Preview Mode Links will not work in preview mode

Idea Machines

Jul 9, 2020

A conversation with Ashish Arora about how and why the interlocking American institutions that support technological change have evolved over time, their current strengths and weaknesses, and how they might change in the future. Ashish Arora is the Rex D. Adams Professor of Business Administration at the Fuqua School of Business at Duke University. His research focuses on the economics of technology and technical change and we spend most of this conversation focused on his recent paper: “The changing structure of American Innovation - some cautionary remarks for economic growth.”

I tried an experiment this episode and wrote notes on the paper before the interview

Key Takeaways

  • Ashish introduces a useful framework by breaking the innovation world down into four players : academia, incumbent companies, inventors, and government and then look at how their relationships evolve over time.
  • The current innovation system is well equipped to enable new products with large technology risks and almost no market risk (like new cancer drugs) or high market risks and almost no technology risks (like most software) but falls short in between those two extremes.
  • A fuzzy one but it’s important to marinate in the constant complexity of the answer to ‘How does technology happen? ’  


Ashish’s Home Page
Ashish on Twitter
The Changing Structure of American Innovation

My notes on the paper
Steve Usselman’s Website

Transcript (Experiment and automatically transcribed)



just to start us off, , would you give a summary of the paper? I'm going to direct everybody to go read it, but just for people who are, are listening, like what, what do you think are the key things that you would want people to take away from reading your paper? So the paper itself is descriptive, but our objective data is to, to make, make one argument, which is that the way in which innovation in America is organized?

Has changed over time. And there's a sense in which the system we have now is closer to what we had say at the turn of the night of the 20th century. So, you know, a hundred years [00:02:00] ago there are important differences. So that's, that's one from a descriptive point of view. There are important differences too.

And we, you know, we can talk more about that, Ken. The part, which I think is, is most interesting. And perhaps also most speculative is, you know, two things. One, why has, why, why, what, what caused this change? What caused this system to evolve? And the second is, well, you know, is it good or bad? And you know, what, what might, what should one do about it?

What could we do about it? , and I suspect we would spend some time on that as well. Yeah. I thought the, the dividing up the paper into different areas was, was really important., and so actually, would you say a little bit more about how,, the way that innovation is structured now resembles the way that it did at the turn of the 20th century?

[00:03:00] So if you think let's start with today, right? If we think about today, we have the, the. Big tech companies. , but most people would say, you know, if you think about the innovation system today, we have sort of three sets of players, maybe four, we have the universities where, which do a lot of the research produce a lot of the fundamental knowledge and importantly, a lot of the, what economists call human capital people that, that do it.

so that's one. The second part is, is the startup community, right? The startups and the VCs that fund them and all that kind of stuff. And the third are the firms, the, the incumbent firms, as we call them in economics, the peanut, the Googles, the Facebooks, but also the IBM's Microsoft and so on. And these, these are the different components.

And if you go back to Adam Smith, He talked about a division of labor as being the quintessential aspect of capitalism. That [00:04:00] capitalism is this relentless force towards specialization. And what we have, you might think of it as a division of labor in innovation there, the universities that produce the research, the startups that take it and make it more commercially applicable.

And then the incumbents that apply it. If you go back. Say two 1860s, that's kind of the system we had. We didn't have the universities, but we had independent inventors and we had people that backed them. And then those inventors would sell that inventions for the most part to companies that were producing, you know, early ones were railroads, for example.

And so there's a sense of, you know, in that sense, it's similar. You could think of this as a splinter or a fragment system. I prefer to think of this as, as specialization and a division of innovative labor. Does that make sense? Yeah, definitely. I think so. Something that, so I completely agree with that.

, those similarities, the thing that strikes me, that's [00:05:00] different between that, like the technology then, and the technology now is. Sort of the level of complexity and the amount that it takes to integrate it. Like something that I noticed about, , 1850s technology, and maybe this is this, this might be like a cognitive bias where it's like a fish in the water sort of thing.

