Latest / What's Your Problem? / How Bubbles Power Breakthroughs

Transcript

How Bubbles Power Breakthroughs · What's Your Problem?

0:15
Speaker 1: Pushkin. There are these moments when people make huge technical
0:23
Speaker 1: advances and it happens all of a sudden, or at
0:26
Speaker 1: least it feels like it happens all of a sudden.
0:29
Speaker 1: You know this is happening right now, most obviously with AI,
0:33
Speaker 1: with artificial intelligence. It happened not too long ago with rockets,
0:37
Speaker 1: when SpaceX dramatically lowered the cost of getting to space.
0:42
Speaker 1: Maybe it's happening now with crypto. I'd say it's probably
0:45
Speaker 1: too soon to say on that one. In any case,
0:49
Speaker 1: you can look at technological breakthroughs in different fields and
0:52
Speaker 1: at different times, and you can ask, what can we
0:56
Speaker 1: learn from these? You can ask, can we abstract certain
1:00
Speaker 1: you know, certain qualities, certain tendencies that seem to drive
1:04
Speaker 1: toward these burths of technological progress. There's a recent book
1:09
Speaker 1: called Boom that asked this question and it comes up
1:12
Speaker 1: with an interesting answer. According to the book, one thing
1:16
Speaker 1: that's really helpful if you want to make a wild
1:19
Speaker 1: technological leap a bubble. I'm Jacob Goldstein, and this is
1:29
Speaker 1: what's your problem. My guest today is Burne Hobart. He's
1:33
Speaker 1: the author of a finance newsletter called The diff and
1:36
Speaker 1: he's also the co author of a book called Boom,
1:39
Speaker 1: Bubbles and the End of Stagnation. When burn talks about bubbles,
1:44
Speaker 1: he isn't just talking about financial bubbles where investors drive
1:47
Speaker 1: prices through the roof. When he says bubble, largely he
1:51
Speaker 1: means social bubbles, filter bubbles, little groups of people who
1:56
Speaker 1: share some wild belief. He really gets it what he
1:59
Speaker 1: means in this one sentence where he and his co
2:01
Speaker 1: author write, quote, transformative progress arises from small groups with
2:07
Speaker 1: a unified vision, vast funding, and surprisingly poor accountability end quote,
2:14
Speaker 1: basically living the dream. Later in the conversation, Burne and
2:19
Speaker 1: I discussed the modern space industry and cryptocurrency and AI.
2:23
Speaker 1: But to start, we talked about two case studies from
2:26
Speaker 1: the US in the twentieth century. Burn writes about him
2:29
Speaker 1: in the book, and he argues that these two moments
2:31
Speaker 1: hold broader lessons for how technological progress works. The two
2:36
Speaker 1: case studies we talk about are the Manhattan Project and
2:38
Speaker 1: the Apollo Missions. So let's start with the Manhattan Project,
2:43
Speaker 1: and maybe one place to start it is with this
2:46
Speaker 1: famous nineteen thirty nine letter from Albert Einstein and other
2:50
Speaker 1: scientists to FDR the President about the possibility of the
2:55
Speaker 1: nazis building an.
2:56
Speaker 2: Atomic bomb, right, So that letter it is it feels
3:02
Speaker 2: like good material for maybe not a complete musical comedy,
3:04
Speaker 2: but at least an act of a musical comedy, because
3:06
Speaker 2: it's kind.
3:07
Speaker 1: Of a springtime for Hitler in Germany.
3:10
Speaker 2: There is this whole thing where you have this brilliant physicist,
3:13
Speaker 2: but he is just kind of the stereotypical professor, you know,
3:17
Speaker 2: crazy hair, very absent minded, always talking about these things
3:19
Speaker 2: where no one, you know, no normal person can really
3:22
Speaker 2: understand what he's talking about. And suddenly, instead of talking
3:27
Speaker 2: about space time and that you know, energy and matter
3:30
Speaker 2: in their relationship, suddenly he's saying someone could build a
3:33
Speaker 2: really really big bomb and that person will probably be
3:36
Speaker 2: a German, and that has some very bad implications for
3:40
Speaker 2: the rest of the world.
3:41
Speaker 1: So now here we are, the president decides, okay, we
3:43
Speaker 1: need to build a bomb, we need to spend a
3:46
Speaker 1: wild amount of money on it. And this is a
3:48
Speaker 1: thing that you describe as a bubble, which is interesting, right,
3:50
Speaker 1: because it's not a bubble in the sense of market prices.
3:53
Speaker 1: It's the federal government in the military, but it has
3:56
Speaker 1: these other bubble like characteristics in your telling, right, maybe
4:02
Speaker 1: other meanings of bubble, the way we talk about a
4:03
Speaker 1: social bubble or a filter bubble. Tell me about that, like,
4:06
Speaker 1: why is the Manhattan Project a kind of bubble?
4:09
Speaker 2: Why is it a bubble? Because there's that feedback loop
4:12
Speaker 2: because people take the idea of actually building the bomb
4:16
Speaker 2: more seriously as other people take it more seriously, and
4:19
Speaker 2: the more that you have someone like more that you
4:22
Speaker 2: have people like Oppenheimer actually dedicating time to the projects,
4:25
Speaker 2: the more other people think the project will actually happen,
4:27
Speaker 2: this is actually worth doing. So you have this group
4:29
Speaker 2: of people who start taking the idea of building a
4:32
Speaker 2: bomb more seriously. They treat it as a thing that
4:35
Speaker 2: will actually happen rather than a thing that is hypothetically
4:38
Speaker 2: possible if if this particular equation is right, if these
4:40
Speaker 2: measurements are right, et cetera, and then they start actually
4:42
Speaker 2: designing them.
4:43
Speaker 1: The man Hen Project seems like this sort of point
4:47
Speaker 1: that gets a bunch of really smart people to coalesce
4:49
Speaker 1: in one place on one project at one time. Right.
4:52
Speaker 1: It sort of solves the coordination problem the way whatever
4:55
Speaker 1: you might say AI today is doing that, Like just
4:59
Speaker 1: brilliant people suddenly are all in one place working on
5:02
Speaker 1: the same thing in a way that they absolutely would
5:04
Speaker 1: not otherwise be.
5:05
Speaker 2: That is true. And this was both within the US
5:08
Speaker 2: academic community and then within the global academic community, because
5:10
Speaker 2: you had a lot of people who were in Central
5:13
Speaker 2: Europe or Eastern Europe who realized that that is just
5:15
Speaker 2: not a great place for them to be and tried
5:18
Speaker 2: to get to the UK, US other Allied countries as
5:23
Speaker 2: as quickly as possible. And so there was just this
5:26
Speaker 2: massive intellectual dividend of a lot of the most brilliant
5:29
Speaker 2: people in Germany and in Eastern Europe and in Hungary, etcetera.
5:34
Speaker 2: They were all fleeing and all ended up in the
5:37
Speaker 2: same country. So, yeah, you have just this this serendipity
5:40
Speaker 2: machine where it was just if you were a physicist,
5:44
Speaker 2: it was an incredible place for just overhearing really novel
5:47
Speaker 2: ideas and putting those ideas, putting your own ideas to
5:50
Speaker 2: the test, because you had all the all the smartest
5:53
Speaker 2: people in the world pretty much in this one little
5:55
Speaker 2: town in New Mexico.
5:57
Speaker 1: Right. So the Los Alamos piece is is the famous part,
6:01
Speaker 1: you know, it's the part one has heard of with
6:05
Speaker 1: respect to the Manhattan Project. There's a less famous part
6:07
Speaker 1: that's really interesting, and that also seems to hold some
6:09
Speaker 1: broader lessons as well, right, And that is the kind
6:12
Speaker 1: of basically the manufacturing part of the how do we
6:15
Speaker 1: enrich enough uranium to build the bomb? If the physicists
6:19
Speaker 1: figure out how to design it, talk about that piece
6:21
Speaker 1: and the lessons there.
6:23
Speaker 2: Yes, so that one, You're right, It is often under
6:26
Speaker 2: emphasized in the history. It was more of an engineering
6:30
Speaker 2: project than a research project, though there was a lot
6:31
Speaker 2: of research involved. The purpose was get enough of the
6:35
Speaker 2: get enough in rich uranium so the isotope that is
6:39
Speaker 2: actually prone to these chain reactions, get it isolated, and
6:44
Speaker 2: then be able to incorporate that into a bomb. They
6:48
Speaker 2: were also working on other physom materials because there were
6:52
Speaker 2: multiple plausible bomb designs. Some use different triggering mechanisms, some
6:56
Speaker 2: use different materials, and there were also multiple plausible ways
7:00
Speaker 2: to enrich enough of the physico material to actually build
7:03
Speaker 2: a bomb. And so one version of the story is
7:07
Speaker 2: you just go down the list and you the one
7:09
Speaker 2: that you think is the most cost effective, most likely,
7:12
Speaker 2: and so we choose one way to get just you
7:16
Speaker 2: two thirty five, and we have one way to build bomb.
