression or in revealing human nature?

   6. NSNeal Stephenson1:26

      Well, the standards of living, uh, life expectancy, all that, have gotten incredibly better within the last, particularly the last 100 years. I mean, just antibiotics, um, modern vaccines, electrification, uh, the internet, um, these are all, uh, improvements in most peoples' standard of living and health and longevity that, um, that exceed anything that was seen before in, in human history. Um, so, um, so people are living longer, they're generally healthier, and so on. Uh, but again, um, we still see a lot of the same behavior patterns, some of which are, uh, not very attractive.

   7. LFLex Fridman2:11

      So some of it has to do with the constraints on resources. Presumably, with technology, you have less and less constraints on resources, so we get to maybe emphasize the better angels of our nature? And in, in so doing, does that not potentially fundamentally alter the sort of th- the experience that we have of life on Earth?

   8. NSNeal Stephenson2:32

      Y- you know, until the last 10 or so years, I would've, uh, taken that view, I think, but, um, you know, uh, uh, people w- w- will find ways to be, um, to be divisive and angry, um, if it scratches a kind of psychological itch that they have got. And, um, we used to look at the Weimar Republic, um, what happened in the economic collapse of Germany prior to, um, the, the rise of Hitler, um, World War II, uh, and kind of, uh, explain Hitler, at least partially, by, um, just the, the misery that people were living in at that time. Um-

   9. LFLex Fridman3:17

      The economic collapse.

   10. NSNeal Stephenson3:18

       Yeah. The hyperinflation and unemployment and, um, the, the decline in standard of, of living. And that sounds like a, a plausible, uh, explanation, but there are economic troubles now for sure. We had the bank collapse in 2008, um, and there's stagnation in some peoples' standards of living. But it's hard to explain what we've seen in this country in the last few years just strictly on the basis of, uh, people are poor and angry and sad. I think they wanna be angry.

   11. LFLex Fridman3:50

       (laughs) So without being political in a divisive kind of way, can we talk about the lessons you can draw from World War II?

   12. NSNeal Stephenson4:01

       Sure.

   13. LFLex Fridman4:01

       This singular event in human history, it seems like.

   14. NSNeal Stephenson4:05

       Yeah.

   15. LFLex Fridman4:05

       And yet, as you say, history rhymes, at the very least.

   16. NSNeal Stephenson4:10

       Yeah. Being who I am, I tend to focus on the curious technological things that happened in, in conjunction with that war.

   17. LFLex Fridman4:19

       Right.

   18. NSNeal Stephenson4:19

       Um, which may not be where you wanna go. But, uh, the-

   19. LFLex Fridman4:22

       Well, there's several things in ... Sorry to interrupt. So one, in Crypto, uh, Nomicon is more like the Alan Turing side of things, right?

   20. NSNeal Stephenson4:29

       Yeah. Right.

   21. LFLex Fridman4:30

       And then, and then there, there's the outside of techno- ... Well, there's, uh, first of all there's the tools of war-

   22. NSNeal Stephenson4:36

       Yeah.

   23. LFLex Fridman4:36

       ... which is a kind of technology.

   24. NSNeal Stephenson4:37

       Mm-hmm.

   25. LFLex Fridman4:38

       But then there's just, like, the human nature, the nature of good and evil.

   26. NSNeal Stephenson4:41

       Yeah. Well, so one of the things that emerges from, uh, from the war and from the, um, the extermination camps is that we were never allowed to have illusions anymore about human nature. So y- you, you have to, to learn that lesson to be a- an educated person, and you have to know that, that even in a supposedly, you know, enlightened, civilized society, people can become monsters quite easily. So that is for sure the big takeaway.

   27. LFLex Fridman5:13

       So do you agree with S- Solzhenitsyn about, uh, what is it? Uh, the line between good and evil runs through the h- the heart of every man?

   28. NSNeal Stephenson5:23

       Yeah.

   29. LFLex Fridman5:24

       That all of us are capable-

   30. NSNeal Stephenson5:26

       Great line. Yeah.
3. 9:28 – 14:06

   Search engine morality

   1. NSNeal Stephenson9:28

      th- this is the theme of a book I wrote called The Diamond Age, which, you know, talks about a book that essentially does that. And, um, I've been sort of watching people try to come at the, the problem of building that thing, uh, from different directions for... e- ever since the book came out, basically. Um, and so, uh, the, uh... A- and so I- I kind of have a... although I haven't worked on it myself, I do get a sense of the, the level of difficulty in, in realizing that, that, that goal. Um...

   2. LFLex Fridman10:01

      So, that book is in, in the '90s, so as Google is coming to be.

   3. NSNeal Stephenson10:07

      Yeah.

   4. LFLex Fridman10:08

      It's essentially, uh... Not Google, but the search engine, the initial search engines and then... which gave birth to Google, essentially, in, in contrast.

   5. NSNeal Stephenson10:16

      Right. Yeah, yeah. That was still in the era of AltaVista and AskJeeves and multiple different, uh, search engines. And yeah, I'm pretty sure I had not heard of Google at that point. That would have been '95, '96. And I think the book came out in '94.

   6. LFLex Fridman10:33

      And then, of course, the social networks followed, which is another form of, um, guidance through the space of information.

   7. NSNeal Stephenson10:41

      Yeah.

   8. LFLex Fridman10:41

      And-

   9. NSNeal Stephenson10:42

      Well, what happens is that these things come along and then people find ways to game them. Um, and so, uh, I s- I saw an interesting thread the other day pointing out that, um, you know, uh, 20 years ago, if you had Googled, um, Pythagorean theorem, uh, chances are you would have been taken directly to a page explaining the Pythagorean theorem. If you do it now, you're probably gonna... the top hits are gonna be from somebody who's, who's got an angle, who's got a scheme, right? They're, like, trying to sell you math tutoring or, you know, they're, um, uh, they're working some kind of marketing plan on you. Um, so the, the traditional engines, um, become actually less useful over time for their original educational purpose. That doesn't mean that they can't... they shouldn't be replaced by newer and better ones.