But you look at like patents from 1850. And like, you could, you could take that. You could take that patent and you could like build the thing., whereas now. Everything is just, is so complex. And like, literally, if you like, even just like downloading software from get hub and try to get it to run, sometimes doesn't work.

and, and so do you, do you think that that comparison breakdown at all there, you know, that's a fair point and that I've struggled with it? So, so there's a sense of what's surely things are much more complex now than they were earlier. , but, but let me offer you. , a sort of a counter example or two, please.

So one, if you think about one [00:06:00] complex industry of the time was agricultural machinery, right? Those mechanical devices were complex and people did, , innovative parts of it. , and at some point, you know, the whole system became integrated. You can just sort of bolt on stuff. The second sort of, probably more compelling one is, is, , the railroads, which if you think about as a technical system, we're quite complex.

And, , Steve  who's at Georgia tech has written up. He he's a historian of study science. He studied this extensively and I'm persuaded by his work that this was really complex. , but somehow the railroads managed to integrated. While, still relying on independent inventions. So if you think about track switching, these all came from different, different parts of, , you know, different people in different parts of the co , of the country.

And, and they didn't really, the railroad companies themselves didn't really have a function whose [00:07:00] job it was to, to, to develop these innovations. This somehow had got managed. Yeah. So, you know, I mean, I find it depending on which side of the bed I wake up, I either agree with you or disagree with you. But yeah, I think the trick with all of this,, that I think is fascinating is that it's so multi causal and so nuanced that it's, it's very, it's tough to say like, okay, like this is.

Exactly the same or exactly different., and so I, I think that conversations like this are actually really important for sort of exploring that nuance., actually like just something that I'm wondering about the railroads is, , my, my sense of modern corporations is that they are very hesitant to integrate.

External systematic change. Right? So it was like in my, my mental model, if we've had a railroad today and someone came and said like, Oh, I have this, this like, great way to change the way that you do tracks, but you need to [00:08:00] do all your tracks this way. It would never adopt that., is, do you have a sense of whether there was like a cultural difference or a good point?

I'm not an expert on this, but again, relying on Steve's work. Steve Musselman's work. There's an interesting case of, , of breaking, you know, when you have a, when you have a locomotive and you've got these, these are these bogeys that are coupled, how do you stop this thing? And so this was a complicated thing and it was, it was a system that had to be installed, , in, in, in all the, all the cars and the railroads were.

So, so there's a sense in which they were very open to the system, , and Westinghouse. Was that was the guy who was one of the people who came up with the whole system. There were others who came up with different ways of accomplishing this. And the railroad said, fine. , you know, we'll, we'll take it, but we want to do it.

And Westinghouse said, no, no, no, I'll supply you the whole system. And just, you just put it in. And there was a lot of friction around that and, and [00:09:00] eventually Westinghouse prevailed, , thanks in part to his, his patent position. And his willingness to take the railroads on. So,, but to go back to your big question, is there a cultural change?

I mean, surely there has to be right. And we were talking about 150 years, right? Yeah. But you know, that, that particular axis. Yeah. I suspect, I suspect that that all particularly since companies now have an autumn D function or an engineering function, that's that, you know, build certain. Builds up such certain sort of preferences or biases or, or views.

It would be hard to adopt something wholesale from the outside and give up what you have internally. If you didn't have such an entrance function, it might be easier. But you know, I'm really speculating on this one. Yeah, absolutely. That's, that's what we're here for. The, the, like, this is not, , we're not doing any sort of peer review or anything.

and, and so I guess I, another. Big [00:10:00] theme that I was wondering about that you didn't. Like, I feel like you hinted at, but didn't quite touch on in the paper was sort of the nature of the technology in these different areas themselves, like, you know, , late 18 hundreds, you have a lot of sort of mechanical inventions and then sort of giving way to chemistry and then electronics,, and then eventually software.

and, and do you, do you have a sense of whether the, the organization of American innovation. Was Al was like, which way does the causality run? Yeah, that's a good point. I mean, so, so, so far I think it is, I think you've got something really important there, which is, it may be that mechanical systems are maybe designed based systems have this kind of a, are different from more integrated systems, like, like a, you know, if you think of.