7:18
Speaker 1: You two thirty five is the enriched uranium. Yes, And that,
7:21
Speaker 1: by the way, is the way normal businesses do things
7:24
Speaker 1: in normal times. You're like, well, we got to do
7:26
Speaker 1: this really expensive thing. We got to build a factory,
7:28
Speaker 1: and we don't even know if it's going to work.
7:29
Speaker 1: Let's choose the version that's most likely to work. Like
7:32
Speaker 1: that is the kind of standard move right, yeah, right?
7:35
Speaker 1: And then the problem though, is that if you try
7:38
Speaker 1: that and you just you got unlucky. You kept the
7:41
Speaker 1: wrong bomb design and the right phissile material or right
7:44
Speaker 1: material of wrong bomb design, you've done a lot of
7:46
Speaker 1: work which has zero payoff, and you've lost time. Right,
7:49
Speaker 1: Like crucially, there is there is a huge sense of
7:52
Speaker 1: urgency present at this moment that is driving the whole
7:56
Speaker 1: thing really right.
7:57
Speaker 2: We could also do more than one of them in parallel,
8:00
Speaker 2: and that is what we did. And on the manufacturing
8:02
Speaker 2: side that was actually just murderously expensive. If you are
8:05
Speaker 2: building a factory and you build the wrong kind of factory,
8:08
Speaker 2: then you you've wasted a lot of a lot of
8:10
Speaker 2: money and effort and time. So they just they did
8:13
Speaker 2: more than one. They did several different processes for enriching
8:17
Speaker 2: uranium and for perstonium.
8:19
Speaker 1: All at the same time, right, And they knew they
8:21
Speaker 1: weren't going to use all of them, they just didn't
8:23
Speaker 1: know which one was going to work. So it's like, well,
8:25
Speaker 1: let's try all of them at the same time and
8:27
Speaker 1: hopefully one of them will work. Yes, Like that is
8:30
Speaker 1: super bubbly, right. That is wild and expensive. That is
8:33
Speaker 1: just throwing wild amounts of money at something in a
8:37
Speaker 1: great amount of haste.
8:38
Speaker 2: Yes. Yeah, And if you so, if you believe that
8:42
Speaker 2: there's this pretty linear payoff, then every additional investment you make,
8:46
Speaker 2: you know has it doesn't qualitatively change things. It just
8:49
Speaker 2: means you're doing a little bit more of it. But
8:50
Speaker 2: if you believe there's some kind of nonlinear payoff where
8:52
Speaker 2: either this facility doesn't basically doesn't work at all or
8:56
Speaker 2: it works really really well, then when you when you
8:59
Speaker 2: diversify a little bit, you do actually get just this
9:02
Speaker 2: this better risk adjusted return even though you're objectively taking
9:05
Speaker 2: more risks.
9:06
Speaker 1: Interesting, right, So in this instance, it's if the Nazis
9:09
Speaker 1: have the bomb before we do, it's the end of
9:12
Speaker 1: the world as we know it, Yes, and so we
9:14
Speaker 1: better take a lot of risk. And that's actually rational.
9:17
Speaker 1: It reminds me a little bit of aspects of operation
9:22
Speaker 1: warp speed. I remember talking to Susan Athea, Stanford economist
9:26
Speaker 1: early in the pandemic who was making the case to
9:28
Speaker 1: do exactly this with vaccine manufacturing in like, you know,
9:32
Speaker 1: early twenty twenty, we didn't know if any vaccine was
9:34
Speaker 1: going to work, and it takes a long time to
9:36
Speaker 1: build a factory to make a vaccine, basically a tailor
9:39
Speaker 1: of factory. And she was like, just make a bunch
9:41
Speaker 1: of factories to make vaccines because if one of them work,
9:43
Speaker 1: is we want to be able to start working on
9:45
Speaker 1: it that day. Like that seems quite similar to this
9:48
Speaker 1: and you work.
9:49
Speaker 2: Yeah, yeah, I think that's absolutely true that you you know,
9:53
Speaker 2: the higher the stakes are, the more you want to
9:55
Speaker 2: be running everything that can plausibly help in parallel. And
10:00
Speaker 2: depending on the exact nature of what you're doing, there
10:01
Speaker 2: can be some spillover effects. It's you know, it's possible
10:04
Speaker 2: that you build a factory for manufacturing vaccine A and
10:08
Speaker 2: vaccine doesn't work out, but you can retrofit that factory
10:11
Speaker 2: and start doing vaccin B. And you know, there are
10:14
Speaker 2: little ways to shuffle things around a bit, but you
10:16
Speaker 2: often want to go into this basically telling yourself, if
10:19
Speaker 2: we didn't waste money and we still got a good outcome,
10:22
Speaker 2: it's because we got very, very lucky, and that we
10:24
Speaker 2: only know we're being serious if we did, in fact
10:26
Speaker 2: waste a lot of money. And yeah, I think that
10:28
Speaker 2: kind of inverting your view of risk is often a
10:30
Speaker 2: really good way to think about these big transformative changes.
10:34
Speaker 2: And this is actually another case where the financial metaphors
10:37
Speaker 2: do give useful information about just real world behaviors because
10:41
Speaker 2: at at head funds, this is actually something that risk
10:43
Speaker 2: teams will sometimes teleportfolio managers. Is you are making money
10:47
Speaker 2: on two higher percentage of your trades. This means that
10:49
Speaker 2: you are not making all the trades that you could
10:51
Speaker 2: and if you made, if you took your hit rate
10:54
Speaker 2: from fifty five percent down to fifty three percent, we'd
10:56
Speaker 2: be able to allocate more capital to you, even though
10:59
Speaker 2: you'd be annoyed that you were losing money on more trades.
11:01
Speaker 1: Interesting because overall you would likely have a more profitable
11:05
Speaker 1: outcome by taking bigger risks and incurring a few more losses,
11:08
Speaker 1: but winds would be bigger and make up for the losses.
11:11
Speaker 2: Yes, and this kind of thing, you know, it's very
11:13
Speaker 2: easy if you're the one sitting behind the desk just
11:15
Speaker 2: talking about these relative trade offs. It's a lot harder
11:16
Speaker 2: if you are the first person working with uranium in
11:19
Speaker 2: the factory and we don't quite know what the risks
11:21
Speaker 2: of that are, but it is. It is just a
11:23
Speaker 2: it's a generally true thing about trade offs that if
11:26
Speaker 2: you about trade offs and risk, that there is an
11:28
Speaker 2: optimal amount of risk to take. That optimal amount is
11:31
Speaker 2: sometimes dependent on what the what the downside risk of
11:34
Speaker 2: inaction is. And so sometimes if you're if you're too successful,
11:38
Speaker 2: you realize that you are actually messing something up.
11:41
Speaker 1: Yeah, you're not taking enough risk. So we all know
11:44
Speaker 1: how the Manhattan Project ends. It worked. I mean, it
11:49
Speaker 1: is a little bit of a weird one to start with.
11:51
Speaker 1: You know, the basic ideas like technological progress is good,
11:54
Speaker 1: risks are good, and we're talking about building the atomic
11:56
Speaker 1: bomb and dropping it on two cities and it's you know,
12:01
Speaker 1: it's morally a much easier question if you think it's
12:03
Speaker 1: the Nazis. Sorry, but the Nazis are absolutely the worst,
12:06
Speaker 1: and I definitely don't want them to have a bomb first.
12:09
Speaker 1: You know, there is the argument that more people would
12:11
Speaker 1: have died in a conventional invasion without the bomb. I don't,
12:14
Speaker 1: I don't know. I mean, how do you what do
12:19
Speaker 1: you make of it? Like, obviously the book is very
12:21
Speaker 1: pro technological progress, this show is basically pro technological progress.
12:25
Speaker 1: But like the bomb isn't isn't a happy one to start?
12:28
Speaker 1: I'm like, what do you make of it?
12:29
Speaker 2: Ultimately, Yeah, it's one of those things where you do
12:32
Speaker 2: it does make me wish that we could we could
12:34
Speaker 2: run the same simulation, you know, a couple of million
12:36
Speaker 2: times and see what the what the net? You know,
12:39
Speaker 2: loss and save lives are different scenarios, but one thing
12:43
Speaker 2: the bomb. If you I guess from like a purely
12:47
Speaker 2: utilitarian standpoint, I suspect that there have been net lives
12:52
Speaker 2: save because of less use of coal for electricity generation,
12:57
Speaker 2: more use of nuclear power, And that is directly downstream
12:59
Speaker 2: of the bomb. That you can build these, you know,
13:02
Speaker 2: buy design uncontrollable releases of atomic energy. You can also
13:07
Speaker 2: build more controllable ones, and then getting the funding for
13:10
Speaker 2: that would be a lot harder.
13:11
Speaker 1: And presumably we got nuclear power much sooner than we
13:14
Speaker 1: otherwise would have because the incredibly rapid project progress of
13:18
Speaker 1: the Manhattan Project. That's yes, there fair.
13:22
Speaker 2: Which I don't I don't think you know you if
13:24
Speaker 2: you let me push the button on what I drop
13:27
Speaker 2: an atomic bomb on a civilian population in exchange for
13:32
Speaker 2: fewer people dying of respiratory diseases over the next couple decades,
13:35
Speaker 2: you know, there, I would have to give it a
13:37
Speaker 2: lot of thought.
