21: Innovation Systems with Ben Reinhardt

Play episode
Hosted by
Will Jarvis

In this episode, we talk with Ben Reinhardt about different innovation systems, how to create more sci-fi technology a reality, and why our research institutions are not as effective as they used to be. You can check out Ben’s work at https://benjaminreinhardt.com/about/.

Some things mentioned in the episode:

Studies on Slack

Don Braben

DARPA

Ben’s Twitter

Transcript:

Will Jarvis 0:04
Hey folks, I’m William Jarvis, along with my dad, Dr. David Jarvis. I host the podcast narratives. narratives is a project exploring the ways in which the world is better than it has been. The ways it is worse in the past or making a better, more definite future. I hope you enjoy it.

Well, Hey, folks, today on the podcast, we’ve got been Reinhardt. Did I get your name? pronunciation? Right. Okay, the German I in the back. Okay, perfect. So Ben, I like to get kicked off, could you just give us a, you know, kind of a brief bio, you know, what your mission statement is? And, and what you’re trying to do in the world? Sure. So So I’ll start with that. Very simply, it’s, I want to be more awesome sci fi in the world. Am I allowed to say that? I forgot to ask. Okay.

Ben Reinhardt 1:04
All right. And

so that’s, that’s it the end of the day, like, that’s, that’s what I’m trying to do. And I guess my bio is, like, if you want to sort of tie narrative around it, it’s me trying out different ways of of doing that, and realizing that they’re not the right way. And so trying a different way. So my, like, I, I went to grad school, because I thought I yes, if I get a PhD, then and like build spaceships, that’s the correct way to make there be more awesome sci fi stuff. And I found like, that wasn’t quite the right thing. And so it’s like, I guess, they go and I join a startup and I build amazing augmented reality technology, that will be the way to bring the sci fi future. And that was not the right way. So it was like I asked if I quit, and start a company, that will be the right way that there was there’s a couple there. And so I was like is, well, if I then instead help other people start companies. And so I went, and I worked at a VC firm and suppose called entrepreneur first. And maybe that will be the way to enable it. It was the that wasn’t quite right, either. So now I am, I’ve come to the conclusion that the current institutional structures that we have right now, are not

Unknown Speaker 2:25
that big, constrain certain activities that have happening, that I think are essential to having there’ll be more awesome sci fi stuff in the world. And so I’m trying to figure out what, what institutional structures might be able to fill that gap? And how to below

Will Jarvis 2:43
that, that’s, that’s awesome. Do you think there has been a change since like, the 70s, and how fast innovation has gone on and like so, you know, we’ve had like, we had these really wacky scientists back in the day, like Peter Thiel has this line that, you know, the the letter from Einstein wouldn’t get to the President’s desk today, you know, and that seems to be true. On some level, like, the it sounds like it’s almost seems like academia has weeded out. The weird people like Isaac Newton, you know, half of this time was like, called, I don’t know, if I was having this time was on like, alchemy, and the other half was gravity. Right.

Unknown Speaker 3:16
That’s simply more than half of it. That’s it? Yeah. So yes, I do sort of buy into the whole great stagnation.

Will Jarvis 3:29
decade. argument.

Unknown Speaker 3:31
Yeah. So so. So I definitely think that that is the case. I think the the more interesting question is like, is why is that the case? And then the even more interesting question is even putting aside like, the Why is only interesting insofar as we can use that to reverse that or, or sort of speed things up again, in my opinion.

Will Jarvis 3:57
That makes a lot of sense. So what have you found so far? Because I know you’ve been you’ve tried a lot of different things. What has been most effective? What do you think the the clearest vision you have for kind of accelerating technology? And, you know, on the back end science as well?

Unknown Speaker 4:13
Yeah. So I’m gonna take sort of one step back, and try to like, disambiguate that a little bit because I think one of the, the sort of meta realizations is that there are many that that sort of lumping it all together beyond saying, like, there’s some problem. It’s not too useful to lump it all together. So I would sort of disambiguate and say, like, I think I’m sort of really good at a certain set of activities, right? So they’re really good at bringing novel solutions to new markets. But then, I think the trick is that a lot of the day activities that need to happen now are not that so In terms of things that are especially effective, like, I think, at the level of granularity of a podcast, yeah, things that I would say are going to sound like, they like things that everybody sort of knows. The trick is like how to make them happen, right? So it’s like, get, like, really good teams of smart people together, like, like, give them like a good focus and, and let them go. But then, like, keep them keep them sort of tightly on the rails. Um, so So I think the interesting thing is like, how do you navigate all of the incentives that people have and that are in the world? In order to allow those things to happen like that, to allow the effective things to happen?

Will Jarvis 5:51
That makes sense. Yeah. I guess so I guess what you’re saying is, like, startups are good at certain problems. Yeah. And so like, and Vint, even when we think about venture capital, it’s like, maybe Peter Thiel, again, someone at Founders Fund has this line, it’s like, venture capitals, you know, between like, a million dollars and like, maybe a little bit less, and then a billion dollars, like, but it’s somewhere between that once you get above that it’s like, this is not where venture capital, you know, would be useful. So like, Are you saying there’s like, things that are too capital intensive? Or they’re just like, if there’s not a clear revenue, opportunity or profit motive immediately? Like, if it’s too long term, there’s no reason for people to go after it?