, a modern chemical process, which is highly optimized in many ways, and everything [00:11:00] is, is interacting with each other. So, so the shortly our differences, and you could make the argument that, that sort of the mechanical systems were more amenable to sort of bolt on parks. Right? You take this part out of the, the, the agricultural machinery and you bought a different type of, , part onto it.

Yeah., A variant of that, which is an argument that, , again, a historians have made economic historians have made, which is the one difference from, from the mechanical technologies of the late 19th century and the, the chemical and electrical technologies of the early 20th century was the latter of a much more science-based and your independent thinkers, you know, had much less of a, , The opportunities were much, much less fruitful for just the tinkerer, the famous Yankee in January, which, which, you know, was that was irresponsible for, for America's [00:12:00] rise to riches prop, you know, in some sense, had, had, had, had to change and evolve to accommodate the new science based industries.

And I think that's probably true. And that may be one reason why. Companies like DuPont had to start doing some of the, the inventing themselves and to, to bring some of this insight there inside the phone. That's certainly, certainly one possibility on the other hand, you know, I'm sorry, this is going to be on the one hand, on the other hand, that's amazing.

On the other hand, think about a petroleum refining, which is. We started out as tinkering, but eventually had a very strong science, scientific and engineering base. Yeah., some of the most far reaching and inventions were made by, , independent inventors, but by a guy called interestingly enough, CP dubs and I've read, I can't [00:13:00] verify that C stands for carbon and P stands for petroleum.

So the guy's name was carbon petroleum dubs, and he came up with a dubs process. And the, that led to, to the technology that is that's used in pretty much every on refinery that you can think of the, the platin , what's called the platform. Other platinum reforming technology uses a platinum catalyst.

So, you know, there was lots of room for independent invention., even in these new science based industries, by the way, dubs was competing with standard oil. Wow. Standardize. It had the song process and this guy that the modern day company, you can look it up. It's called UOP universal oil products build itself as the supplier of technology to the independence, the independency of being the independent oil refining companies, independent of the standard oilbrella.

Yeah. Wow. And, and [00:14:00] so yeah, the, the, the relationship between like a science and tinkering is. I feel like there's, they're, they're like the people on the science side and they're the people on the tinkering side. But my, like the hypothesis that I've sort of been coming more and more towards is that it's almost, almost like a cycle where like, everything goes through cycles where it's like very tinkering heavy and then very science heavy.

And then maybe back to tinkering heavy,, depending on, on where it is., and so. I think what's, what's interesting to me now. So like let's pull to the present day is that the structure of the American innovation system feels to me at least very geared towards software, right? Like this, the whole started started a software company in your garage.

You have like these really cheap startup costs., You know, high, high capital expenditures, low operating [00:15:00] expenditures,, really does seem to lend itself to venture capital, , acquisitions by large companies, sort of this externalized R and D model that you talk about., what I wonder is like, have you looked at how, but at the same time we still do have all of these other industries, right.

And. At least to my eye. It feels like they, that model, which is really good for like the call it the like the hot, or like most top of mind industry then gets applied to all the other industries., is that like, does that ring true to you? And like, did that sort of happen in other places as well, where you would see like when corporate labs started rising up.

Get the corporate lab started get applied to industries that previously didn't need them. So, so, so let's, let's break this question up into two. One. Is, does the VC model work elsewhere? [00:16:00] It's certainly being was started elsewhere., I think the other place, what it arguably works is is, is biotech, which is a very different kind of sector.

Right. Very science heavy. Yeah., Capital-intensive as heck, , at least in terms of sort of paying for, for equipment and reagents and people., and it, I would say on the whole, it works well there as well. So it works with two very different, you know, almost like two extreme sides of the economy. Yeah. I think, and you know, I, I want to be cautious.

I think it, it sort of breaks down in the middle. And we have a way of thinking about it that it's, it's speculative at this point. , so, so, so that's the answer to your first question is, you know, yet it works at extreme ends of the spectrum is sort of breaks down in the middle. You know, if you think about materials, technologies, , energy, climate change related stuff, it's, it's difficult.