   10. LFLex Fridman11:41

       F- first of all, to defend the people with the angle.

   11. NSNeal Stephenson11:43

       Mm-hmm.

   12. LFLex Fridman11:44

       Right? Th- they're trying to find business models-

   13. NSNeal Stephenson11:47

       Yeah.

   14. LFLex Fridman11:47

       ... to fund oftentimes... w- which is funny you went with Pythago- (laughs) like, you went at math. (laughs) Those greedy bastards. (laughs)

   15. NSNeal Stephenson11:56

       Yeah, I know.

   16. LFLex Fridman11:57

       But it's gr- it's great because-

   17. NSNeal Stephenson11:58

       How can we monetize the Pythagorean theorem? Yeah.

   18. LFLex Fridman12:02

       (laughs) Well, I mean, e- education, right?

   19. NSNeal Stephenson12:04

       Yeah.

   20. LFLex Fridman12:04

       It's to figure out, like, w-... people who love math education, for example, love it purely, not purely, but very often love it for itself, for j- just teaching math.

   21. NSNeal Stephenson12:17

       Yeah.

   22. LFLex Fridman12:17

       But then they start, you know, when coming face to face with, for example, like, the YouTube algorithm, they start to try to figure out, "Okay, how can I make money off of this?" The, the primary goal is still that love of education, but they also want to, to make that love of education a f- their full-time job. But, I, I see that, sort of that dance of humanity with the algorithms as, uh, it finds this kind of local pocket of optimality that's ... Or sub-optimality, whatever.

   23. NSNeal Stephenson12:50

       Yeah.

   24. LFLex Fridman12:50

       It's, gets stuck in it.

   25. NSNeal Stephenson12:51

       Pocket, anyway.

   26. LFLex Fridman12:51

       Well, it's a pocket of some sort. And, but I see that pocket as way better than what we had before-

   27. NSNeal Stephenson12:57

       Mm-hmm.

   28. LFLex Fridman12:57

       ... in the '80s, right?

   29. NSNeal Stephenson12:58

       Okay.

   30. LFLex Fridman12:58

       Or '90s, before the internet. But like, and now we're now, this is, this is also human nature, we start, uh, writing very eloquent articles about how this pocket is clearly a pocket, it's not very good, and we can imagine much better lands far beyond. And, but the reality is, it's better than before-
4. 14:06 – 31:07

   Space exploration

   1. NSNeal Stephenson14:06

      the barrier.

   2. LFLex Fridman14:06

      And eventually we'll be, uh, as you've talked about, I mean, we'll be ... We'll colonize the solar system and then, uh, well, we'll be stuck in the solar system and then, uh, people will say, "Well, we're screwed when, 'cause when the sun energy runs out, there's no way to get to the, the next solar system," and then th- and so on, it goes on until we colonize the entirety of the observable universe.

   3. NSNeal Stephenson14:28

      Yeah.

   4. LFLex Fridman14:29

      (laughs)

   5. NSNeal Stephenson14:29

      I, I think, I think getting out of the solar system is going to be a hard one, but, uh ...

   6. LFLex Fridman14:33

      So can you, you mentioned this, can you elaborate why you think, back to sort of a serious question, why do you think it's hard to get outside of our solar system?

   7. NSNeal Stephenson14:43

      It's just a energy calcul- I mean, you, you can do it slowly, m- uh, whenever you want, um, but, uh, the idea of getting there in, you know, uh, one lifetime or multiple, a few lifetimes is, uh, requires huge amounts of energy to, to accelerate. Um, and then you, as s- soon as you get halfway there, you need to expend an equal amount of energy to decelerate, or you'll just go shooting by. Um, and so, um, that means carrying a lot of energy, and there's, there's, uh, ideas, like Yuri Milner c- I, I think is still funding the, the idea to use laser propulsion to send something, uh, to another star system, a small object. Um, but it'll have no way to slow down, as far as I know.

   8. LFLex Fridman15:35

      Right. They never talk about that part.

   9. NSNeal Stephenson15:36

      Yeah.

   10. LFLex Fridman15:37

       Like, how do we slow down?

   11. NSNeal Stephenson15:38

       Yeah. Um, (laughs) so-

   12. LFLex Fridman15:40

       It's a quick flyby. You take a good picture, I guess.

   13. NSNeal Stephenson15:43

       Yeah, you better take some good pictures on your way by. So, and that's great if it happens, I'm not knocking it, um, but the amount of energy is, is, uh, that's needed is just staggering, and there's, there's other issues like just how do you maintain, uh, uh, uh, an ecosystem for that long in isolation? Uh, how do you prevent people from going crazy? What happens if you hit something while traveling at a significant fraction of the speed of light?

   14. LFLex Fridman16:10

       What about sort of, um, some combination of expanding human lifespan, but also just good old-fashioned stable society on a spaceship?

   15. NSNeal Stephenson16:21

       Yeah, yeah, the, the generation-

   16. LFLex Fridman16:23

       Noah's Ark.

   17. NSNeal Stephenson16:23

       ... ship, yeah, yeah.

   18. LFLex Fridman16:24

       Yeah.