Unknown Speaker 6:32
Yeah, I would say all the above, I think what the way that I would frame it is that venture capital has a certain set of constraints. And I think the, the interesting thing is that you can sort of think of all institutional structures as having a certain set of constraints. And when you start looking at through that lens, then you start to see like, okay, like, what things are sort of constraint advocating. So for example, venture capitalists have the constraint that they need to raise a new fund every couple of years. In order to raise a new fund, you need to go to your LPs, right, surely turn on capital. Exactly. Exactly. So and so it’s like, even if you’re making a 10 year bet, which which VCs do, you need to be able to show that that that is sort of continuously increasing? And so you need an A, the only way that you can price a venture capital investment, is if you raise a round. Right, right. And so your, as a venture capitalist, you’re then sort of constrained out of investing in things that don’t have sort of a legible regular price increase. Gotcha. And so it’s like, a little bit more settled than just the amount of capital. But yeah, so so it’s like,

Will Jarvis 7:55
things like, like, difficult to value in the short term, at least.

Unknown Speaker 7:58
Yeah, or like, and that, like, Don’t, that don’t sort of stick follow like a nice, a nice growth curve. Right. So gotcha.

Will Jarvis 8:07
Yeah. So what would be a good example of that? Like, like a problem like that? Um, so I guess, like, from, from my, like, I’ll just like my, my personal experience. Yeah, definitely. Um, sort of like trying to build a, an entirely new system, I think is really sort of falls into that. So like, like, I worked at a company called magically for a couple of years, and we were trying to build, you know, like, and basically, they were trying to build like an entirely new, like computer interface. And really cool demos, I remember that

Unknown Speaker 8:44
super cool demos, and like, the technology was super cool. The trick was that there were so many pieces that needed to be built, that it wasn’t like the rate at which they needed to, like, get a product out there and start showing increasing numbers, right? prevented the kind of approach that I think that they needed to successfully introduce, like an entirely new way of interacting with computer. So that’s, that’s one example. Um,

Unknown Speaker 9:17
I think it’s

Unknown Speaker 9:20
also just like places where, like, more abstractly, where you want to build a new system. And in order to build that new system, you actually need to create several components. And any one of those components on their own is fairly worthless, or like, isn’t that valuable, it’s only in combination that they’re they’re really valuable. The pressure from to to sort of get a product out there to start increasing in value makes you start trying to monetize each of those components, which then sort of like punts the eventual system farther and farther off until it Like never exists,

Will Jarvis 10:01
right? That makes us focus on the short term. You know, and I’ve worked in startups I graduated, and this is true, like you got this grand vision you’re trying to accomplish. And then in the short term, it’s like, well, you know, we got to make it a next week, right? And the next week always dominates it. Like no matter what, yeah. And you can get some slack. Like, I think Scott had a post on slack. Maybe I love that post

Unknown Speaker 10:21
that Oh, yes. I mean, sure. So So Scott, Alexander, author of star Codex has a post, it’s called studies on slack. And it is amazing. And you the listener should pause and go read it, I’ll put a link. But it’s short, it’s effectively makes the argument that in order to get sort of, like, if you think of yourself as like, a, like, I like some water, and you’re trying to get to the the lowest possible point, then the like, if you’re like purely just doing optimization, you go to the lowest local point, but that might not be the lowest global point. So like, the global minima, in in a sort of optimization parlance. And in order to sort of like, get to a global minimum, you need some amount of slack. And you could imagine that it’s like water analogy, get imagine that as like, the ability to sometimes go against gravity. sort of get into another puddle. That’s a poor explanation. But that’s,

Will Jarvis 11:40
I think, yeah, that’s great. I think you also mentioned the the development of like an AI. Oh, yeah. As a great example, like, this is incredibly complex thing. And you really couldn’t just, you know, through these super hyper competitive processes, create something as complex as an AI. Yeah, exactly. So that’s, that’s really interesting to me. So do you think some of what happened is, is things have gotten super competitive, and some weird senses is like, like, so I think of like, the academic like grant process where, like, essentially, you know, so my mother in law writes grants. And it seems like and there’s actually econ papers on this where all the profits from grant writing get competed away in the end. Yeah, like, it’s like this complete net loss. Is that a real phenomena? That just?

Unknown Speaker 12:26
Oh, that’s absolutely, that’s absolutely real phenomena. I think if you calculate the amount like, like, if you price, people’s time, at some, some amount of money, and then you can like multiply that by the amount of time that people in aggregate spent writing grants, it comes out to be some number that’s like, way bigger than the amount of the grant itself. Yeah. So so I think you’re spot on. Early in, in my, my analysis, you’re spot on about the competition. And so like in academia? Absolutely. There’s a pretty strong argument. I think, that the reason that sort of the great r&d Labs of the past went away, is at least in large part because of increased competition on on companies. If you if you look at it, like both Xerox PARC and Bell Labs are sort of like the the Paragons of r&d labs were both part of monopolies. Right? And so these people were getting monopoly rents. And that’s, that’s sort of what they needed to, to to run those those labs. And so, this isn’t me. Like, it’s a tricky thing. Because like, culturally, sort of as a 21st century Americans, we definitely are inculcated to think that competition is good, right? Like competition is what, like, forces people to improve and get better. And like, I agree with that, but at the same time, you sort of had to put on this other hat and be like, well, you also need a little bit of slack. And I think that we don’t have that in in sort of our current, like innovation ecosystem.