[00:17:00] I mean, we haven't really seen very much coming out of it., and Peter deals sort of famous quote. Oh Quip, you know, we wanted flying cars and we got 140 characters. It is, , it's in a sense, has it has a element of bitter truths there that the system, for whatever reason, hasn't really worked in the middle to go back to the other, you know, how did it work earlier?

I suspect we didn't have, you know, professional investors investing other people's money, which is what VCs are for the most part. Right.,, But we did have people who, who backed independent in ventures., you know, if you think about, , I, I spent a long time at Carnegie Mellon in Pittsburgh and Alcoa was a homegrown company and it was, it was funded by, by wealthy individuals today.

We will, we might call them angel investors, but really that, that involvement went much farther than, than a typical angel investor would do. And you, there are lots of other examples. Well, you know, the, the [00:18:00] people who backed Tesla, for example, the Nikola Tesla, you know, Westinghouse back, back to him out of personal funds.

So, so we, we had, we had people willing to, to, to, to back in independent inventors., you know, obviously things were never quite the same, you know, history never quite repeats itself, but, but yeah, there's certainly, you know, you can see the, you can hear the resonance. Of the past in what we see today. Yeah.

I think the, where I'm, I'm interested in, whether you could make sort of a broad sweeping statement that the, and, and I, and I realized that this is like a very, sort of like a big statement, and I'm not asking you to like, endorse her undergoes, but like, what would you say about the, the, the thought that,, the structure of the American innovation system.

Sort of follows what's best for the most profitable, [00:19:00] , industry at the time, and then sort of applies itself to all industries regardless of applicability, like at that time. Yeah. You know, so it's, it's, it's, it's, there's an interesting point. I think I would put it in a different way, which is it's certainly the case that.

If, if you're interested in, in the VC type model, there are a couple of sectors, you know, the VC is like a particular model and they're willing to go to go for that. The thing that's striking to me about the American system. Has been its diversity and its, you know, the incredible diversity, , and the, , willingness to experiment in many different forms.

So even, even within the VC sector, you'll find VCs who are specializing in science-based. , thing, , you know, startups, they're not whether it's showing the, that they're sort of, they're saying we won't follow the heart and just do you know, SAS, or we just do, you [00:20:00] know, B to C companies or platforms or whatever it is.

You can find people to back. You pretty much, no matter what you're doing, , maybe not, not enough, , in some sense to meet societal needs. So I would say, in fact, it's the opposite that the American system has been very good in terms of diversity in large measure because of its scale. I mean, America is in some sense, , and this is a tangent, but it is just a giant exception, right?

It's a continent, which is one which has had a unified currency, a largely unified set of rules for commerce, , for trade. A common system of law. It's, it's, it's really quite, quite amazing, , what, what we have here. , so for countries who want to emulate America, I always have this caution is that is only one, you know, , , and, and the fact, you know, unlike Yoda, it didn't really have to, to kind of reinvent something from all plot and wait for [00:21:00] 200 years.

I mean, I'm overstating it, right? I mean, it's an important point. Right., so, so we have this diversity, there's a sense in which the VCs, you know, a lot of them are chasing the trends as you talk as you, as you say., and perhaps for good reasons. I mean, maybe this was, this was getting us off point, but maybe if you think about it, right, , we're.

Work w America is a really rich country. And, you know, it seems odd to say this in, in the situation we're in, but we pretty much solved a lot of the basic problems that signed some technology could solve., you know, if you go back 250, 300 years ago, the big problem was getting enough to eat, fighting off jumps and microbes, and just basically the sheer drudgery of daily living.

And for the most part. Again, I don't want to sound, , like, like a, like a nut, but to some, to, to, to a large extent, we've solved [00:22:00] those problems. So if I, if I look, I have teenage kids, I have a, you know, a kid college rising, rising senior, and I look at their lives and their lives are so different from what I grew up in India.

And it strikes me that that biggest problem is, is boredom. That we've, we've ultimately reached the stick stage in, in, in, in America where people people's people have to fight the boredom. And you could argue that many of the so-called innovations that the VCs are funding are a solution to that problem, which was the problem.