13:38
Speaker 1: I'm not going to I'm not going to push that button,
13:41
Speaker 1: but I'm never going to have a job where I
13:42
Speaker 1: have to decide because I can't deal. Okay. Then you
13:47
Speaker 1: mentioned in the book kind of in passing that was
13:50
Speaker 1: really interesting and surprising to me was that nuclear power
13:54
Speaker 1: today accounts for eighteen percent of electric power generation in
14:00
Speaker 1: the US. Eighteen percent, Like that is so much higher
14:03
Speaker 1: than I would have thought, given sort of how little
14:06
Speaker 1: you hear about existing nuclear power plants, right.
14:08
Speaker 2: Like, that is a lot. Yeah, yeah, it is. It
14:13
Speaker 2: is a surprisingly it's a president hand number. But also
14:16
Speaker 2: nuclear power it is it is one of the most
14:18
Speaker 2: annoying technologies to talk about, in the sense that it
14:21
Speaker 2: doesn't do anything really really exciting other than provide essentially
14:25
Speaker 2: unlimited power with minimal risk.
14:28
Speaker 1: And some amount of some amount of scary tail risk,
14:34
Speaker 1: right ye, Like, I mean that is what is actually
14:36
Speaker 1: interesting to talk about. Sort of unfortunately for the world,
14:39
Speaker 1: given that it has a lot of benefits, there is
14:41
Speaker 1: this tail risk, and once in a while something goes
14:44
Speaker 1: horribly wrong, even though on the whole it seems to
14:47
Speaker 1: be clearly less risky than say, a caul fired power.
14:50
Speaker 2: Plant, right, And the industry has they they're aware of
14:54
Speaker 2: those risks, and nobody wants to be responsible for that
14:58
Speaker 2: kind of thing, and nobody wants to be testifying before
15:00
Speaker 2: Congress about ever having cut any corner whatsoever in the
15:04
Speaker 2: event that a disaster happens. So they do actually take
15:07
Speaker 2: that incredibly seriously. So nuclear power does end up being
15:11
Speaker 2: in practice much safer than other power sources. And then
15:15
Speaker 2: you add in the externality of doesn't really produce emissions,
15:18
Speaker 2: and uranium exists in some quantities just about everywhere.
15:23
Speaker 1: No climate change, no local air pollution, has a lot
15:25
Speaker 1: going for it. Always on. Okay, let's go to the moon.
15:31
Speaker 1: So you write also about the Apollo missions us going
15:36
Speaker 1: to the moon. It's the early sixties, right, was it
15:40
Speaker 1: sixty one? Kennedy says we're gonna go to the moon
15:44
Speaker 1: by the end of the decade. There's the Cold War context. Yes,
15:48
Speaker 1: Kennedy announces this goal. What's the response in the US
15:52
Speaker 1: when Kennedy says this, Yeah, so a lot of the response,
15:55
Speaker 1: you know, at first people are somewhat hypothetically excited.
15:59
Speaker 2: As they start realizing how much it will cost, they
16:01
Speaker 2: go from not especially excited to actually pretty deeply opposed.
16:06
Speaker 2: And you know this shows up in there was a
16:09
Speaker 2: someone coined the term moondoggle.
16:10
Speaker 1: Yeah, moondoggle. I loved moondoggle. I learned that from the book.
16:14
Speaker 1: It was Norbert Wiener, like a famous technologist, not a
16:17
Speaker 1: not a not a crank, right, somebody I knew what
16:20
Speaker 1: he was talking about. Was like, this is a crazy idea.
16:23
Speaker 2: It's a moondoggle, right, And you know, this really works
16:26
Speaker 2: its way to popular culture. Like if you if you
16:28
Speaker 2: go on Spotify and listen to the Tom learror song
16:32
Speaker 2: Verner von Braun, the recording that Spotify has, it opens
16:35
Speaker 2: with a monologue that is talking about how stupid the
16:38
Speaker 2: idea of the Apollo program is. It's a it and
16:42
Speaker 2: you know this is a This is again someone who
16:43
Speaker 2: is in academia, who's a very very sharp guy and
16:47
Speaker 2: who is just feels like he completely sees through this
16:50
Speaker 2: political giveaway program to big defense contractors and knows that
16:54
Speaker 2: there's no point in doing this.
16:56
Speaker 1: You write that NASA's own analysis found a ninety percent
17:01
Speaker 1: chance that that a failure of failing to reach the
17:05
Speaker 1: moon by the end of the decade. Like, it wasn't
17:07
Speaker 1: just outside people being critical, it was NASA itself didn't think.
17:10
Speaker 2: It was good work.
17:12
Speaker 1: There's a phrase you use in the book to talk
17:15
Speaker 1: about talk about these sort of bubble like environments that
17:18
Speaker 1: are of interest to you, and I found it really
17:21
Speaker 1: interesting and I think we can talk about it in
17:23
Speaker 1: the context of Apollo. That phrase is definite optimism. Tell
17:27
Speaker 1: me about that phrase.
17:28
Speaker 2: Yes, So definite optimism is the view that the future
17:32
Speaker 2: can and will be better in some very specific way,
17:36
Speaker 2: that there will be there is something we cannot do
17:39
Speaker 2: now we will be able to do in the future,
17:40
Speaker 2: and it will be good that we can do it.
17:42
Speaker 1: And why is it important, Like it's a big deal
17:45
Speaker 1: in your telling in an interesting way. Why is it
17:48
Speaker 1: so important?
17:50
Speaker 2: It's important because that is what allows you to actually
17:53
Speaker 2: marshal those resources, whether those are the people or the
17:57
Speaker 2: capital or the political pull, to put them all in
18:00
Speaker 2: some specific direction and say we're going to build this thing.
18:03
Speaker 2: So we need to actually go step by step and
18:05
Speaker 2: figure out, Okay, what specific things have to be done,
18:08
Speaker 2: what discoveries have have to be made, what laws have
18:10
Speaker 2: to be passed in order for this to happen. And
18:13
Speaker 2: that is so it's definitely optimism in the sense that
18:16
Speaker 2: you're saying, there is a specific thing we're going to build.
18:18
Speaker 2: It's the kind of thing that can keep you going
18:20
Speaker 2: when you encounter temporary setbacks. And that's that's where the
18:23
Speaker 2: optimism part comes in, because if you have a less
18:28
Speaker 2: definitely optimistic view of about that project, you might say
18:31
Speaker 2: the goal of the Apollo program is to figure out
18:33
Speaker 2: if we can put a person on the moon. But
18:36
Speaker 2: I think what that leaves you open to is the
18:38
Speaker 2: temptation to give up at any point, because at any
18:41
Speaker 2: point you can have you know, a botched launch or
18:45
Speaker 2: an accident, or you are you're designing some component and
18:48
Speaker 2: the math just doesn't pencil out. You need you know
18:51
Speaker 2: it's going to weigh too much to actually make it
18:53
Speaker 2: onto craft. And you could say, okay, well that's how
18:56
Speaker 2: we figured out that we're not actually doing this. But
18:58
Speaker 2: if you do just have this kind of delusional view
19:01
Speaker 2: that know that if there's a mistake, it's a mistake
19:04
Speaker 2: in my analysis, not in the ultimate plan here, and
19:07
Speaker 2: that it is physically possible, we just have to figure
19:08
Speaker 2: out the all the details. Then I think that does
19:11
Speaker 2: set up a different kind of motivation, because at that
19:12
Speaker 2: point you can view every mistake as just exhausting the
19:16
Speaker 2: set of possibilities and letting you narrow things down to
19:19
Speaker 2: what is the correct approach. What you sort of needed
19:22
Speaker 2: was this sort of very localized definite optimism where someone
19:27
Speaker 2: could you could imagine a researcher thinking to themselves, or
19:30
Speaker 2: an engineer or someone throughout the project thinking to themselves that, Okay,
19:34
Speaker 2: this will probably not work overall, but the specific thing
19:38
Speaker 2: I'm working on, whether it is designing a spacesuit or
19:41
Speaker 2: designing this rocket or programming the guidance computer, that one
19:45
Speaker 2: I can tell that my part is actually going to work,
19:48
Speaker 2: or at least I believe that I can make it work.
19:49
Speaker 2: And two, this is my only chance to work with
19:51
Speaker 2: these really cool toys. So if the money is going
19:54
Speaker 2: to be wasted at some point, let that money be
19:56
Speaker 2: wasted on me. And I think that that kind of
19:59
Speaker 2: attitude of just you know that you have one shot
20:01
Speaker 2: to actually do something really interesting, you will not get
20:04
Speaker 2: a second chance. If everyone believes that, it does become
20:07
Speaker 2: a coordinating mechanism where now they're working extremely hard, they
20:10
Speaker 2: all recognize that the success of what they are doing
20:13
Speaker 2: is very much up to them, and then that ends
20:16
Speaker 2: up contributing to this group success.