Will Jarvis 14:06
That makes a lot of sense. And there’s this great I can’t remember give this talk or find the link but they were talking about how it was in the context of free trade and Japanese automakers and how they were they use protectionist policies to allow like, you know, Japanese cars to get really good and out compete American cars. And and like, there’s these weird effects that you know, like, yes, like free trade does get us better products at the end of the day. And like more competition gives us better cheaper products, but they’re also back in effects that can happen and we should think about them.

Unknown Speaker 14:35
Yeah, absolutely. I think the the really good book on that is this book called how Asia works. That sort of talks about interesting, sort of the difference between like, why South Korea and Japan managed to rebound from World War Two pretty handily, while countries like Malaysia sort of have not, and they really sort of go into that, like, I mean, at the end of the day, it’s like protection. ism, which is like a dirty word. But it gave the companies the slack they needed to, like become Toyota. Right, like become right.

Will Jarvis 15:08
Largest auto manufacturers. Yeah. Which is, which is quite valuable for to have that in your country. That’s really interesting. So I wanted to jump back. And so you mentioned Bell Labs, you know, Xerox PARC, these places that no longer exists a lot, a lot of because competition came through and you know, got rid of all their monopoly rents, and they, they don’t have the money to do it anymore. Or they, you know, get competed away, don’t exist anymore. Sorry, I didn’t say that rule. But technically they do. Yeah, that’s

Unknown Speaker 15:39
the interesting thing is both Bell Labs and Xerox PARC are still places they still have employees. They still do research. Really? Yeah, absolutely. They just, yeah. And that’s, they’re just not. It’s not like, they’re not there. They’re their former selves. Right. So

Will Jarvis 15:54
it seems like Google right. So they should they are the ones that did the most clear example to me that should be doing this because they have cash equivalents that when I checked a couple, this was years, a couple years ago, was it was like 100 billion in cash equivalents. I mean, they’re sitting on 100 lots, but they’re burning, like, you know, some percentage just inflation every year, right? Because they don’t know what to do with it. And it seems like this clear chance, and they they speak about being like this highly innovative company, you know, like I have their phone, right? I use their search engines, but they and maybe they’ve done that with like, maybe their life sciences stuff a little bit where they’ve invested in anti aging resorts, but it seems like they should be doing more. And often wonder if like the backlash against big tech, at its core is some complain about them. not living up to the promise at the end of the day? I don’t know.

Unknown Speaker 16:44
Yeah, there’s so there’s sort of, can I can I take that two directions? Absolutely. So So one is I would argue that to some extent they are, and the way that they’re doing that is you look at the results that came out of DeepMind recently, right, right. That’s Google money. But like alpha fold? Yeah, so we’re talking in December 2020. And DeepMind, recently released alpha fold, and protein folding stuff. Exactly. Exactly. Like, so. Um, and like, similarly, there’s like, really good work that comes out of Google brain. So I would, I would argue that those organizations actually are sort of doing that like doing the the Bell Labs equivalent for Google. The trick is that I have like, a wacky theory, that in order for corporate r&d to be really effective, it needs to address some amount of like, existential threat to the parent company. Gotcha. And, like, Google’s existential threats come in the form of like, software, like they write, they need better software. And so you could expect that their corporate art, like their high quality corporate r&d to be in software, and that’s what you see. And so I think the the thing is, like, what we want is for Google to have really high quality r&d in, in, in atoms, but because that is not going to address an existential threat to them, except in places like, you know, like, tensor processing units like that, that’s, that’s not going to happen. So that’s the thought there and then I think, oh, to your point about like, people like the the sort of like the tech lash, against them for protecting Beckett, I think that’s that’s absolutely spot on. And, like, arguably, the, like, the whole reason Bell Labs existed was sort of as a as a as a peace offering to the American public to for like, at&t to say, like, Look, we’re good guys. We make technology and like spread to the world. So yeah, I’d love to see a little more of that. That’s, that’s,

Will Jarvis 19:04
that’s really interesting. I’ve got two responses that made me that I’ve just thought about so one is we talk to us v. Marshall. What’s up read his blog?

Unknown Speaker 19:18
Yeah. Actually, I know him. Because I played magic and gathering very competitively. When I was younger.

Will Jarvis 19:27
Oh, yeah. That’s awesome. Yeah, he’s a big deal, man. It’s cool. Yeah. So we were talking to me and I’ve had this off for a while and what she repeated what you just told me confirms that my pre existing belief here is that and and I’d love to get your, your take on this. Is there anything that’s that we can do to motivate people that is not an existential threat? Because like, you know, like we had World War Two, right, the Cold War, and these are like existential threats and we rose the challenge. COVID we’ve had a lot of trouble, right? It’s like not big enough? Like, that’s kind of scary. And then like you mentioned, like these research labs, if it’s not an existential threat to the company do do people ever really work very hard? I don’t know. Is there a way to get around that? Yeah. I

Unknown Speaker 20:14
don’t know. It’s like, the real answer is, I don’t know, I would love to figure it out. I think perhaps another way of framing it is like, how can we make existential threats be more in line with the sorts of things that we want to see like? And so where I go with that is like you think about, like medieval cathedrals, and arguably medieval cathedrals, actually, were addressing an existential threat, because people thought that their souls were on the line. And that if they did not build things to the glory of God, then they were going to go to hell.

Unknown Speaker 20:58
Forever.