How do, how do, how do you stay off boredom? Yeah, I think now, no, it's my turn to do the on the other head, which is so I, I actually, I completely agree with you about how far we've come and how, how many amazing things we have. And I like don't, , I don't want to understate that. And I think that people do, I think that people are not [00:23:00] grateful for like what, how different it was even a hundred years ago.

and yet I think my. My point is, or my hope is that there's so much more to be done. Right like that, that it's not, we're not at some tapering off point. Right. But ideally there's, there's so much more to be done. So ideally not only would we not worry too much about bacterial infections, , but we would also not have to worry as much about viral infections.

Right. And instead of going like really the maximum speed limit being, you know, Couple hundred miles per hour on an airplane, it would be, you know, a couple thousand miles per hour on a rocket ship or something. So, yeah. Well look, I, I completely agree with you. I'm, I'm, I'm an optimist and we will find that the, the challenges we're dealing with now, to some extent we've created, right?

If you think of the big challenge we face or a big [00:24:00] challenge we face, which is, which is climate change, it's, it's, it's something that essentially is a fruit of our own success. Yeah, that the arts can now support so many more people. It's, it's fundamentally changing the earth itself. Right. All of the fossil fuels that took so to, to build we're we're, we're consuming and we're dumping all the carbon back in the, in the environment.

So to CO2 back in the, yeah, that's a tough one. It definitely is. And so I guess actually sort of looking to the future,, you, you touch on this sort of lightly in the paper, cause it's mostly focused on, , the, the history and sort of like how we got to where we are today., where do you see sort of like in your mind, where is the American innovation system?

Sort of like less equipped to handle sort of featured things then [00:25:00] maybe it could be. So, so that's, that's a great point. And, you know, we've between my Cortez and me, we've had a, , , , I would say a very spirited debate on both this and therefore what, what might happen., so, , So I'm not going to represent them.

I'll represent my views. Oh, okay. Well, if you, if you, if you could like mention their views as well, that would be amazing. Right? So, so, so let's, let's start with sort of, you know, where we might be going. One view is, and th their view is in some ways market forces or the profit motive. Has entered so deeply into the innovation system that it's taking us away from, from pressing important problems to, as I said, you know, for example, solving the problem with bored teenagers.

Right? Okay., so this, this is one view I don't disagree with, you know, [00:26:00] when we are in terms of what were the, what big parts of the innovation system are doing, but. The question is, you know, what is it? Is it, is it profit motive or something else? You know? Cause every it's, it's natural to look back at, just roughly say the period between 1930 and up to 1980, what we had these, you know, the DuPonts, the GEs, the Kodaks the IBM of course, the bell labs, Xerox park, you know, one after the other, these companies great companies that did great things,, did one for themselves.

What also did get, did great things produce, you know, fantastic, , innovations., you know, there's a sense in which people want to go back and sort of go back to that golden past, , in, in many ways, , really possible. And my view is that's just not going to happen that, you know, Much for whatever it's, it's, it's far, there's a sense in which we're not, I don't think it's possible to go back there.

[00:27:00] So this is the system we have for, for better or for worse., which is the, you know, they mentioned the universities, the startups, and then the incumbents. And the question is, , where might we go and what, what might we be able to do with it? I think that this kind of system could be improved., You know, if you look at the current pandemic, , which is an interesting case in point, we find ourselves hopelessly under prepared.

And if you look at, for example, the CDC has guidance on what universities should do to, to reopen the CDC does not recommend at this point,, widespread and regular testing, which I find absolutely. Definitely it's absolutely baffling. I do it baffled given that we really don't have any, any, any prophylactic, any way to prevent.

If we don't have a vaccine, we're not going to get a vaccine for the next 18 months, no matter who and whatever, you know, on a widespread or wide-scale. Yeah, we [00:28:00] don't have a cure. The only thing we could do is to sort of test and isolate and prevent people. , you know, from, from infecting others and the fact that this wonderful innovation system has left us six months after this, why this was first discovered, why are the CDC is still not prepared to say you should test and test regularly?