20:18
Speaker 1: So it's like this, if I'm going to do this,
20:22
Speaker 1: I got to do it now. Everybody's doing it. Now,
20:24
Speaker 1: we got the money. Now this is our one shot.
20:27
Speaker 1: We better get it right, We better do everything we
20:29
Speaker 1: can to make it work.
20:30
Speaker 2: Yes, fear of missing out?
20:33
Speaker 1: Yeah, fomo right, so fomo it's funny. People talk about
20:35
Speaker 1: that as like a dumb investment thesis basically right, it's
20:39
Speaker 1: like a meme stock idea, but you talk about it
20:41
Speaker 1: in these more interesting context basically right, more meaningful. I
20:46
Speaker 1: would say, yes, yeah.
20:48
Speaker 2: That it's so in the purely straightforward way. The idea
20:52
Speaker 2: is there are sometimes these very time limited opportunities to
20:55
Speaker 2: do something, and if you're capable of doing that thing,
20:58
Speaker 2: this may be your only chance, and so missing out
21:00
Speaker 2: is actually something you should be afraid of. So you know,
21:02
Speaker 2: if you actually have a really clever idea for an
21:06
Speaker 2: AI company, this is actually a time where you can
21:08
Speaker 2: at least the attempt. So yeah, we do argue that
21:11
Speaker 2: missing out is something you should absolutely fear.
21:13
Speaker 1: So what happens with the Apollo project, just even brief
21:16
Speaker 1: like talk about just how big it is and how
21:19
Speaker 1: risky it is, like it's it's striking right.
21:22
Speaker 2: Right, Yeah, so it is. It was running. The expenses
21:25
Speaker 2: were running on like a low single digital percentage of
21:27
Speaker 2: GDP for a while.
21:29
Speaker 1: So a couple percent of the value of everything everybody
21:33
Speaker 1: in the country does is going into the Apollo mission.
21:36
Speaker 1: Just this one plainly unnecessary thing that the government has
21:41
Speaker 1: decided to do.
21:42
Speaker 2: Right, And this is one of the cases where there
21:44
Speaker 2: was a very powerful spillover effects because the Apollo Guidance
21:49
Speaker 2: computer needed the most lightweight and least power consuming and
21:54
Speaker 2: most reliable components possible. And if you were building a
21:57
Speaker 2: computer conventionally at that time and you had a budget,
22:00
Speaker 2: you would probably build it out of vacum tubes. And
22:02
Speaker 2: you knew that the vacuum tubes they're bulky, they consume
22:05
Speaker 2: a lot of power, they throw off a lot of heat,
22:06
Speaker 2: they burn out all the time, but they are fairly cheap.
22:10
Speaker 2: But in this case, there there was an alternative technology.
22:15
Speaker 2: It was extremely expensive, but it was it was lightweight,
22:20
Speaker 2: didn't use a lot of power, and did not have
22:22
Speaker 2: moving parts. And that's the integrated circuit, so transistor based computing.
22:26
Speaker 1: The chip, well we know today as the chip.
22:29
Speaker 2: Yes, the chip.
22:30
Speaker 1: You read that. In nineteen sixty three, NASA bought sixty
22:34
Speaker 1: percent of the chips made in the United States, just NASA,
22:38
Speaker 1: not though government, just NASA sixty percent.
22:41
Speaker 2: They actually bought more chips than they needed because they
22:44
Speaker 2: recognized that the chip companies were run by, you know,
22:48
Speaker 2: very very nice electrical engineering nerds who just love designing tiny,
22:51
Speaker 2: tiny things, and that these people just don't know how
22:54
Speaker 2: to run a business, and so they were worried that
22:57
Speaker 2: fair Child Semiconductor would just run out of cash at
23:00
Speaker 2: some point and then NASA would have half of a
23:03
Speaker 2: computer and no way to build the rest of it.
23:05
Speaker 2: So they actually over ordered the US. They used integrated
23:08
Speaker 2: circuits for a few applications that actually were not so
23:10
Speaker 2: dependent on the power consumption and weight and things. So
23:13
Speaker 2: that critique of the Apollo program was directionally correct. It
23:17
Speaker 2: was money being splashed out to defense contractors who were
23:19
Speaker 2: favored by the government, but in this case it was
23:21
Speaker 2: being done in a more strategic and thoughtful way and
23:25
Speaker 2: kind of kept the industry going.
23:26
Speaker 1: So you talk a fair bit in the book about
23:31
Speaker 1: the sort of religious and quasi religious aspects of these
23:35
Speaker 1: little groups of people that come together in these bubble
23:38
Speaker 1: like moments to do these big things, and that's really
23:42
Speaker 1: present in the Apollo section, like talk about the sort
23:46
Speaker 1: of religious ideas associated with the Apollo mission that the
23:51
Speaker 1: people working on the mission had.
23:52
Speaker 2: Yeah, I mean, you name it after a Greek god
23:56
Speaker 2: and you're already starting starting a little bit religious. So
24:00
Speaker 2: there were people who worked on these missions who felt like,
24:03
Speaker 2: this is part of mankind's destiny is to explore the stars,
24:07
Speaker 2: and that there's this whole universe that is universe created
24:09
Speaker 2: by God and it would be kind of weird. You know,
24:11
Speaker 2: we can't second guess the divine, but it's a little
24:14
Speaker 2: weird for God to create all of these astronomical bodies
24:16
Speaker 2: that just kind of look good from the ground and
24:18
Speaker 2: that you're not actually meant to go visit.
24:20
Speaker 1: You talk about somewhat similar things in other kind of
24:24
Speaker 1: less obviously spiritual dimensions of people coming together and having
24:29
Speaker 1: it kind of more than rational. You use this word
24:34
Speaker 1: thimos from the Greek meaning spirit, like what's going on there?
24:37
Speaker 1: More broadly, why is that important more generally for technological progress,
24:43
Speaker 1: because the so th amos is part of this tripartite
24:47
Speaker 1: model of the soul where you have your appetites.
24:50
Speaker 2: And your reason and then your themost like you're longing
24:53
Speaker 2: for glory and honor and this kind of transcendent transcendent
24:58
Speaker 2: achievement and logos reasoning it only gets you so far.
25:03
Speaker 2: You can you can reason your way into some pretty
25:05
Speaker 2: interesting things, but at some point you do decide that
25:08
Speaker 2: the reason thing is probably to take it a little
25:12
Speaker 2: bit easier and not take not take certain risks. And
25:15
Speaker 2: it still is just this this pursuit of something, something greater,
25:19
Speaker 2: and you know, something something beyond the ordinary, something really
25:22
Speaker 2: beyond the logos, right, like beyond what you could get
25:25
Speaker 2: to just by reasoning one step at a time. And
25:30
Speaker 2: I think that that is, Yeah, that is that is
25:32
Speaker 2: just a deeply attractive proposition to to many people. And
25:36
Speaker 2: it's also a scary one because at that point, you know,
25:38
Speaker 2: if you are if you're doing things that are beyond
25:40
Speaker 2: what is the rational thing to do, then of course
25:43
Speaker 2: you have no rational explanation for what you did wrong.
25:46
Speaker 2: If you mess up. Yeah, and you are sort of
25:48
Speaker 2: betting on some historical contingencies.
25:50
Speaker 1: That's the definite optimism part, right, Betting on historical contingencies
25:54
Speaker 1: is another way of saying definite optimism. Right. So back
26:00
Speaker 1: to the moon. So we get to the moon. In fact,
26:03
Speaker 1: against all odds, we make it. Uh, And there's this
26:08
Speaker 1: moment where it's like, you know, today the moon tomorrow
26:12
Speaker 1: the solar system, but in fact it was today the
26:15
Speaker 1: moon tomorrow, not even the.
26:18
Speaker 2: Moon, right, Like what happened? Well, you know, you had
26:22
Speaker 2: asked about what these megaprojects have in common with financial bubbles,
26:25
Speaker 2: and one of the things they have in common is
26:26
Speaker 2: sometimes there's a bust, and sometimes that bust is actually
26:29
Speaker 2: an overreaction in the opposite direction. And people people take
26:34
Speaker 2: everything every true, you know, everything they believed in say
26:38
Speaker 2: nineteen you know, nineteen sixty nine about humanity's future in
26:41
Speaker 2: the stars, and they say, okay, this is this is
26:43
Speaker 2: exactly the opposite of where things will actually go in
26:46
Speaker 2: the exact opposite of what we should care about. That
26:48
Speaker 2: we have plenty of problems here on Earth, and why
26:50
Speaker 2: would we you know, do we really want to turn
26:52
Speaker 2: Mars into just another planet? That also has problems of
26:56
Speaker 2: racism and poverty and nuclear war and all that stuff.
26:58
Speaker 2: So so maybe we should stay home and fix fix
27:00
Speaker 2: our stuff. Then public policy, you'd actually need for there
27:03
Speaker 2: to be some kind of resurgence in belief in space.
27:07
Speaker 2: You need some kind of charismatic story, and perhaps to
27:09
Speaker 2: an extent, we have that right now. Yes, maybe Elan's
27:13
Speaker 2: not the perfect front man for all of this, but
27:14
Speaker 2: he is certainly someone who demonstrates that space travel it
27:17
Speaker 2: can be done, it can be improved, and that it's
27:21
Speaker 2: just objectively cool. That it is just hard to watch
27:24
Speaker 2: a SpaceX launch video and not feel something.