Unknown Speaker 21:00
And, and sort of, um, and so like, you could, you can still put that in the existential threat framework. It’s just that existential threats, were a very different thing to those people. And so I just sort of wonder if, like, Can we shift culturally so that the things that are existential threats to companies or individuals, just so happen to align with the things that we sort of want to see in the world? I don’t know, that’s just some guy with a Nissan know that that’s, it seems

Will Jarvis 21:35
really important, it does seem to be a huge challenge. Because, you know, we just have to, we have to figure out ways to motivate people, where else, these problems don’t get solved. That does remind me to talk to Don Braben. Yeah, yeah. And so actually, I’m talking to him in like, a month or two. Yeah, so I cuz I read his book. And then I was like, I’m gonna email him because I found out like, Okay, he’s still alive, you know, and like, here’s his email. And then I found your podcast, which was super awesome with him, and very informative. And what do you think about his idea of just, you know, essentially giving really, really smart, talented, weird people money and just saying, Go?

Unknown Speaker 22:15
Yeah, I think sort of get, like, going back to a point earlier is, is this this point of disambiguation? I think that the sort of Fincher research idea, that’s sort of his thing, is really good for a certain set of activities that are essential. So I think it’s really good for like, what what he calls like the plot cut club, so people who are really trying to sort of like pry new knowledge from the like, the grips of nature itself. Right. And I think that that’s, it’s really good for that. What I don’t think is that it’s like, I don’t think that that’s what we should do for like, everybody who has any sort of innovative idea at all. Gotcha. That makes sense. Yeah.

Will Jarvis 23:04
So like, so maybe good for, like, pure research or something. But this very select group of people, maybe, but not Yeah,

Unknown Speaker 23:12
like, a select group of people like it. And I think it’s actually a really a, a fascinating challenge to try to think about, sort of, like, what are the parameters of, like, that sort of area of like, activity space? Right? where it’s like, like, it does it have to do with like, the sort of, like the size of the group that you need to do the discovery, right? Because he was funding sort of, like, very small groups of like, one to three people. It’s like, is that the key thing? Is it the time scale? Is it like that? These were things that these people were like, devoting their lives to work. But, yeah,

Will Jarvis 23:55
that’s really yeah, it’s really a shame. So maybe like a partial solution to one problem.

Unknown Speaker 24:01
Yeah, exactly. And I, my, my hypothesis is that that’s sort of what we need is like, there isn’t going to be like the solution, right, like capital T, capital S. Um, it’s, it is going to be lots of, like, specific solutions for specific parts of, of activity space.

Will Jarvis 24:26
That makes a lot of sense. Do you think peer review is as evil as he says it is? Yes. I think I’m not in academia, so I don’t know.

Unknown Speaker 24:36
Yeah, um, I think that he makes the strongest argument for it. Um, but I don’t think that he’s wrong. I guess maybe,

Unknown Speaker 24:53
maybe what

Unknown Speaker 24:54
I would argue like, I would argue that it’s bad, but what I’m not as confident as he is. Is that it is sort of like the the sole sole perpetrator of the problems with academic funding. I think that that’s where I might soften it a little bit.

Will Jarvis 25:14
Gotcha. That makes sense. It’s terrible. It’s bad. Yeah. So it’s what do you think some of the other problems aren’t like just the grant process? You know, all the politicians win? Or are there other things?

Unknown Speaker 25:26
Yeah, I think there are. There are many, many things just so some, like, I’ll just like, throw out some of them is like, and the thing to keep in mind is, it’s like, well, at the end of the day, we really want to find try to find these are like the root causes. And it’s really tricky, because like, all of these factors sort of like feed into each other. And it’s really unclear of like, where to start untangling or the ball, right. But, like, some, some problems are the fact that academia is sort of become like, highly overloaded. Like, we expect academia to both sort of like, train like, train new researchers, educate do like general undergraduate education. So educate like, citizens of a, like liberal Republic, Joe Schmo. Yeah, exactly. To do like, sort of, like deep exploratory, natural sciences. And at the same time, do basically Industrial Research. And at the same time, basically do like military research. And at the same time, like, be like thought leaders, right, like, it’s like, oh, it’s like, like, the scientists that we listened to. So like, they need to be like, have like, so. You know, so it’s like, highly overloaded, I talked to friends who are professors, and we’re like, counting off the number of jobs that they actually do. Oh, and they, the professors are expected to be administrators as well. So, so, so like, professors are doing like, four or five jobs at once. And each one involves a very different way of thinking, right? So like, there’s like, like, my friend literally had to go from like, like arguing with an undergraduate about the grade, the midterm, to like, what is the research program that I need to set up the next five to 10? years? Like, in an afternoon?

Will Jarvis 27:38
Like, like, how are you gonna actually? Yeah, how do you actually make discoveries when you’re trying to juggle all that?

Unknown Speaker 27:44
Exactly. So that’s like, that’s one of the problems. Another is just like, the sheer number of people who hope to become professors versus the number of people who are professors. And then the sort of the way that professors are judged, right? So it’s like, like, and just like, the way that tenure works, is another problem, because it’s like hair come in, we’re gonna judge you on like, how many papers you publish, and how many how much grant money you bring in, in your first couple of years at the university. And, like, the people who pass through this filter that filters for like, really aggressively publishing papers and getting doing research are the ones who then get to stick around forever, and are expected to do like 20 year projects. Exactly. Sounds like. They’re

Will Jarvis 28:40
Right, exactly. No, that’s, that’s really fascinating. And then you’re right, so then there’s this huge cullet, right, and so that just drives comp, it makes it more competitive. And then you get more of these weird effects. Exactly. It

Unknown Speaker 28:52
goes goes right back to that competition thing, where there’s, there’s no slack in academia, like zero.