I suspect it's because our testing capacities, woefully inadequate. That's the only charitable reason I can, I can describe as to why they're doing it. Yeah. But it's, it's, it's, it's a huge, I think we're putting thousands of lives at danger, , because we haven't developed this, this ability to deploy, destined, which uses technologies that by and large exist, right.

You're you're using PCR based antibody testing or, or automate [00:29:00] testing, whichever. I, you know, I'm not an expert, but certainly those technologies exist. We know how to do these things. And the fact that the richest country in the world, the technically the most sophisticated country in the world is, is unable to deploy.

It tells us something about the innovation system that I think is not flattering for the system. It's, it's failed us in important ways. And I suspect one could make a similar claim as regards climate change. I think the system is failing us., what I think we will have to do is imagine a more constructive role for the government and perhaps, , private philanthropy as well, that can fill in some of the gaps that the cotton system leaves, where we can have more of our minds, the bright, wonderful minds that America produces on that it attracts.

Use to S you know, employ to solve what I, at least I consider to be more pressing problems and [00:30:00] more important problems., and, you know, we, we, we could, we could talk about what, what shape those things might take, but at least, you know, at a high level, that would be my answer to your question is I it's, it's, it's a great system.

There are some weaknesses that need to be fixed. Yeah, I, I would, I would, would you mind sort of digging into, into those weaknesses? , cause I mean, I have, I have my own opinions on that, but I 100 w yours and I can react to them. Okay, sure. So I think mine actually, , you, you, you illustrate this very well in the paper where you talk about.

I believe it was a DuPont. I think it was DuPont, , tried to acquire, they, they bought the patent, , for, I think it was synthetic silk,, yes, Korean and, , they just, they couldn't get it to work. And so they, they eventually needed to actually bring the people in and start doing things in house. And that [00:31:00] was one of the reasons that they started, , Like that, that corporate labs took off was because,, there was, there was like a lot of sort of integration work that needed to be done.

And at least in my mind, we've sort of returned to. That weakness now. And so I see at anything that can either, , sort of stand on its own or be very modular and quickly,, become part of a larger thing. , it does well, like th the system does really well for it, but,, technologies that sort of are like improvements to systems or replacements for systems.

they, they sort of wither on the vine. Yeah, no, I think that's right. I mean, one way to think about this is that, you know, if you go back to where the VC based system does really well, you see the two opposing ends. If you think about software, I would argue to a first approximation software is mostly about [00:32:00] figuring out what consumers want, will enough people buy it?

Do I, can I find them? So if you think, if you see what, what, what a lot of VC money gets spent. It's not in, in solving the technical problems. It's in solving the commercial problems. Yeah. How do I find the market? What is the market? You know, how do I sell it? And, and, and I did those things out, which is why VCs don't care about whether you make money, but they're care.

They care about top line growth. Can you acquire the customers? Can you acquire them cheaply enough? So the system is well tailored to that. On the other hand, other end, if you think about the biotech side, The problem that is not do customers want it? Yes. We want a cure for, for, you know, for Colbert. We want a cure for cancer.

The big problem is the scientific and technical problems. Can we find something that will do it? And once again, I would argue the VC system for the most part is, is, is well suited or VC system. The startup [00:33:00] based system is well suited to that. Yeah. You have patient capital people willing to put in money where they will not see.

The final outcome for another 20 years, people are, you know, are spending money in, in how you could make a human beings immortal and you know, just think, think about, you know, there's no, there's no doubt. There's a market for that. And we also know, we also know who, who, who will want it. Right. I'm in that category.

I like living too, right. It's the stuff in between it, where these two things interact and important place where, you know, if you go back to the question of, of, , nylon, let's say our, our rail on the question of what are you going to do with it? So you have a new material, what should it be used for? So did we used to make.

Make underwear for women, or should it be used to make parachutes, could it be used to make a billiard balls, which is one of the earliest, , not for nylon, but for something else. [00:34:00] There's a myriad of possible ways so that, you know, you could have many different kinds of markets, many different kinds of consumers, and depending on each, you would have to take this and get to a different price point, have different sort of performance characteristics.