   19. NSNeal Stephenson16:24

       No, I think that's the only way. It would, it would have to keep going for a long time. Um, and they might get to where they're going and find, uh, a, a shitty, um, solar system. Like, w- we can try to-

   20. LFLex Fridman16:40

       (laughs)

   21. NSNeal Stephenson16:40

       We can try to do some advanced survey, but, I mean, if, if you get there and all the planets in that solar system are just garbage planets, then it's kind of a big letdown for, for this, like, thousand-year voyage that you've just, uh, you've just been on.

   22. LFLex Fridman16:56

       Mm-hmm.

   23. NSNeal Stephenson16:57

       Right? So, I mean, we have a pretty narrow range of, of parameters w- that we need to stay, uh, between in order to survive, um, in terms of the, the gravitational, uh, field that we can deal with, um, so that such a, sets a bound on the size of the, the planet, and, um, what we need in the way of temperature and atmosphere and so on. So, when you look at all those complications, then, um, basically building a, uh, sort of, uh, exactly the environment we want out of available materials in this solar system starts to look a hell of a lot better. Um, as, it's hard to make, uh, um, an economic argument, let's say, for, for, for making that journey. Uh, one of the things I like about The Expanse is the fact that the people who are trying to build the starship to go to the other solar system are doing it for religious reasons, and I think that's the only reason that you would do it, um, 'cause economically, it just makes more sense to build rotating cylindrical space habitats and make them perfect.

   24. LFLex Fridman18:10

       Well, is-isn't everything done for religious reasons? Like why do we... Exploration?

   25. NSNeal Stephenson18:15

       Yeah.

   26. LFLex Fridman18:15

       Like what, why- why did we go to the moon again?

   27. NSNeal Stephenson18:17

       Mm-hmm.

   28. LFLex Fridman18:18

       And do the other things? Uh, what has JFK said? It's because, not because they're easy but because they're hard.

   29. NSNeal Stephenson18:24

       I knew a-

   30. LFLex Fridman18:24

       Isn't that kind of a religious reason?
5. 31:07 – 39:30

   Aliens and UFOs

   1. NSNeal Stephenson31:07

   2. LFLex Fridman31:07

      Mm-hmm.

   3. NSNeal Stephenson31:07

      But...

   4. LFLex Fridman31:08

      Can I ask you, um, a question on propulsion that's a little bit more out there? So, I don't know if you've, uh, seen quite, uh, a lot of recent articles and reports and so on about, uh, UFOs.

   5. NSNeal Stephenson31:23

      Mm.

   6. LFLex Fridman31:24

      Like the Tic Tac aircraft.

   7. NSNeal Stephenson31:26

      I keep seeing a lot of chatter about it, but I haven't gone deep into it.

   8. LFLex Fridman31:32

      So the DoD released footage filmed by, um, pilots, and there's a lot of reports about objects that moved in ways they haven't seen before that seemed to, uh, defy the laws of physics if we consider the aircraft that we have today. And so, the, the reason I ask you that is because it kind of, um... To, to me, whatever the heck it is, it's inspiring for the possibilities of i- ideas for propulsion.

   9. NSNeal Stephenson32:07

      Hmm. Mm-hmm.

   10. LFLex Fridman32:08

       If it's like, um, secret projects from foreign nations or it's physical phenomena that we don't yet understand, like ball lightning, all those kinds of things-

   11. NSNeal Stephenson32:18

       Mm-hmm.

   12. LFLex Fridman32:18

       ... or if it is aliens or objects from an alien civilization, I most likely believe it's fr- if it's an object from an alien civilization, it's gotta be like a really dumb drone that just like-

   13. NSNeal Stephenson32:31

       Yeah.

   14. LFLex Fridman32:31

       ... got lost.

   15. NSNeal Stephenson32:32

       (laughs)

   16. LFLex Fridman32:32

       It's, it's definitely not-

   17. NSNeal Stephenson32:34

       Yeah.

   18. LFLex Fridman32:34

       ... like the pinnacle (laughs) of intelligence. It's like, some like teenager's like, uh-

   19. NSNeal Stephenson32:41

       Science fair experiment.

   20. LFLex Fridman32:42

       Yeah, it just flew for-

   21. NSNeal Stephenson32:43

       Yeah.

   22. LFLex Fridman32:43

       ... for a few centuries out and just landed, and then we humans are all like really excited about this.

   23. NSNeal Stephenson32:49

       (laughing) Yeah.

   24. LFLex Fridman32:50

       (chuckles) This, uh, this wild thing. I mean, what do you, what do you think about those, um... First of all, like the millions of reports of UFOs, right? There's some psychology there that's deeply cultural, uh, but also the possibility of aliens having visited Earth.

   25. NSNeal Stephenson33:05

       Yeah, I mean, I'd like to see some better pictures. Uh, for the reason I mentioned earlier, having to do with the difficulty of traveling between star systems, it's really hard for me to believe it's aliens. Uh, I, I just can't understand why you would, um, go to all that trouble to transport something across light years and then do what these UFOs are allegedly doing. Like, how is that interesting? How, how does that justify the trip?

   26. LFLex Fridman33:36

       So, if you travel across, you know, those kinds of distances, you would make a bigger splash?

   27. NSNeal Stephenson33:45

       Th- first of all, I, I would expect that the, the arrival of these things would be something we'd notice. It's gotta, you know, decelerate into, into our solar system by... Un- unless it got here really, really, really slowly, so I guess that's a-

   28. LFLex Fridman34:00

       (laughs)

   29. NSNeal Stephenson34:01

       ... that's a possibility, and just kind of snuck in.

   30. LFLex Fridman34:04

       So at the end, we would detect some kind of footprint in terms of energy?
6. 39:30 – 46:52

   SpaceX and Blue Origin

   1. NSNeal Stephenson39:30

      forms.

   2. LFLex Fridman39:30

      I have to ask you, um, in terms of space, just looking at where Blue Origin is, looking at where SpaceX is today-

   3. NSNeal Stephenson39:37

      Mm-hmm.