Will Jarvis 29:00
Yeah, it just, it seems so screwed up on so many levels. So you’ve recently been writing a little bit about another organization that does things a little bit differently, and we’ve done a little bit of work with them in the past. So DARPA, so what makes makes DARPA special, maybe the program manager model, it’s a little bit different.

Unknown Speaker 29:20
Yeah. So, the the thing that sort of defines the DARPA model is that they is that it sort of revolves around these these what are called empowered program managers. And so instead of either having some sort of like, like manager level person or like an executive say like Okay, do and then at the same time, and also contracted to have a researcher say, Okay, this is what I want to do. You have these, these program managers, who are given like an incredible amount of leeway To go to the researchers try to understand what the researchers think and then like, then go to people in the military and see what they need. And then basically design a program that coordinates the efforts of several research groups that go across academia and industry to try to accomplish fairly or audacious goals, and I think it’s so the thing that really makes them different different is one, just the structure of the program managers, and then to the the amount of leeway that those those perfect managers are given.

Will Jarvis 30:41
Gotcha, that makes a lot of sense. It does seem like, freedom does seem to be really important in these in these like, and and giving, like, one guy can go wring his neck if it doesn’t work out in 10 years. And like, like, just go manage it. Yeah. With COVID. Right, the best things that have happened in the US are like, okay, we appoint some Army Corps of Engineers guy, and we say, go do it. And here’s $26 billion, you get a vaccine?

Unknown Speaker 31:05
Yeah, let’s, let’s record about. So

Unknown Speaker 31:06
I think a really consistent thing that that I’ve observed is that the thing that seems crucial to all these things is like trust that you need to do, like trust that, like, this person, if you give them a lot of power, they’re actually going to go do the right thing. And so like, I think a almost like a systematic lack of trust may be another thing that like, like, I think one of the core reasons that we’re seeing all this slowdown, because like, if you don’t trust people, then you either need to restrict their action, or like, force them to, like, give you lots of metrics,

Will Jarvis 31:48
that that makes a lot of sense. And you see this in like the defense contractors, so so our CTO, really good friend of mine, he, he won the technologies of the year at Lockheed Martin, and he just he will sit there and describe sometimes, okay, so like, what is Lockheed Martin manufacturer will like, I don’t know, you know, f 35. You know, he’s like, no, it’s billable hours. That’s what matters. And it’s like, it’s firm, fixed price, and you’ve got billable hours. And that’s because the contractors, you know, at one time, they screwed over the government, and they were selling these cokes in Iraq for like 10 bucks a pop? Well, you know, they were actually out, it probably cost them like seven bucks a pop. But you know, there’s a big, big news story. And so you go in and you it’s like, because one bad thing happens, you create this role for everybody. And that just piles on and piles and piles on, and then it just gets worse and worse and worse. Exactly.

Unknown Speaker 32:36
That nails it. And I think you see that sort of everywhere. Yeah, I would, I would argue that DARPA actually could not like in its current form, and the government would not be started today. And that it’s basically been get, like grandfathered in, because it was started in in the 50s. And had such a good track record.

Will Jarvis 32:57
Interesting. Yeah. It could just, yeah, that’s really. And that’s a real problem. Right? That’s a real problem. Yeah, absolutely. And the only solutions I’ve seen that seem to work at all I have you seen this? So you’re in North Carolina right now? Do you see the DMV? in North Carolina? This is really interesting. So the DMV, and I’m not sure exactly why they do this. But this has actually worked pretty well. So if you ever do anything with the DMV now, it’s really cool. We can do it all through an app at this point. Nice. It works much better than it used to, and how did they get there? Well, they actually they took the DMV, which was in Raleigh, and moved it 60 miles east to my hometown, Rocky Mountain, in the middle of nowhere. And what that did was, essentially 70% of people quit, and you can rebuild the organization. And so I do wonder if you know, the answer is, well, DC goes to Kansas City or something? I don’t know, just somewhere in the middle. Yep. You know what, and we’re just going to restart it. Right, and Bozeman, Montana or something?

Unknown Speaker 34:00
Yeah. Well, that’s another thing about DARPA, that I think actually is, like, does really help is that there’s just like, program managers are out after five years. Like, it doesn’t matter. How, like, doesn’t matter how good a job that you did, it doesn’t matter how, if you if you do a really, really bad job, you’re out before five years. But yeah, and so I think that that there really is something to that sort of like refresh.

Will Jarvis 34:27
That short tenure. Yeah. Do you know how they had the idea for DARPA originally? Yeah.

Unknown Speaker 34:33
It was both me basically. Like, Dwight Eisenhower, freaking out over. Like, like, like so what it was is, if I get the story correctly, it was it was that there were that like, this was like height of the Cold War. Like the Russians are Just starting to touch test h bombs, we need to be able to understand like, what like what they’re doing, and defend ourselves from them. And so there’s like sort of all these, like crazy programs going on in, in the military to try to apply science to these, like these critical national defense problems. Right. And so they’re like, okay, we’re just gonna, like, stick it all together in this in this one agency.

Will Jarvis 35:26
Wow. Yeah, that’s it’s also interesting. Are there any other keys? You think that they’ve had to keep it avoiding entropy in DARPA itself?