27:28
Speaker 1: Yes, so so good. I want to talk more about
27:31
Speaker 1: space in a minute. So it's interesting these two stories
27:37
Speaker 1: that are kind of in the middle of your book.
27:39
Speaker 1: They're kind of the core of the book, right, these
27:40
Speaker 1: two interesting moments that are non financial bubbles, when you
27:44
Speaker 1: have this incredible technological innovation and a short amount of
27:47
Speaker 1: time that seems unrealistic, unrealistically you know, fast impressive outcome.
27:53
Speaker 1: And they're both pure government projects. They're both you know,
27:58
Speaker 1: command and control economy. It is not the private sector,
28:02
Speaker 1: it does not capitalism. What do you make of that?
28:06
Speaker 2: I would say, there's a very strong indirect link for
28:10
Speaker 2: a couple of reasons. One is just the practical, kind
28:15
Speaker 2: of the practically enough reason that personnelis policy, and that
28:19
Speaker 2: the US government in the nineteen thirties, the US government
28:23
Speaker 2: was hiring and the private sector mostly wasn't, and so
28:26
Speaker 2: all the ambitious people, basically all the ambitious people in
28:28
Speaker 2: the country tried to get government jobs. And that is
28:31
Speaker 2: usually not the case, and there are there's certainly circumstances
28:34
Speaker 2: where that's a really bad sign, but in this case
28:36
Speaker 2: it was great. It meant that there were a lot
28:38
Speaker 2: of New Deal projects that were staffed by the people
28:40
Speaker 2: who would have been rising up the ranks at RCA
28:43
Speaker 2: or General Electric or something a decade earlier. Now they're
28:46
Speaker 2: running New Deal projects instead, and they're again rising up
28:49
Speaker 2: their ranks really fast, having a very large real world
28:50
Speaker 2: impact for early in their careers. And those people had
28:53
Speaker 2: been working together for a while, and you know, they
28:56
Speaker 2: knew each other. There was a lot of just institutional
28:58
Speaker 2: knowledge about how to get big things done within the
29:01
Speaker 2: US government, and a lot of that institutional knowledge could
29:04
Speaker 2: then be redirected. So you have the New Deal and
29:07
Speaker 2: then the war effort, and then you have the post
29:09
Speaker 2: war economy where there's still you know, it takes a
29:11
Speaker 2: while for the government to fully relax its control, and
29:13
Speaker 2: then very soon we're into the Korean War. So yeah,
29:17
Speaker 2: there was just a large increase in state capacity and
29:20
Speaker 2: just in the quality of people making decisions within the
29:22
Speaker 2: US government in that period.
29:28
Speaker 1: We'll be back in a minute to talk about bubble
29:30
Speaker 1: esque things happening right now, namely rockets, cryptocurrency, and AI. Okay,
29:44
Speaker 1: now to space today. Burn and I talked about SpaceX
29:48
Speaker 1: in particular, because you know, it really is the company
29:51
Speaker 1: that launched the modern space industry, and there's this one
29:55
Speaker 1: key trait that SpaceX shares with the other projects Burn
29:58
Speaker 1: wrote about in the book. It brought together people who
30:01
Speaker 1: share a wild dream. If you go to work at SpaceX,
30:05
Speaker 1: it's probably because you believe in getting humanity to mar Yeah.
30:10
Speaker 2: Yeah, it's not just that you believe in the dream.
30:12
Speaker 2: But when you get the job, you're suddenly in an
30:15
Speaker 2: environment where everyone believes in the dream. And if you're
30:17
Speaker 2: working one of those organizations, you're probably not working nine
30:20
Speaker 2: to five, which means you have very few hours in
30:22
Speaker 2: your day or week where you are not completely surrounded
30:25
Speaker 2: by people who believe that humanity will people will be
30:29
Speaker 2: living on Mars, and that this is the organization that
30:32
Speaker 2: will make it happen. And that just has to really
30:34
Speaker 2: mess with your mind, Like, is what is normal to
30:38
Speaker 2: an engineer working at SpaceX in two thousand and six
30:40
Speaker 2: is completely abnormal to ninety nine point nine percent of
30:44
Speaker 2: the human population. And you know most of the exceptions
30:46
Speaker 2: are like six year old boys who just want Star
30:48
Speaker 2: Wars for the first time.
30:49
Speaker 1: Mars's crazy, Yeah, I mean really, as I went through
30:52
Speaker 1: the book, I was like, oh, really, the bubble you're
30:53
Speaker 1: talking about is a social bubble, like the meaningful bubble.
30:57
Speaker 1: Like maybe there's a financial bubble attachment, maybe there isn't,
30:59
Speaker 1: But what really matters is you're in this weird little
31:02
Speaker 1: social bubble that believes some wild thing together, that believes
31:07
Speaker 1: it is not wild, that believes it is going to happen,
31:09
Speaker 1: like that's the thing. Yeah, and has money and has
31:14
Speaker 1: the money to act on their wild belief yes, And
31:18
Speaker 1: so you know, getting the money does mean interacting with
31:21
Speaker 1: the normy sphere, interacting with people who don't.
31:23
Speaker 2: Quite buy into all of it, but they when you
31:26
Speaker 2: have these really ambitious plans and you're taking them seriously,
31:30
Speaker 2: you're doing them step by step. Some of those steps
31:33
Speaker 2: do have other practical applications, and so that is the
31:36
Speaker 2: space story. It was not just street shot. We are
31:38
Speaker 2: going to invest all the money Elon got from PayPal
31:41
Speaker 2: into going to Mars and hopefully we get to Mark
31:43
Speaker 2: before we run out. Yeah, it was you know, we're
31:46
Speaker 2: going to build these prototypes, We're going to build reusable rockets.
31:49
Speaker 2: We're going to use those for existing use cases, and
31:52
Speaker 2: we will probably find new use cases. And then once
31:55
Speaker 2: we get really really good at launching things cheaply, well,
31:57
Speaker 2: there are a lot of satellites out there, and perhaps
31:59
Speaker 2: we should have some of our own, and we can
32:01
Speaker 2: do it at sufficient scale, then maybe we can just
32:03
Speaker 2: throw a global communications network up there in the sky
32:06
Speaker 2: and see what happens next. So yeah, that's you know,
32:09
Speaker 2: the intermediate steps. Each one it's basically taking the themost
32:13
Speaker 2: like the spirited. You know, here's our grand vision of
32:16
Speaker 2: the future, and you know, here's my destiny and I
32:19
Speaker 2: was put on earth to do this and saying okay, well,
32:21
Speaker 2: the next step is have enough money to pay rent
32:23
Speaker 2: next month, right.
32:24
Speaker 1: Tomorrow to get to Mars. So is there a space
32:29
Speaker 1: bubble right now. I think so.
32:34
Speaker 2: I think there is I think there are people who
32:37
Speaker 2: look at SpaceX and say this is achievable, and that
32:41
Speaker 2: more is achievable. They also look at SpaceX and say
32:43
Speaker 2: this is a kind of infrastructure, that there are things
32:46
Speaker 2: like doing manufacturing in orbit or doing manufacturing on the Moon,
32:50
Speaker 2: where in some cases that is actually the best place
32:53
Speaker 2: to build something.
32:54
Speaker 1: Basically because SpaceX has driven down the cost so much
32:58
Speaker 1: of getting stuff into orbit, new ideas that would have
33:02
Speaker 1: been economically absurd twenty years ago, like manufacturing and space
33:06
Speaker 1: are now plausible. And so this is the sort of
33:08
Speaker 1: bubble building on itself, and like, why is it not
33:11
Speaker 1: just an industry now? Why is it a bubble?
33:13
Speaker 2: In your telling, it is the feedback loop where what
33:17
Speaker 2: SpaceX does makes more sense if they believe that there
33:20
Speaker 2: will be a lot of demand to move physical things
33:24
Speaker 2: off of Earth and into orbit, and perhaps further out
33:27
Speaker 2: that if they believe that there's more demand for that,
33:30
Speaker 2: they should be investing more on R and D. They
33:32
Speaker 2: should be building better and bigger and better rockets, and
33:35
Speaker 2: they should be doing the you know, big fixed cost
33:38
Speaker 2: investment that incrementally reduces the cost of launches and only
33:41
Speaker 2: pays for itself, you do a lot of them. And
33:42
Speaker 2: then if they're doing that and you have your dream
33:46
Speaker 2: of we're going to manufacture drugs in space and they
33:49
Speaker 2: will be you know, like the marginal cost is low
33:51
Speaker 2: once you get stuff up there, well that dream is
33:53
Speaker 2: a little bit more plausible if you can actually plot
33:56
Speaker 2: that curve of how much does it cost to get
33:57
Speaker 2: a kilogram into space and say this, you know, there
34:00
Speaker 2: is a specific ear at which point we would actually
34:03
Speaker 2: have the cost advantage versus terrestrial manufacturing.