   4. LFLex Fridman39:39

      ... and maybe looking out 10, 20 years out from now, are you impressed with what's happening? We just saw William Shatner go up to space.

   5. NSNeal Stephenson39:47

      Yeah, I was ju- I was just watching his video th- this morning before I came here.

   6. LFLex Fridman39:51

      (laughs)

   7. NSNeal Stephenson39:51

      Yeah. Yeah.

   8. LFLex Fridman39:52

      Are you impressed at where things stand today?

   9. NSNeal Stephenson39:54

      Yeah. I mean, uh, I mean, SpaceX in particular is, has d- done things that are just unbelievable. Um, and, um, uh, uh, I don't think anyone was anticipating, um, 20 years ago, let's say, when this all started, just the, uh, the speed with which they'd be able to, um, rack up these incredible achievements. I- if you've kind of, uh, even seen a little bit of how the sausage is made and, and sort of the, the, the difficulty of, of doing any kind of space travel, um, what they've achieved is, uh, is just, uh, is, is unbelievable.

   10. LFLex Fridman40:36

       What about, uh, maybe a question about Elon Musk? Um, even more than Jeff Bezos, he has a very kind of ambitious vision-

   11. NSNeal Stephenson40:46

       Mm-hmm.

   12. LFLex Fridman40:47

       ... of, um, this project that we're on as a species-

   13. NSNeal Stephenson40:51

       Yeah.

   14. LFLex Fridman40:51

       ... of becoming a multiplanetary species, and becoming that quickly, like-

   15. NSNeal Stephenson40:57

       Yeah, I-

   16. LFLex Fridman40:57

       ... as soon as possible. Landing on Mars, colonizing Mars.

   17. NSNeal Stephenson41:01

       Yeah.

   18. LFLex Fridman41:01

       What do you think of that project? There's two questions to ask. First, the question is what, wha- what do you think about the project of colonizing Mars?

   19. NSNeal Stephenson41:08

       Mm-hmm.

   20. LFLex Fridman41:09

       And second, what do you think about a human being who is so unapologetically ambitious at achieving the impossible, at what a lot of people would say is impossible?

   21. NSNeal Stephenson41:23

       I think that colonizing Mars is the kind of, of goal that's, uh, it's easily stated, uh, it's, um, it's catchy, it's, it's, it's the kind of thing that it can inspire people to get involved in a way that some other programs might not. Um, so I think it's well-chosen in that way. Um, I have technical questions about, um ...

   22. LFLex Fridman41:49

       (laughs)

   23. NSNeal Stephenson41:50

       You know, there's, there's a, a problem of perchlorates, uh, on the surface of Mars that, that's gonna be big trouble. Um, and there's, there's radiation, so, and it ... peop- this is known, I'm, uh, b- but, um-

   24. LFLex Fridman42:04

       What about business questions?

   25. NSNeal Stephenson42:06

       Mm.

   26. LFLex Fridman42:06

       Do you think, 'cause you mentioned sort of, uh, going outside of the solar system would, would best be done for religious reasons. Um, what about colonizing Mars? Can you spin it into a business proposition?

   27. NSNeal Stephenson42:20

       It's hard to think of a resource that's on Mars that could be brought back here cheaply enough to compete with, um, uh, with stuff we could just dig out of the ground here or grow here.... um, so I- I don't know if there is a business plan for that or if it's just strictly we're gonna go there and, and see what happens, um, you know.

   28. LFLex Fridman42:48

       (sighs) Maybe again we need communism to kinda-

   29. NSNeal Stephenson42:51

       Yeah.

   30. LFLex Fridman42:51

       ... to get us going. (laughs)
7. 46:52 – 51:19

   Social media

   1. LFLex Fridman46:52

      Do you think that there's a lot to be done in the digital space? That's, uh, we mentioned sort of Wikipedia and knowledge. Don't you think there could be a lot of flourishing in the space of innovation, in terms of innovation in the digital space?

   2. NSNeal Stephenson47:07

      Yeah, I mean, I'd like to see that. I think it's where a lot of the brain power went during the last couple of generations, um, because people who might previously have been building rockets or other kinds of sort of hard technologies, uh, ended up instead going into programming, computer science, uh, which is understandable and great. Um, we've got structural problems right now in the way social media works that are pretty severe, and so I certainly hope that we're not... You know, 10 years from now that we're not exactly where we are today when it comes to that stuff. We need to move on.

   3. LFLex Fridman47:51

      The beautiful thing about problems is they show you how not to do things.

   4. NSNeal Stephenson47:57

      Yeah.

   5. LFLex Fridman47:57

      And they, uh, give you, give opportunity to, uh, new ideas to flourish and to beat out the ideas of the old.

   6. NSNeal Stephenson48:06

      Mm-hmm.

   7. LFLex Fridman48:06

      Which is, uh, a dream for me in, in... To see, um, new social media.

   8. NSNeal Stephenson48:13

      Yeah.

   9. LFLex Fridman48:13

      That beats out the ways of the old. So I, I tend to... You perhaps agree that it's not... That it's impossible to do social media well.

   10. NSNeal Stephenson48:22

       Oh, not at all. I mean, I listened to your, uh, interview with Jaron a couple weeks ago and I, I know Jaron and we've, you know, we've talked about this and-

   11. LFLex Fridman48:31

       He went, he went hard on me. He basically said, like-

   12. NSNeal Stephenson48:34

       In his very-

   13. LFLex Fridman48:34

       ... "It's impossible." (laughs)

   14. NSNeal Stephenson48:35

       ... in his very nice way... Well, the last time I kind of paid attention to Jaron's thoughts on it, he was thinking in terms of...... that basically there should be, you know, micropayments, uh, such that if I, by clicking the like button on something, I am essentially giving, um, valuable intellectual property to Facebook or Twitter or whatever. Uh, it's not a very large amount of IP, but it's definitely a transfer of information that, that when they aggregate it, it is beneficial to them. So, and now I, now I do remember that he, uh, on, on his interview with you was talking about what data unions or-

   15. LFLex Fridman49:21

       Mm-hmm.