Unknown Speaker 35:37
Yeah. Well, so I would accept that it was like, I don’t think that they have completely. Okay. So I would like, I think that it has like, slowly become more and more of a government agency. And so it’s like, I don’t want anybody could give that it’s perfect. But still, like really, like, considering how well they’ve done over the course of 60 years. Like, it was really impressive. But like, ways to avoid entropy, I think it’s one. A big one that I think maybe is understated, is the fact that it’s state really small. It’s interesting. It’s about 100 program managers. And like, my hunch is that that number is not a coincidence, and it like happens to be right around Dunbar’s number. And my hunch is that there is something to the fact that like, everybody in the organization can kind of know what everybody else is up to, that actually helps build that trust that we were talking about. Interesting. So so I think that that’s, that’s really clutch. And then also what it means is that

Unknown Speaker 36:45
they

Unknown Speaker 36:47
because it’s small, and because it actually has a relatively small amount of money, in the grand scheme of things with God funding like it’s, it’s, like just over a little more than billion dollars a year. Gotcha. And that’s, that’s a lot of money, but not a lot compared to deity funding. And so what that means is that there’s a little bit less pressure. Gotcha, to to deliver. And the the anecdote that I love is there was a director in the 90s actually went to Congress and petitioned to have DARPA’s budget lowered. Because, yeah, he made the argument that like, if you give us too much money, you’re gonna expect too much out of us. And I think this is actually another way of creating like, a little bit of slack is, is that if you have a bot, like, if you’re overfunded, then you have no slack to sort of, like try crazy stuff. So I think that that, like the size is, is another way that they’ve managed to avoid entropy. And, and I think it’s also just like, this is like, really sort of trophy, but like, they get really good people like they interesting. Like they really focus on just like, they don’t care about backgrounds very much. It’s just like, okay, like, okay, we get in here. Okay,

Will Jarvis 38:19
we’ll do a good job. Yeah, exactly. That makes a lot of sense. And it also seems like keeping it small. Like that means that you don’t have this problem where you’re forcing money in places where, like, maybe it’s like, I don’t know, this is not like a great idea. But we’re doing it just because we got to keep this and you want to keep your budget, right? You always want to try that. You see this, like corporate environments at all time. It’s like, well, we got $100,000 left. Gotta go buy something, right, because they’re, they’re gonna cut our budget next year. Yeah, exactly. That’s interesting. So I want to move on to another topic, do you think problems and this is a very large, very large questions, so I want to preface it. Do you think problems have gotten harder in general in science and technology?

Unknown Speaker 39:02
Yeah, that’s a really good question that I am not sure I have a good answer to it’s a very hard question.

Unknown Speaker 39:09
Um, I?

Unknown Speaker 39:13
Yeah. Well,

Unknown Speaker 39:16
I think that

Unknown Speaker 39:18
it depends on where we’re looking. Right. Like, it’s like, like, I think that looks certainly like problems in physics, like, like any, any field with a name. I think that the problems have gotten harder, but that’s like, kind of the definition of a field.

Unknown Speaker 39:38
But then,

Unknown Speaker 39:39
I think that the, the, I think the problem is that we haven’t been

Unknown Speaker 39:46
doing

Unknown Speaker 39:48
as fast as we used to. And so it’s like, that’s, that’s almost what what I think the real problem is more so than like, the problems have gotten harder because it’s like, let’s think through this.

Unknown Speaker 40:00
If, like, the problems were

Unknown Speaker 40:05
harder, like, Look, the problems are like, exactly as hard as people can solve almost by definition. where it’s like, like, Yeah, I don’t know, how would you compare, like, the hardness of two problems?

Will Jarvis 40:21
That’s difficult, isn’t it? That’s really difficult. Yeah. It’s always Yeah. Yeah. Yeah. Where do you even start? Right? Yeah. I mean, it’s like, definitely, um, you know, it’s like, maybe you know it, if you see it, like, maybe that’s it, but it’s like, obscenity or something like that. I

Unknown Speaker 40:39
don’t know. Well, let’s,

Unknown Speaker 40:41
let’s let’s, I think maybe a proxy is like, so equipment is definitely like to in order to do a lot of cutting edge experiments. equipment is definitely more expensive in some areas. Right, like, right. Yeah, that’s, that’s definitely true. But at the same time, we’re much wealthier society than we were 100 200 years ago. Right. So does that that balance out? And then the cost of equipment has fallen in, in other areas, right, like sequencing a genome is much, much less expensive than it was years ago. So it’s just like, yeah, I, I struck like, I think it’s, it’s, it’s, it’s a good question to think about, but a possibly impossible one to answer. But just because it’s not. Interval does not mean it’s not worthwhile to like, talk about or think about together. Yeah, definitely.

Will Jarvis 41:39
I do really like your heuristic. If it has a name, by definition, the prot, these problems will be more difficult. And I think that it’s almost because there are more people looking right.

Unknown Speaker 41:51
Yeah, well, yeah. Well, that’s, that’s, that’s another thing is like, yeah, like, like, once, once you say like, okay, like, it’s physics, and like, then you have like, a whole bunch of physicists and physicists Russian. Yeah, exactly. Um, so. So yeah.