34:06
Speaker 1: So it's this sort of coordinating mechan and that like
34:09
Speaker 1: you also write about with Microsoft and Intel in the
34:12
Speaker 1: eighties and nineties, where it's like, oh, they're building better chips,
34:15
Speaker 1: so we will build better software, and then because they're
34:18
Speaker 1: building better software, will build better chips. So this is
34:21
Speaker 1: like a more exciting version of that, right because it's
34:24
Speaker 1: going to get even cheaper to send stuff to space.
34:28
Speaker 1: We can build this crazy factory to exist in space,
34:31
Speaker 1: and then that tells SpaceX, oh, we can in fact
34:34
Speaker 1: keep building, keep innovating, keep spending money.
34:37
Speaker 2: Yes, And so someone has to do just half of that,
34:41
Speaker 2: like the half of that that makes no sense whatsoever.
34:43
Speaker 1: That was SpaceX at the beginning, right, that was like, yes,
34:46
Speaker 1: just a guy with a lot of money in a crazy.
34:49
Speaker 2: Yeah, it just really helps to have someone who's eccentric
34:52
Speaker 2: and it has a lot of money and is willing
34:54
Speaker 2: to throw it at a lot of different things. Like Musk.
34:56
Speaker 2: He spent some substantial fraction of his network right after
34:59
Speaker 2: this PayPal sale on a really nice sports car and
35:03
Speaker 2: then immediately took it for a drive and wrecked. It
35:06
Speaker 2: had no insurance and was not wearing a seat belt.
35:08
Speaker 2: So the Elon mus story could have just been this
35:10
Speaker 2: this proverb about dot com access and what happened when
35:14
Speaker 2: you finally gave these people money as they immediately bought
35:17
Speaker 2: sportscrossing Rerexiam. Instead, it's it's a story about a different
35:20
Speaker 2: kind of success, but it's still I guess know what
35:23
Speaker 2: that illustrates really risky.
35:25
Speaker 1: Yeah.
35:26
Speaker 2: Yeah, there's there's a risk level where you you are
35:31
Speaker 2: going on going for a joy ride in your two
35:33
Speaker 2: million dollar car and you haven't bothered to fill out
35:35
Speaker 2: all the paperwork or by the insurance, and that is
35:37
Speaker 2: the risk tolerance of someone who starts a company like SpaceX.
35:41
Speaker 1: Okay, enough about space let's talk about Crypto, formerly known
35:46
Speaker 1: as cryptocurrency. Let's talk about bitcoin, and let's talk about
35:51
Speaker 1: bitcoin especially at the beginning, right before it was number
35:55
Speaker 1: go up, when it was it really was true believers, right,
35:59
Speaker 1: it was people who had a crazy worldview like you're
36:01
Speaker 1: talking about in these in these other contexts.
36:04
Speaker 2: Yes, so we we still don't know for sure who
36:08
Speaker 2: said to A Nakamoto was, and I think everyone in
36:10
Speaker 2: crypto has at least one guest, sometimes many guesses. But
36:14
Speaker 2: whoever Satoshi was, whoever they were.
36:16
Speaker 1: This is the creator of bitcoin for the one person
36:18
Speaker 1: who doesn't know.
36:19
Speaker 2: Yeah, they had this view that one of the fundamental
36:24
Speaker 2: problems in the world today is that if you are
36:28
Speaker 2: going to transfer value from one party to another, you
36:31
Speaker 2: need some trusted intermediary.
36:32
Speaker 1: You need a trusted intermediary like a government and a bank. Right.
36:37
Speaker 1: Typically in money you need both governments and banks the
36:40
Speaker 1: way it works in the world today, right, Yes, And
36:42
Speaker 1: Satoshi happened to publish the Bitcoin White Paper in October
36:45
Speaker 1: two thousand and eight, which was a great moment to
36:47
Speaker 1: find people who really didn't want to have to deal
36:50
Speaker 1: with governments and banks when they were dealing with money
36:53
Speaker 1: at the financial crisis, right right in the teeth of
36:55
Speaker 1: the financial crisis.
36:57
Speaker 2: Yes, so hes So it is in one sense just
36:59
Speaker 2: this this technically clever thing, and then in another sense
37:03
Speaker 2: it's this very ideological project where he doesn't like central banks,
37:06
Speaker 2: he doesn't like regular banks. He feels like all of
37:09
Speaker 2: these institutions are corrupt, and you know, your money is
37:12
Speaker 2: just an entry in somebody's database, and they can update
37:15
Speaker 2: that database tomorrow and either change how much you have
37:18
Speaker 2: or change what it's worth, and we need to just
37:20
Speaker 2: build something new from a clean slate. And there's also
37:23
Speaker 2: I think there's this tendency among a lot of tech
37:25
Speaker 2: people too. When you look at any kind of communications
37:29
Speaker 2: technology and money broadly defined as a communication technology, you're
37:32
Speaker 2: always looking at something that has evolved from something simple,
37:34
Speaker 2: and it has just been patched and altered and edited
37:37
Speaker 2: and tweaked and so on until it works the way
37:40
Speaker 2: that it works. But that always means that you can
37:43
Speaker 2: easily come up with some first principles view that's a
37:46
Speaker 2: whole lot cleaner, easier to reason about amidst some mistakes,
37:49
Speaker 2: and then you often find that, Okay, you emitted all
37:52
Speaker 2: the mistakes that are really really salient about FIAT, but
37:53
Speaker 2: then you added some brand new mistakes or added mistakes
37:56
Speaker 2: that we haven't made in hundreds of years. So there
37:59
Speaker 2: it's full of trade off.
38:00
Speaker 1: It gets complicated, but at the beginning, right, so the
38:02
Speaker 1: white paper comes out and you know, I covered I
38:06
Speaker 1: did a story about Quinn in twenty eleven, which was
38:08
Speaker 1: still quite early. You know, we'd shocked that it had
38:12
Speaker 1: gone from ten dollars a bitcoin to twenty dollars a bitcoin.
38:15
Speaker 1: Thought we were reading it wrong. And at that time,
38:17
Speaker 1: like I took to Gavin Drieson, who was very early
38:21
Speaker 1: in the bitcoin universe, like he was not in it
38:23
Speaker 1: to get rich, right, Like he really believed, he really
38:26
Speaker 1: believed in it, and that was the vibe then, and
38:29
Speaker 1: like he thought it was gonna be money, right. The
38:33
Speaker 1: dream was people will use this to buy stuff. And
38:37
Speaker 1: one thing that is interesting to me is, yeah, some
38:40
Speaker 1: people sort of use it to buy stuff, but basically
38:44
Speaker 1: not right like that it would go from twenty dollars
38:48
Speaker 1: a bitcoin to one hundred thousand dollars a bitcoin without
38:51
Speaker 1: some crazy killer app, without becoming the web, without becoming
38:54
Speaker 1: something that everybody uses whether they care about it or not.
38:57
Speaker 1: That I would not have guessed. And it seems weird
39:01
Speaker 1: and plainly now crypto is full of some people who
39:03
Speaker 1: are true believers and a lot of people who just
39:04
Speaker 1: want to get rich, and some of them are pretty scammy.
39:08
Speaker 2: Yeah. Yeah, there's like the grifter coefficient always goes up
39:11
Speaker 2: with the price, and then you know, the true believers
39:14
Speaker 2: are still there during the next eighty percent draw down.
39:16
Speaker 2: And I'm sure there will be a drawdown something like
39:18
Speaker 2: that at some point in the future. It's just that
39:20
Speaker 2: that's kind of the nature of these kinds of assets.
39:24
Speaker 2: The bitcoin, it was originally conceived as more of a currency,
39:27
Speaker 2: and so so she talked about some hypothetical products you
39:30
Speaker 2: could buy with it, and then one like the first
39:35
Speaker 2: bitcoin killer app to be fair was e commerce. It
39:37
Speaker 2: was specifically drugs. Yes, it is.
39:40
Speaker 1: It is a very very.
39:41
Speaker 2: Libertarian product in that way. So so it doesn't it
39:44
Speaker 2: doesn't work very well as a dollar substitute for many reasons,
39:46
Speaker 2: you know, most of the obvious reasons. But it is
39:50
Speaker 2: interesting as a gold substitute, where part of the point
39:52
Speaker 2: of gold is that it is very divisible and your
39:55
Speaker 2: gold is the same as my gold, and we've all
39:58
Speaker 2: kind of collectively agreed that gold is worth more than
40:02
Speaker 2: the value than its value as just an industrial product.
40:05
Speaker 2: And then and the neat thing about gold is it's
40:09
Speaker 2: really hard to dig up anymore. Gold supply is extremely anelastic.
40:12
Speaker 1: And so and bitcoin is designed to have a finite supply, right, Yeah,
40:16
Speaker 1: important analogy, Yeah, yes, more generally, like it's what, it's
40:21
Speaker 1: a long time out now, it's you know, seventeen years
40:25
Speaker 1: or something since the white paper. What do you make
40:30
Speaker 1: of the sort of costs and benefits of cryptocurrency so far?