   16. NSNeal Stephenson49:21

       ... yeah.

   17. LFLex Fridman49:22

       Those are a lot of interesting ideas, but for me, the biggest disagreement was in the level of cynicism. He has a distrust and cynicism towards people in Silicon Valley being able to do these kinds of things. And I'm really ... Okay, when you have a large crowd of people who are doing things the wrong way-

   18. NSNeal Stephenson49:44

       Mm-hmm.

   19. LFLex Fridman49:45

       ... you should nevertheless maintain optimism because what's important is to find the one person in that room that's going to do things the right way. Cynicism is going to completely silence out the whole room. So, he was saying, "I've, I've been here a long time."

   20. NSNeal Stephenson50:00

       Oh, yeah.

   21. LFLex Fridman50:01

       "I, I've known, you know, I, I under- like how these folks work. They think they're gods and they know the right way to do things and they will tell you how to do those things. And that kind of hubris is going to always lead you astray when you are the one who is engineering the algorithms." And there's a lot of deep truth to that because algorithms are powerful and, uh, many people when given power do not do the best of things. I mean, most, w- what is it, uh, the old Lincoln line, "If you want to test a man's character, give him power."

   22. NSNeal Stephenson50:39

       Yeah.

   23. LFLex Fridman50:40

       Yes, but that doesn't mean that some people are not able to handle the power, that some people are not able to come up with good, uh, ideas that create better social media.

   24. NSNeal Stephenson50:51

       Yeah, I didn't interpret Jaron's statements as, as being entirely cynical and hopeless. I mean, he's, he's definitely raising, you know, issues of concern, um, but he wouldn't be out, you know, writing the books that he's written and talking about this stuff if he didn't think there was a way.

   25. LFLex Fridman51:09

       If, if he didn't think there was hope, yeah. And part of it, as you probably know with Jaron, he just loves a good argument. (laughs)

   26. NSNeal Stephenson51:15

       Yeah.

   27. LFLex Fridman51:16

       (laughs) He just loves to have a little bit of fun.

8. 51:19 – 1:03:09

   Climate change

   1. LFLex Fridman51:19

      Well, I have to ask you about, uh, I mean we talked about taking all big bold risky ideas. So, in your new book, Termination Shock, it's set here in Texas.

   2. NSNeal Stephenson51:32

      Part, part of it is, yeah.

   3. LFLex Fridman51:34

      Yeah.

   4. NSNeal Stephenson51:34

      Most, most of it.

   5. LFLex Fridman51:35

      Yeah, it's a great place to set it. So, in it, the main character, T.R. McCulligan, a Texas billionaire oil man and truck stop magnate decides to solve climate change, take on climate change by himself. So, this is an interesting philosophical exploration of how to solve climate change from a perspective that's perhaps different than we've been thinking about. So-

   6. NSNeal Stephenson51:57

      Yeah, I wouldn't, I wouldn't use the word solve, but let's say ameliorate-

   7. LFLex Fridman52:02

      Ameliorate.

   8. NSNeal Stephenson52:02

      ... the temporary effects. But please.

   9. LFLex Fridman52:05

      Take on.

   10. NSNeal Stephenson52:06

       Yeah.

   11. LFLex Fridman52:07

       Take on the challenge. So, it's, it's very interesting but as, s- so there's a gradual nature to this process.

   12. NSNeal Stephenson52:15

       Yeah.

   13. LFLex Fridman52:15

       And I mean just like in, in your book, um, the power of innovation is something that has, uh, saved us quite a few times in history. So, what role does that play as, in this gradual process?

   14. NSNeal Stephenson52:34

       Right. So ultimately, we don't solve the problem until we get the CO2 out of the atmosphere, um, but that is going to take a while. Um, we're still adding more, uh, we haven't even started to, to reduce the amount. So, um-

   15. LFLex Fridman52:54

       So there's two possibilities, and sorry to interrupt, is reduce the amount that we're putting in the atmosphere and two is removing what we got in the atmosphere.

   16. NSNeal Stephenson53:03

       We have to do both.

   17. LFLex Fridman53:04

       Right. And th- those are two different kind of, uh-

   18. NSNeal Stephenson53:07

       Yeah.

   19. LFLex Fridman53:08

       ... efforts in terms of like what's involved.

   20. NSNeal Stephenson53:10

       Because it stays up there. So I think just last week China announced that they're going to try to level off their CO2 emissions in like 2030. So 2031 they'll only put as much CO2 into the atmosphere as they did in 2030, which is still a lot of CO2. In 2060 they're saying we'll be net zero. So, if everyone in the world does that and the PPM of CO2 in the atmosphere by then is, say, 450 parts per million, it will stay at 450 parts per million until we take it out. And taking it out, um, is hard, it's a, you know, it's a big... We, we, it took us a long time. We had to empty out huge coal mines and oil reservoirs and burn all that stuff. We had to chop down forests and dig up peat bogs, um, in order to create all of that CO2. And so, we have to reverse all of those processes, uh, somehow in order to remove the CO2 and get it back down hopefully into the 200 and some parts per million range where it used to be.

   21. LFLex Fridman54:26

       So, how about you get a, a single Texas billionaire to have a massive gun that blasts huge quantities of sulfur into the up- upper atmosphere?