Will Jarvis 42:07
Do you think some of the problems we’ve seen with slower? Well, I, what I see to be slower advances in like biology, or, like, I have this weird thought that people get selected into physics, you just mentioned physics. So like, physics is like, that is the hardest thing. That’s the most interesting. And I had a really smart roommate in college, you know, like, this is what I want to get to. And I was always like, man, like, I think like, they’re more gains to be made in other places, right? And not like the top end of like, you know, string theory or something where only 100 people understand it. And there’s no way for me to evaluate whether it has any solid or at at all right, like, I don’t know, I’m sorry, like the, I think I missed the Yeah. Do you think we we select, we’ve got a poor selection mechanism, in that, perhaps like physics is high status, and that it sucks up a lot of talent that could be going to like, I particularly I think about biology,

Unknown Speaker 43:06
I actually think we might have the opposite problem. In that, like, I think biology, like at least now, biology is like the the the new hotness? I gotcha II look at, I don’t know, it’s like, I try to keep my pulse on like, what are the smart people interested in it? Like, they’re all interested in biology. And it feels like, I don’t know, for me, at least I ran across. So many people who went like originally went towards physics, but then sort of like, realized that it was a good show. And then they like, went into biology or finance or computer science. So I think the problem might actually be that, like, we don’t have enough smart people in physics, because it is, it’s become such like a esoteric

Will Jarvis 43:57
mess. Right, right. Less applications. And more. Yeah, okay. Essentially. That’s really, yeah. I don’t know that like it. It’s. Yeah. Very cool. So I wanted to move on now. So you studied medieval history in mechanical engineering? Yes. So you’re talking about, you know, fields that don’t have a name? Well, this one has one now, when you think about Clio dynamics, or in Peter turchin, and things like yeah,

Unknown Speaker 44:25
I, I want to like it. Like, you know, it’s like we all want psycho history to be a thing, right. But as far as I can tell, it is no more a predictive science than story and pattern matching. Like, it’s like,

Unknown Speaker 44:52
I

Unknown Speaker 44:53
so I’ve not studied it deeply. So I’m not well, well qualified to sort of opine on it. I haven’t read the books. I’ve read several those aren’t goes there, Fred, sort of commentary on it. But But as far as I can tell, like,

Unknown Speaker 45:07
it’s,

Unknown Speaker 45:09
it’s for the most part, it’s just like narrative pattern matching, similar to any other kind of narrative pattern matching, and like, there’s value to that. And I think that that’s, that’s the thing is like, I don’t want to denigrate that, and I think that the look, what we don’t do is we don’t actually assign enough value to like non scientific knowledge. And I think that there’s, there’s probably a lot of non scientific knowledge that he’s poking at. And it’s worth understanding. But then sort of like saying that it’s it’s scientific, like predictive knowledge would also be incorrect.

Will Jarvis 45:43
Gotcha. It almost seems like the using the science aspect is almost like a sales tool. Right? Like, like, wow, like, Look, what we’re finally making even empirical, these variants running around, you know, just like,

Unknown Speaker 45:54
I might, I don’t know, I would, I would actually take the charitable view and say, like, I think that they actually do see it as a scientific. I don’t think it’s just a sales tool. I just, I don’t think I just don’t think that it’s there’s sort of like no theory behind the hypotheses, right? It’s just like, okay, like, let’s match some patterns, and then create some post hoc explanations for why we see these patterns. Gotcha. Yeah, what do you think?

Will Jarvis 46:25
It’s tough to tell? I mean, it’s really, like, it’s so complex, right? I mean, at the end of the day, and I do applaud them for trying, I mean, I think it’s a it’s a great thing to try to do.

Unknown Speaker 46:33
Yeah. And it might, it might be cut, like, Look, that’s the thing is, like, I think that the like, we we want to render these like immediate judgments, but like, like going back to Newton, and, and like, astronomy, like, like and alchemy, right? Like, yeah, like, we got chemistry, because we had a bunch of people, like, trying alchemy for hundreds of years. Right? And if, like, if we gone and said, like, Oh, you shouldn’t do alchemy, because it’s not science, then we’d like, which would have been correct. Right? Exactly. We would never get chemistry. So I guess, like, my, it’s like, I occupied this weird space where it’s like, I both want to say like, no, it’s it’s not really science, but then that doesn’t necessarily mean that we should denigrate.

Will Jarvis 47:21
Right, exactly. Makes a lot of sense. That’s cool. So I had a fun question for you that I did. So David, we had david Friedman on and I asked him a similar question. The question is, is what most lay people not understand about medieval history? Oh, boy. I think the the sort of, there’s one,

Unknown Speaker 47:47
I’ll give the, the, more like, the more common like answer, I think, would be that people, people regularly pretty, like assume that sort of like nothing happened. And then I’ll write like that, it was just sort of like, you know, it’s like, Roman Empire declined. And then there’s like, a whole bunch of people like digging around, they’re hitting each other with swords for, like, for, you know, almost 1000 years, and then the Renaissance happened, right. And so it’s like a, and it’s like, actually, people, like they did a lot of inventing, it’s like, they figure out how to do windmills. And like, actually, a lot of philosophical breakthroughs happened in in the Middle Ages. And so I think that that’s, that’s one. I think, the really interesting thing that I love about medieval history is that

Unknown Speaker 48:41
people,

Unknown Speaker 48:41
I think, actually take this really charitable, like, this plate thing. They think this way that that really comes to their hearts. And I think that people in the Middle Ages thought like we do, like, it was just like, like, it’s just like, they’re just like us. And I think that that is actually a misconception. Like people literally, like in their, the core of their beings, they thought very differently from us. Like,

Unknown Speaker 49:07
like, like,

Unknown Speaker 49:10
I think, like, relics are a really good example of this, where like, relics are like, literally, it’s like a body part from a seat, right? And they’d like carry it through the streets. And people would like, flip out and get really excited. And like, think of it as if that saint was literally like walking through their street. And like to the modern interpretation would be like, Oh, well, they’re really excited about it because other people are excited about it, or they’re really excited excited about it. Because it was like that with like the like getting to like have a party with the saint. And it’s like no, like even you read the primary sources as far as like, as far as I can tell it like they actually just thought of it as the saint was watching. Through the street. And so I think that that, yeah, like that’s it. It’s really interesting. And it’s like, I like to say that, like other people are both much more similar to you than you expect. But at the same time, they’re like, much more different. You can’t even imagine.