40:34
Speaker 1: The costs are more obvious to me, Like, there's a
40:36
Speaker 1: lot of grift. It's you know, by design, very energy intensive,
40:41
Speaker 1: Like I'm open to like better payment systems. There's lots
40:45
Speaker 1: of just like boring efficiency gains you would think we
40:48
Speaker 1: could get that we haven't gotten, right, Yeah, what do
40:51
Speaker 1: you think about the cost versus the benefits so far?
40:54
Speaker 2: I think I think in terms of the present value
40:57
Speaker 2: of future gains, probably better off. I think in terms
40:59
Speaker 2: of yeah, realize gains so far, worse off? Huh so
41:03
Speaker 2: basically worse off so far, but in the long run
41:05
Speaker 2: will be better off. We just haven't got the payss yet.
41:08
Speaker 2: This is actually something that general purpose technologies. It is
41:10
Speaker 2: a feature of general purpose technologies that there's often a
41:13
Speaker 2: point early in their history where the net benefit has
41:15
Speaker 2: been negative.
41:16
Speaker 1: What what what wouldn't make it clearly positive? Like what's
41:20
Speaker 1: the killer return you're hoping to see from cryptocurrency?
41:24
Speaker 2: Yeah, so I think I think the killer return would
41:26
Speaker 2: be if there is a a financial system that is
41:29
Speaker 2: open in the sense that starting starting a financial institution,
41:32
Speaker 2: starting a bank or an insurance company or something is
41:34
Speaker 2: basically you write some code and you click the deploy
41:38
Speaker 2: button and your code is running. You have capitalized your
41:41
Speaker 2: little entity, and now you can provide whatever it is.
41:44
Speaker 2: Like mean, tweet insurance. You're selling people for a dollar
41:47
Speaker 2: a day. You will pay them one hundred dollars if
41:48
Speaker 2: there's a tweet tweet that makes them cry, you know
41:53
Speaker 2: in your insurance business, Yes, you know, you get to
41:56
Speaker 2: speed run all kinds of financial history. I'm sure you
41:58
Speaker 2: learn all about adverse selection. But like a financial system
42:00
Speaker 2: where anything, anything can be plugged into something else and
42:04
Speaker 2: basically everything is an API call away is just a
42:07
Speaker 2: really interesting concept. And there's the field system is moving
42:10
Speaker 2: in that direction, but slowly, and just.
42:12
Speaker 1: To be clear, like why is it? Why is that
42:16
Speaker 1: better on balance? So for it to for it to
42:18
Speaker 1: be net positive that has to be not only interesting,
42:21
Speaker 1: but that has to like lead to more human flourishing
42:24
Speaker 1: and less suffering than we would have in its absence.
42:26
Speaker 2: Right, Yeah, markets provide large positive externalities. There's a lot
42:32
Speaker 2: of effort in those markets that feels wasted. But it
42:35
Speaker 2: is like markets transmit information better than basically anything else
42:39
Speaker 2: because what they're always transmitting is the information you actually
42:42
Speaker 2: care about. So like oil prices, you don't have to
42:46
Speaker 2: know that oil prices are up because there was a
42:50
Speaker 2: terrorist attack or because someone drilled a dry hole or whatever.
42:53
Speaker 2: You what you respond to is just gas is more expensive,
42:56
Speaker 2: and therefore I will drive less, or you know, energy
42:59
Speaker 2: is cheaper or more expensive, and so I need to
43:01
Speaker 2: change my behavior. So it's always transmitted the actually useful
43:04
Speaker 2: information to the people who would want to use it.
43:06
Speaker 2: And the more complete markets are and the more things
43:10
Speaker 2: there are where that information can be instantaneously transmitted to
43:13
Speaker 2: the people who want to respond to it, the more
43:15
Speaker 2: everyone's real world behavior actually reflects whatever the underlying material
43:19
Speaker 2: constraints are on doing what we want to do.
43:21
Speaker 1: The sort of crypto dream there is just more more
43:26
Speaker 1: finance markets, more feedback, more market feedback, better financial services,
43:32
Speaker 1: as a result. That's the that's the basic view you're
43:35
Speaker 1: arguing for it, and it's just.
43:37
Speaker 2: A really interesting way to build up new financial products
43:41
Speaker 2: from first principles. And sometimes you learn whether those first
43:43
Speaker 2: principles are wrong, but that itself is valuable, Like there
43:46
Speaker 2: there is actual value in understanding something that is a
43:50
Speaker 2: tradition or a norm and understanding why it works and
43:53
Speaker 2: therefore deciding that that norm is actually a good norm.
43:56
Speaker 1: Good last one? All right, you know what it's going
44:00
Speaker 1: to be. You tell me what the last one is?
44:03
Speaker 2: Is AI a bubble?
44:05
Speaker 1: Yes? But you sound so sad about it? Of course
44:08
Speaker 1: we've got to talk about AI, right, are you sad?
44:11
Speaker 1: Talk about AI? Like it's exactly like what what you're
44:14
Speaker 1: writing about. Yeah. When you hear Sam Altman talk about
44:20
Speaker 1: creating open AI, starting open AI, it's like we basically said,
44:27
Speaker 1: you know, we're going to make a GI artificial general intelligence.
44:32
Speaker 1: Come work with us. And when he talks about it's
44:34
Speaker 1: like there was a universe of people who were like
44:36
Speaker 1: the smartest people who really believed, oh that's what they
44:40
Speaker 1: wanted to do, so they came and worked with us,
44:42
Speaker 1: which seems like exactly your story.
44:45
Speaker 2: Yes, it turns out that a lot of people have
44:47
Speaker 2: had that dream and for a lot of people, and
44:50
Speaker 2: maybe it wasn't what they were studying in grad school,
44:53
Speaker 2: but it was why they ended up being the kind
44:55
Speaker 2: of person with major computer science and then try to
44:56
Speaker 2: get a PhD in it and you know, would go
44:59
Speaker 2: go into a more researchy end of the software world.
45:02
Speaker 2: So yeah, there were there were people for whom this was.
45:05
Speaker 2: It was incredibly refreshing to hear that someone actually wants
45:07
Speaker 2: to build a thing.
45:08
Speaker 1: So you have that kind of shared belief. I mean,
45:11
Speaker 1: at this point, you have these other elements of what
45:13
Speaker 1: you're talking about, right, like a sense of urgency, an
45:20
Speaker 1: incredible amount of money, elements of spiritual or quasi spiritual belief.
45:29
Speaker 2: Yes, there are pseudonymous open Aye employees on Twitter who
45:32
Speaker 2: will tweet about things like building God. So yeah, they're
45:35
Speaker 2: they're taking it in a weird spiritual direction. But there,
45:39
Speaker 2: I think there there is something you know, it is
45:42
Speaker 2: it is interesting that a feature of the natural world
45:45
Speaker 2: is that you can actually if you put enough of
45:48
Speaker 2: a you know, you you arrange refined sand and a
45:51
Speaker 2: couple of metals in exactly the right way and type
45:54
Speaker 2: in the right incantations and add a lot of power
45:57
Speaker 2: that you get something that appears to think and that
46:00
Speaker 2: can trick someone into thinking that it's a real human being.
46:03
Speaker 1: The is it good or is it bad? Question is
46:05
Speaker 1: quite interesting here, obviously too soon to tell well, but
46:11
Speaker 1: striking to me in the case of AI that the
46:13
Speaker 1: people who seem most worried about it are the people
46:16
Speaker 1: who know the most about it, which is not often
46:19
Speaker 1: the case, right. Usually the people doing the work building
46:22
Speaker 1: the thing just love it and think it's great. In
46:24
Speaker 1: this case, it's kind of the opposite.
46:26
Speaker 2: Yeah, I think the times when I am calmst about
46:30
Speaker 2: AI and least worried about it taking my job are
46:33
Speaker 2: times when I'm using AI products to slightly improve how
46:39
Speaker 2: I do my job that is better natural language search,
46:42
Speaker 2: or actually most of it is processing natural language. When
46:46
Speaker 2: there are a lot of pages I need to read
46:48
Speaker 2: which contain you know, if it's like a thousand pages
46:50
Speaker 2: of which five sentences matter to me, that is a
46:53
Speaker 2: job for the API and not a job for me.
46:55
Speaker 2: But it is now a job that the API and
46:57
Speaker 2: I can actually get done. And my function is to
47:00
Speaker 2: figure out what those five sentences are and figure out
47:03
Speaker 2: a clever way to find them and then the AI's
47:04
Speaker 2: job is to do the grant work of actually reading
47:07
Speaker 2: through them.
47:07
Speaker 1: That's AI as you full tool, right, that's the happy
47:10
Speaker 1: AI story.
47:11
Speaker 2: Yeah, And I actually think that preserving preserving your own
47:14
Speaker 2: agency is a pretty big deal in this context. So
47:17
Speaker 2: I think that if you are, if you're making a decision,
47:20
Speaker 2: it needs to be something where you have actually formalized
47:23
Speaker 2: it to the extent that you can formalize it, and
47:25
Speaker 2: then you have made the call. But for for a
47:28
Speaker 2: lot of the gruntwork, AI is just it's a way
47:31
Speaker 2: to massively parallelize having an intern.