   22. NSNeal Stephenson54:35

       So this is-

   23. LFLex Fridman54:35

       Like that's idea number one.

   24. NSNeal Stephenson54:37

       That's, uh, this is called solar geoengineering and it's, uh, we know that it's a possibility on a technical level because volcanoes have been doing it forever. Um, so many times in human history we've seen a volcanic eruption that was followed by a global cooling trend that lasted for a couple of years. And, uh, one of these things happened, I think in the '60s or '70s in Indonesia and, um, and Australians sent a, a plane up into the stratosphere to take some samples of the plume, and when it came back down, the windscreen of the plane had sort of a deposit on it. So one of the Australian scientists licked it (clears throat) and reported that it was painfully acid.

   25. LFLex Fridman55:25

       Mm-hmm.

   26. NSNeal Stephenson55:26

       So that was our first kind of clue that what was being injected into the stratosphere was sulfur dioxide. Um, so, um, and, and so we know b- then, well, Pinatubo came along in the '90s and, and did this experiment for us. So we know that sulfur in the, uh, in the stratosphere, it forms little, uh, spherical droplets of sulfuric acid after it combines with water, and those bounce back some of the sun's rays and, um, reduce the amount of, of solar energy entering the troposphere, which is where we live. So, um, so we know that it works and we f- we also know that the stuff goes away after a couple of years, so it gradually washes out, and so it's not a permanent thing. You have to c- uh, the, it's, uh, the good news, bad news is, um... Good news is, it's not permanent so if you don't like what's happening, you can just stop and wait a couple of years and you'll get back to where you started. And the, the, the bad news, if, if you're in favor of this kind of thing is that you have to keep doing it forever.

   27. LFLex Fridman56:37

       Right.

   28. NSNeal Stephenson56:37

       Or, um... So, so this guy's one of those... he, he's read these papers, he, he under- the TR, the character in the book, he knows all this. And all, all people who, uh, are familiar with climate science are, kind of know this. It's a pretty well-established fact. And so, um, he just decides he's gonna take action unilaterally and, and do this. Um, and so, uh, there's different ways to get the sulfur up there but because it's Texas, he builds the biggest gun in the world.

   29. LFLex Fridman57:12

       Yeah.

   30. NSNeal Stephenson57:13

       Uh, it's just six barrels pointed straight up and he begins firing shells loaded with sulfur into the stratosphere. And so the book is about not so much that as how people react to his doing that, uh, what the political ramifications are around the world, 'cause, you know, this is a extremely controversial idea and not everyone's on board with it. And even if you are willing to consider using a technological intervention, the, the fact is that it's gonna have different effects on different parts of the world. So some areas may suffer, um, negative, uh, even more negatives than positives, uh, and they're not gonna be happy.
9. 1:03:09 – 1:07:50

   Consequences of big ideas

   1. LFLex Fridman1:03:09

      Um, what do you make of sort of big bold ideas that have, uh, a, that are a double-edged sword? Are all ideas like this, a- all big ideas like this, they have, uh, they have the potential to have, um, highly beneficial consequences and a potential to have highly destructive consequences?

   2. NSNeal Stephenson1:03:34

      I wouldn't say all. I think, you know, going back to the, what we were talking about earlier, you know, how technology developed in the 50s and 60s, there was a, a period of time there when, um, people maybe had unrealistic ideas about new technology and weren't sufficiently attentive to the possible downsides. So, um, so we got, um, and, and there's a reason why. I mean, uh, the, the, there's th- uh, uh, you know, in, in the mid-20th century we saw, you know, antibiotics, we saw the polio vaccine, we saw, um, just simple things like refrigerators in the home, you know. Um, my, my grandmother, to her dying day, called the refrigerator the ice box because when she grew up it was a box with ice in it. So, you see all that change and it's largely for the benefit of people, and so if somebody comes along and says, "Hey, we're gonna build nuclear reactors to, to make energy," or, um, "Here's a new, um, chemical called DDT that's gonna kill, uh, mosquitoes," then, um, it's easy to, uh, to just buy into that and not be alert to the possible downsides. And, of course we know that, um, the, the way that those r- early reactors were built and the way that the, the supply chain, uh, was built to, to g- to create the fuel, um, and deal with the, the waste, um, was, was poorly thought out and, uh, and, and we're still dealing with the, uh, the, the resulting problems at places like Hanford in, in the state of Washington. And we know that, uh, DDT, although it did kill a lot of insects, um, also had terrible effects on bird populations. Um, so the, the kind of backlash that happened in the 70s that's, is still kind of going on, is, is to sort of assume that everything is a double-edged sword and always to look for the, the, you know, we have to absolutely convince ourselves that the, the downside, uh, isn't gonna come back and, and bite us, uh, before we can adopt any new technology. And I, I think the, the, people, um, people are overly sensitized to that now.

   3. LFLex Fridman1:06:10

      (laughs) Yeah, i- it's just funny, depending on the technology people are a little bit too terrified of certain technologies, like artificial intelligence is one. My sense is, that the things that they're afraid of aren't the things that are likely going to happen in terms of negative things. It's probably impossible to predict exactly the unintended negative consequences. But what's also interesting is for AI as an example, not, people don't think enough about the positive things. I mean, the same is true with social media. It's very popular now, for some reason, to talk about all the negative effects of social media. We've immediately forgotten ...... how incredible it is to connect-

   4. NSNeal Stephenson1:06:57

      Yeah.

   5. LFLex Fridman1:06:57

      ... across the world. How, there's a, there's a deep loneliness within all of us. We long to connect, and social media, at least in part, enables that, even in its current state.

   6. NSNeal Stephenson1:07:07

      Mm-hmm.