Will Jarvis 50:20
That is. Yeah, that is incredibly interesting. I’ve had that that thought. So I really like to read about the Civil War. And I especially like to read newspapers that I read now, like, how did they report? Cool? So like, I read about the economist, the Atlantic, and the New York Times, were all like, you know, very well established, they have archives. Yeah, go back and read these things. And like, what people thought, and it’s just, like, incredibly bizarre to like a modern part like that, you know, does that make sense? Like, it’s absolutely accurate. Like, it’s like, it’s hard to grok how different we thought about these things.

Unknown Speaker 50:57
Yeah, I guess. Exactly. Exactly. And I think it’s a really humbling thing about it, for me, at least is like, well, in like 50 or 100 years, like how bizarre is the way that we think about things now going to be? And it sort of gives you a little bit of humility, where it’s like, I feel like some people kind of have this attitude that it’s like, okay, like we’ve like, like, the morality that we have now. And like, our priorities are like, the correct thing. Right, like, well, maybe, maybe, exactly.

Will Jarvis 51:29
Well, and even even within our lifetimes. I mean, like, gay marriage was not like a majority opinion. You know, like, and that’s almost, It’s unthinkable nowadays. Yeah, exactly. For me and everyone I interact with. And that’s just bizarre. That’s just bizarre to think about. And that’s a short time domain. That’s not, you know, hundreds of years. Yeah, that’s like our lifetimes. Yeah. That’s weird. Interesting. So So Ben, I have one last question for you for today. Totally. So So what are the next 10 years look like for you in working on innovation systems? Well, it’s a big

Unknown Speaker 52:03
question. So I can tell you what, let’s see. Um,

Unknown Speaker 52:12
so

Unknown Speaker 52:13
I will give you the optimistic scenario that I’m shooting for Hawaii. And obviously, like, this could very easily change. But the sort of vague gameplan is, I hope to put my time and money where my mouth is, and try to create an organization that sort of like riffs on the ARPA model. Like, that’s nice, what I hope to do. And, obviously, that can, like, what that ends up looking like, is sort of, like, can change, but like, my, my working hypothesis right now, is that it’s, like, create a private organization that does revolve around this, this program manager driven model. And sort of use it to both enable more, more awesome stuff, but also, hopefully, sort of, like, demonstrate a new institutional model. Because I, sort of unlike startups, I don’t like I don’t think that it’ll scale, like to the point about, like, DARPA, I think one of the things that’s kept it working well is that it’s small. And so So, hopefully, it’ll be like, okay, like, we need many more of these things. We need many more people doing that. So and hopefully, that I will still be working on that in 10 years. Because I see it as a multi decade project.

Will Jarvis 53:57
Definitely. Yeah. And what do you see as the biggest challenges to making it happen? Who boy I think the big bake it like that. So very abstractly, I think the biggest challenge is almost like surviving long enough without being pushed into something suboptimal, like into sub optimal form, chasing a product rabbit hole or something.

Unknown Speaker 54:39
Yeah, choosing a product rabbit hole or becoming like a consultancy, or is like sort of without being like distract becoming distracted. So like surviving long enough to actually sort of like, become self sustaining. Like I think that that’s, that’s the Like to become a self catalyzing thing. I think that’s that’s one challenge. And then I think the other is like staying really honest towards, like, doing awesome, like awesome stuff without like drinking your own cooling. Right. Yeah, I think that those are like the obviously because visually challenged I like challenges like, like how do you make money work like, right make the organizational structure work. It’s like these are like, huge challenges, but I think they sort of all are sort of kind of the obvious ones,

Will Jarvis 55:39
right? Like, yeah, yeah, you want to think about but that’s that’s that’s really important is how do you keep the mission at the forefront? Yeah. Throughout all the distractions, right.

Unknown Speaker 55:48
Yeah, exactly. But it’s like, okay, like, Well, you know, keeping the lights on.

Unknown Speaker 55:52
Exactly. Keep

Will Jarvis 55:53
that bad, right. Yeah, exactly. That’s cool. Well, Ben, thank you so much for coming on. Where can people find your work? And where should they go looking for you? And

Unknown Speaker 56:03
yeah, I mean, I’m, I’m on Twitter. I’m sort of at at Ben underscore, Reinhardt. And then my website is just my name is Benjamin Reinhardt calm. Those are sort of the two places on the internet that you can usually find me. Very cool. Well, thanks, Ben. We appreciate it. This is great. This is super fun. Thank you for having me on. Had a great time. All right, take care.

Will Jarvis 56:31
Well, that’s our show for today. I’m Bill Jarvis. And I’m will join us next week for more narratives.

Transcribed by https://otter.ai

Join the discussion

More from this show

Subscribe