47:34
Speaker 1: Plainly, it's powerful, and you're talking about what it can
47:37
Speaker 1: do right now. I mean, the smartest people are like, yes,
47:41
Speaker 1: but we're gonna have AGI in two years, which I
47:43
Speaker 1: don't know if that's right or not. I don't know
47:45
Speaker 1: how to evaluate that claim. But it's a wild claim.
47:48
Speaker 1: It's plainly not obviously wrong on its face, right, it's possible.
47:54
Speaker 1: Can you even start to parse that. You're giving sort
47:56
Speaker 1: of little things today about oh, here's a useful tool,
47:59
Speaker 1: and here's the thing I don't use it for. But
48:00
Speaker 1: there's a much bigger set of questions that seem imminent.
48:03
Speaker 2: You know, there's certain kinds of radical insert day there.
48:05
Speaker 2: You know, I think it increases wealth inequality, but also
48:10
Speaker 2: means that intelligence is just more abundant and is available
48:14
Speaker 2: on demand and is baked into more things. I think
48:17
Speaker 2: that it's you know, you can definitely sketch out really
48:20
Speaker 2: really negative scenarios. You could sketch out, you know, not
48:23
Speaker 2: end of the world, but maybe might as well be
48:26
Speaker 2: for the average person. Scenarios where every white collar job
48:29
Speaker 2: gets eliminated and then a tiny handful of people have
48:32
Speaker 2: just unimaginable wealth and you know, rearrange the system to
48:35
Speaker 2: make sure that doesn't change. But I think there are
48:37
Speaker 2: a lot of intermediate stories that are closer to just
48:41
Speaker 2: the story of say, accountants after the rise of Excel,
48:44
Speaker 2: where there were parts of their job that got much
48:47
Speaker 2: much easier and then the scope of what they could
48:49
Speaker 2: do expanded. It was the.
48:50
Speaker 1: Bookkeepers who took it on the chin. It turns out, yeah,
48:53
Speaker 1: like Excel actually did drive bookkeepers out of work and
48:56
Speaker 1: it made accountants more powerful.
49:00
Speaker 2: Yeah, So, you you know, within I think within a
49:03
Speaker 2: kind of company function, you'll have specific job functions that
49:06
Speaker 2: do mostly go away, and then a lot of them
49:08
Speaker 2: will evolve, and so the way that AI seems to
49:11
Speaker 2: be rolling out in big companies in practice is they
49:16
Speaker 2: generally don't lay off a ton of people. They will
49:19
Speaker 2: sometimes end outsource contracts, but in a lot of the case,
49:22
Speaker 2: a lot of cases, they don't lay people off. They
49:25
Speaker 2: change people's responsibilities. They ask them to do less of
49:29
Speaker 2: one thing and a whole lot more of something else.
49:31
Speaker 2: And then in some cases that means they don't have
49:33
Speaker 2: to do much hiring right now, but they think that
49:34
Speaker 2: a layoff would be pretty demoralizing, so they sort of
49:37
Speaker 2: grow into the new cost structure that they can support.
49:41
Speaker 2: And then in other cases there are companies where they realize, wait,
49:44
Speaker 2: we can ship features twice as fast now, and so
49:46
Speaker 2: our revenue is going up faster, So we actually need
49:48
Speaker 2: more developers because our developers are so much more productive.
49:55
Speaker 1: We'll be back in a minute with the lightning round. Okay,
50:05
Speaker 1: let's finish with the lightning round. The most interesting thing
50:10
Speaker 1: you learned from an earnings call transcript in the last.
50:12
Speaker 2: Year, most interesting thing from a transcript in the last year,
50:17
Speaker 2: I would say there was a point, this might have
50:21
Speaker 2: been a little over yeargo. There was a point at
50:23
Speaker 2: which Satia Nadella was talking about Microsoft's AI spending, and
50:28
Speaker 2: he said we are still at the point, and I
50:30
Speaker 2: think he and Zuckerberg both said something to the same
50:32
Speaker 2: effect in the same quarter, which is very exciting for
50:35
Speaker 2: Nvidia people. But it was like, we're at the point
50:37
Speaker 2: where we see a lot more risk to underspending than
50:40
Speaker 2: to overspending on AI specifically.
50:43
Speaker 1: That really speaks to your book, right, that really is
50:46
Speaker 1: like bubbly as hell in the context of your book,
50:49
Speaker 1: like overspending, like the Apollo missions, like the Manhattan Project,
50:54
Speaker 1: like the big risk is that we don't spend enough.
50:56
Speaker 2: And also they know that their competitors are listening to
50:59
Speaker 2: these calls too, so they were also seeing that this
51:02
Speaker 2: is kind of a winnable fight that they do think
51:05
Speaker 2: that there is a level of capital spending at which
51:08
Speaker 2: Microsoft can win simply because they took it more seriously
51:10
Speaker 2: than everybody else.
51:11
Speaker 1: So he's like, yes, We're going to spend billions and
51:15
Speaker 1: billions of dollars on AI because we think we can win.
51:20
Speaker 1: Zuckerberg implicitly, what's one innovation in history that you wish
51:30
Speaker 1: didn't happen?
51:36
Speaker 2: I wish there were some reason that it was infeasible
51:40
Speaker 2: to have really, really tight feedback loops for consumer facing apps,
51:45
Speaker 2: particularly games.
51:48
Speaker 1: Is that a way of saying you wish games were
51:50
Speaker 1: less addictive?
51:51
Speaker 2: Yeah, I wish games were less addictive, or that they
51:53
Speaker 2: didn't get as they weren't as good at getting more addictive.
51:57
Speaker 2: So I wrote a piece in the newsletter about this recently,
51:59
Speaker 2: because there was that wonderful article in the Loneliness Economy
52:01
Speaker 2: in the Atlantic a couple of weeks back that was
52:04
Speaker 2: talking about it. We just spend One of the pandemic
52:07
Speaker 2: trends that has mean reverted the least is how much
52:09
Speaker 2: time people spend alone. And I think one of the
52:11
Speaker 2: reasons for that is that all the things you do alone,
52:14
Speaker 2: they are things that produce data for the company that
52:17
Speaker 2: monetizes the time that you spend alone. And so the
52:20
Speaker 2: fact that we all washed a whole lot of Netflix
52:23
Speaker 2: in the spring of twenty twenty means that Netflix has
52:25
Speaker 2: a lot more data on what our preferences are.
52:27
Speaker 1: So they got better at making us want to watch Netflix,
52:30
Speaker 1: and all the video games we played on our phones
52:34
Speaker 1: got better at making us addicted to keep playing video
52:37
Speaker 1: games on our phones. Yeah, that's a bummer. It's a bummer.
52:42
Speaker 1: What was the best thing about dropping out of college
52:44
Speaker 1: and moving to New York City at age eighteen.
52:48
Speaker 2: So I would say that it was it really meant
52:53
Speaker 2: that I could could and had to just take full
52:56
Speaker 2: responsibility for outcomes, and that I get to get to
53:01
Speaker 2: take a lot more credit for what I've done since then,
53:04
Speaker 2: but also get a lot more blame. Where there isn't
53:07
Speaker 2: really a brand named fall back on and so if
53:10
Speaker 2: someone hires me, they can't say this person got a
53:12
Speaker 2: degree from institution X. You know I didn't eve. I
53:15
Speaker 2: dropped out of a really bad school too, So there's
53:18
Speaker 2: there's not even like the not even the the extra
53:21
Speaker 2: upside of you know, I startup was so great, I
53:24
Speaker 2: just had to leave Stanford after only a couple of semesters. No,
53:27
Speaker 2: it was it was Arizona State and I didn't even party,
53:29
Speaker 2: so they yeah, but yeah, it's it's that it's just
53:35
Speaker 2: being being a little more in control of the narrative
53:37
Speaker 2: and also just knowing that it's it's a lot more
53:41
Speaker 2: up to me.
53:42
Speaker 1: What was the worst thing about dropping out of college
53:44
Speaker 1: and moving to New York at eighteen?
53:46
Speaker 2: So one time I went through a really really long
53:48
Speaker 2: interview process for a job that I really wanted and
53:52
Speaker 2: at the end of many, many rounds of interviews and
53:55
Speaker 2: you know, a work session and lots of stuff. The
53:59
Speaker 2: hiring committee rejected because I didn't have a degree and
54:01
Speaker 2: that was on my resume, so that was kind of inconvenient.
54:04
Speaker 2: I guess another another downside, like it might have been
54:09
Speaker 2: to spend more time with fewer obligations and access to
54:14
Speaker 2: a really good library.
54:22
Speaker 1: Burn Hobart is the co author of Boom Bubbles and
54:25
Speaker 1: the End of Stagnation. Today's show was produced by Gabriel
54:29
Speaker 1: Hunter Cheng. It was edited by Lydia Jeane Kott and
54:32
Speaker 1: engineered by Sarah Brugier. You can email us at problem
54:36
Speaker 1: at Pushkin dot FM. I'm Jacob Goldstein and we'll be
54:39
Speaker 1: back next week with another episode of What's Your Problem.