   7. LFLex Fridman1:07:08

      And the, all the negative things we see with social media currently are also in part just revealing the basics of human nature. It didn't make us worse, it's just reve- it's bringing it to the surface. And step one of solving a problem is bringing it to the surface. The fact that we are divis- there's a division, the fact that there, we're easily angered and upset and all, all of that, the witch hunts, all of those kinds of things-

   8. NSNeal Stephenson1:07:33

      Mm-hmm.

   9. LFLex Fridman1:07:34

      ... that's human nature, and it just reveals that, allowing us to now work on it. It's therapy. (laughs)

   10. NSNeal Stephenson1:07:40

       Yeah.

   11. LFLex Fridman1:07:40

       And so, th- that's another example of a technology that's just, w- we're, we're not considering the sort of, the pos- positive effects now and in the future enough of.

10. 1:07:50 – 1:30:58

    Virtual reality

    1. LFLex Fridman1:07:51

       I have to ask you about, um ... There's a million things I can ask you about, but virtual reality, I got- gotta ask you. Uh, you've thought about virtual reality, mixed reality, uh, quite a bit. W- what are the, uh, interesting trajectories do you see for the proliferation of virtual reality or mixed reality in the next few years?

    2. NSNeal Stephenson1:08:13

       Yeah, so I was, um, I was at Magic Leap for, what, five years? Um-

    3. LFLex Fridman1:08:19

       With the best title of all time.

    4. NSNeal Stephenson1:08:21

       Oh, thanks. Ch- chief futurist?

    5. LFLex Fridman1:08:24

       Yeah.

    6. NSNeal Stephenson1:08:24

       Yeah. And so I sort of had a, a little squad of people in, in Seattle doing what you might call content R&D. So, we were trying to make content for AR, but, um, (sighs) because it's such a new medium, uh, w- there's l- it's more of an engineering R&D project almost than a, than a creative project. So-

    7. LFLex Fridman1:08:49

       (sniffs)

    8. NSNeal Stephenson1:08:49

       ... it was fascinating to see, um, everything that goes into m- making, uh, an AR system that runs. Um, so AR, um, an AR device, if it's really gonna do AR, needs to be running SLAM in real time, and that alone is a big...

    9. LFLex Fridman1:09:11

       And s- so for people who don't know, first of all, virtual reality is creating a- almost fully artificial world and putting you inside it. Augmented reality, AR, is taking the, the real world and putting top on top ... putting stuff on top of that real world. And when you say SLAM, that means in real time, the device needs to be able to sense, accurately detect everything about that world sufficiently to be able to reconstruct it, the, the 3D, uh, structure of it so you can put stuff on top of it. And doing that in real time, presumably not just real time, but in a way that's creates a pleasant experience for the human perception system is, uh, yeah, that's a, that's an engineering project.

    10. NSNeal Stephenson1:10:01

        Right. Yeah, well said, and it's just one of the things that the system has to do. It's also tracking your eyes so it knows what you're looking at, uh, how far away what you're looking at is. Um, i- it's, uh, um, it's performing all those functions, um, and it's gotta, uh, keep doing that without, you know, burning up the, the CPU or, or, or, uh, depleting the battery, uh, unreasonably fast. And that's, that's just table stakes, it's just the basic functions of the, the operating system. And then any content that you wanna add has to sit on top of that. It's gotta be rendered by the optics, um, at a sufficiently low latency that, um, it looks real and you don't get sick. So, it's an amazing thing, and, um, you know, Magic Leap shipped a device that can do that in 2019. Um, and they're about to ship the ML2.

    11. LFLex Fridman1:11:04

        Mm-hmm.

    12. NSNeal Stephenson1:11:05

        Um, but I don't know any more about that than anyone else 'cause I d- don't work there anymore.

    13. LFLex Fridman1:11:11

        (laughs)

    14. NSNeal Stephenson1:11:11

        Um...

    15. LFLex Fridman1:11:12

        W- does it still in s- to some degree boil down to a killer app? A c- a content question? L- like you said, it's kind of a wide open space. Nobody knows exactly what's going to be the compelling thing.

    16. NSNeal Stephenson1:11:24

        Yeah.

    17. LFLex Fridman1:11:25

        So, doesn't a super compelling experience of some sort alleviate some of the need for engineering perfection?

    18. NSNeal Stephenson1:11:36

        Well, there's a, a base layer of engineering that you have to have no matter what, um, but you're certainly right that people, like in the early days of video games, put up with kind of low frame rate and, and what we would now call crappy graphics because they were having so much fun playing Doom or, or whatever.

    19. LFLex Fridman1:11:55

        Right. Even Tetris.

    20. NSNeal Stephenson1:11:57

        Yeah. Yeah. So, um, so for sure that's true, and so, um, you know, uh, uh, I was, uh, working on consumer-facing content. Um, th- there was a great team in Wellington, New Zealand that, that made a game, uh, called, uh, Dr. Groybrot's, uh, Invaders that, um, that, uh, r- realized the, the potential of, uh, of AR gaming in a way that I don't think anything else has, uh, before or since. Um, and, um, so that was definitely the strategy, um, (clears throat) until, uh, what, April 2020, which is when the company decided to, uh, pivot to commercial industrial applications instead. Um, so, um, and...... you know, I- I- I haven't seen their, their, their financial projections, but I assumed they had good reasons for, for making that strategic decision. Um, it just means that it's no longer, uh, necessarily targeted at, at just end users who want to play a game or, or be entertained. But it's, you know ...

    21. LFLex Fridman1:13:16

        That, to me, from a sort of a, a dreamer, futurist perspective is heartbreaking 'cause I, I, I don't know necessarily from in the VR space, but I see this kind of thing with, uh, with robotics where, to me, the future of robotics is consumer-facing. Uh, and a lot of great roboticists, Boston Dynamics and, uh, companies like that are focused on sort of industrial applications.

    22. NSNeal Stephenson1:13:47

        Yeah. Yeah, it's-

    23. LFLex Fridman1:13:48

        Because for financial business reasons.

    24. NSNeal Stephenson1:13:50

        Yeah. No, I can see the parallels for sure. You know, we'll see. It was a fun project, you know. We, uh, uh, we worked on um, an app for example called Baby Goats which just populated your room with, with baby goats.

    25. LFLex Fridman1:14:07

        That seems like a killer app right there.

    26. NSNeal Stephenson1:14:09

        Well, we thought highly of the, of the idea for sure.

    27. LFLex Fridman1:14:13

        Yes.

    28. NSNeal Stephenson1:14:13

        Um, so but because of the SLAM, uh, the, the, um, the system knew, for example, here's a table, here's a little end table. We know the heights, uh, we know how high our animated baby goat can jump. Um, and so, um, so our engineers had to, to build a system for converting the SLAM primitives into, um, game engine objects, um, that, that the, uh, the game... Uh, the AIs in the game could navigate around.

    29. LFLex Fridman1:14:49

        Mm-hmm.

    30. NSNeal Stephenson1:14:49

        Um, so, um, and that ended up shipping as more of a dev kit or a sort of how to, a sample app-
11. 1:30:58 – 1:45:57

    Artificial intelligence

    1. NSNeal Stephenson1:30:58

    2. LFLex Fridman1:30:58

       Let me ask you a little bit about AI.

    3. NSNeal Stephenson1:31:01

       Okay.

    4. LFLex Fridman1:31:01

       Uh, what are some likely interesting trajectories for the proliferation of AI in society over the next couple of decades? Do you think about this kind of stuff?

    5. NSNeal Stephenson1:31:15

       ... do not think about it a lot 'cause it's a deep topic and I'm not, I don't consider myself super well-informed ab- about it. And AI seems to be a term that is applied to a lot of different things. So, I've messed around just a tiny little bit with, with neural nets, with, uh, what's it called? PCA, Principal Component Analysis. So, I, I guess I tend to think in terms of sort of granular bottom-up, um, ideas, rather than big picture top-down, you know?

    6. LFLex Fridman1:31:49

       Oh, got it. So, like, very specific algorithms, like, how are they going to... W- what problem are they, are they going to solve in society such that, that has, like, a lot of big ripple effects? See, I, I mean, we could talk a, a, a particular successful AI systems and success defined in different ways of recent years. So, one is language models with GPT-3. Uh, most importantly, they're self-supervised, meaning they don't require much supervision from humans, which means they can learn by just reading a huge amount of content created by humans. So, read the internet, and from that, be able to generate text and do all kinds of things like that.

    7. NSNeal Stephenson1:32:31

       Mm-hmm.

    8. LFLex Fridman1:32:31

       It's possible they have a big enough neural network that's going to be able to have conversations with humans based on just reading human language. That's an interesting idea.

    9. NSNeal Stephenson1:32:43

       Mm-hmm.

    10. LFLex Fridman1:32:43

        To me, the very interesting idea that people don't think about it as AI because it's, they're kind of dumb currently, is actual embodied robots. So, robotics-

    11. NSNeal Stephenson1:32:54

        Mm-hmm.

    12. LFLex Fridman1:32:54

        ... like Boston Dynamics. I have downstairs and upstairs, uh, legged robots.

    13. NSNeal Stephenson1:32:59

        Oh.

    14. LFLex Fridman1:33:00

        Uh, you know, uh, the currently Boston Dynamics robots and most legged robots, most robots period, are pretty dumb. So, s- most of the challenges have to do with the actual, first of all, the engineering of making the thing work.

    15. NSNeal Stephenson1:33:17

        Mm-hmm.

    16. LFLex Fridman1:33:17

        Getting, uh, a sensor suite that allows you to do, it's the same thing as with Magic Leap-

    17. NSNeal Stephenson1:33:21

        Yeah.

    18. LFLex Fridman1:33:21

        ... that base layer of, like-

    19. NSNeal Stephenson1:33:22

        Where is it stuff.

    20. LFLex Fridman1:33:23

        ... where am I-

    21. NSNeal Stephenson1:33:24

        Yeah.

    22. LFLex Fridman1:33:25

        ... and, uh, what, what am I looking at? (laughs)

    23. NSNeal Stephenson1:33:29

        Yeah.

    24. LFLex Fridman1:33:29

        I don't need to deeply understand, uh, my surroundings at a level of, like, like, uh, at a level beyond of what will hurt if I run into it?

    25. NSNeal Stephenson1:33:40

        Yeah, yeah.

    26. LFLex Fridman1:33:41

        Yeah.

    27. NSNeal Stephenson1:33:41

        But even that is hard.

    28. LFLex Fridman1:33:43

        That's, that's hard, but the thing that I think people don't, uh, in, in the robotics base explore enough is the human/robot interaction part of the, of the picture, which is how it makes humans feel, how robots make humans feel.

    29. NSNeal Stephenson1:34:01

        Mm-hmm, mm-hmm.

    30. LFLex Fridman1:34:01

        And I think that's going to have a very significant impact in, uh, in the near future in society, which is wh- the more you integrate AI systems of whatever form into society where humans are, uh, in contact with them regularly, so that could be embodied robotics or that could be social media algorithms, I think that has a very significant impact. And people often think, like, AI needs to be super smart to have an impact. I think it needs to be super integrated with society to have an impact. And more and more, that's happening, even if they're dumb.

Episode duration: 2:39:51