And all of those have considerable technical uncertainty. So for example, when nylon first comes in, you can dye it, you can call it it. So it's this really crappy looking, Dell gray kind of material. Well, who would want to buy stuff? You know, clothes may not have that if you could only buy gray, so you have to solve the problem of, okay, how do I diet?

But that depends on what you use it for. If you go to make it to use rope, you don't cash, right? Right. So it's when these two things need to interact in important ways and that are important, you know, funding decisions that have to make, should I invest in, in changing the performance characteristics of this material?

Well, that depends on what market I'm going after. Well, which market should I go after? [00:35:00] That's when I think the system does much, you know, it's not clear at all, , that the system is, is, is weak and it, I would argue. , the, the integrated system would work much better. Yeah. That's really insightful. , I hadn't, I, I hadn't put those pieces together before.

and that, that sounds really correct., and so do you have any thoughts about,, How instead of like, I guess without saying like, Oh, let's just make corporate labs again. , do you, do you have a sense of what steps we could take from where we are now towards, , a system that supported those kinds of innovations?

Yeah. So I would say that they're probably one just to come back to, , the question of, of the role of government. One another place where the current system works breaks down is when you [00:36:00] have what economists call externalities. Whereas, you know, it's hard to capture all the benefits that you're producing and some cases, those benefits could be, could be quite significant.

The government is already involved in one sector in a big way that solves this, which is a university sector. We we're now, you know, , , the university sector just would not survive. Without the NIH and the NSF funding and the other source of the funding over the last two decades, maybe three, I certainly three decades.

I think that the system has the government system has become more short term. I think they're willing to take fewer bets, , perhaps for good reasons. , but, but I think. If you think about many of the things that, that we got many of the breakthroughs we got, certainly after following world war II, the government not only subsidize some of the upstream [00:37:00] research, they were happy to support the training of students and they were happy to, to buy the things that came out at the backend.

what's. Interesting. So let's go back to covert. One of the things that would have been very useful is for the government have also announced early on that they would, they would be ready to buy, you know, 250 million units of vaccines for the next so many years. I couldn't agree more. And if he has to write and test and forget, government buying government is actively hindering.

The use of, of tests, right? They won't allow people to do tests. The FDA has become such a force, , such a destructive force and disrespect, I would say, , it's it's, it's, I'm, I'm shocked and saddened by, by how, at least in this particular case, the FDA has behaved., so, so, so the, but just not to get too [00:38:00] far off on my pet hobby horse right now is.

So the government, I think we need to find a way where, where the government can, can not only solve some of the upstream, scientific uncertainty, commercial, you know, technical uncertainty part, which they're currently doing through funding research and so on, but maybe also try to do something at the, at the backend where they stand ready to, to procure or to, to, to be important customers for, or somehow.

Help mitigate some of the commercial uncertainty through the procurement. So I'm, , I feel like very torn on this subject because so, , to, to expose my biases, like I, I'm a big fan of markets. And so,, I'm very, I buy the arguments that, , government is not. Well, especially when, when it's, they're not buying something for themselves, they're not, they [00:39:00] will always make the best choice about what to buy.

But at the same time, I, I find the argument you just made compelling., do you have any sense of like how to, to balance those two? Yeah, I think it's a good point. So, I mean, The government as, as the user, as a lead customer, for many of the technologies and electronics computing, et cetera, kind of makes sense because I don't know if they bumped the best ones, but they certainly bought a lot of them.

And it was justified by, you know, by the, by the space race, you know, the Sputnik and then the core war that followed. And in some sense, even if you bought the wrong things as the government. In those days, you could console yourself by saying, well, at least we trained a whole bunch of PhDs and engineers.

, and so we, we got something in the body game, right? We, we developed a capacity. We developed a way to test these systems. So [00:40:00] I'm, my biases are the same as yours. I am generally having grown up in India. I'm very suspicious of government. , You know, actively interfering in the marketplace because usually no good comes out of that.

Yeah., but at the same time, it's hard not to look at the, the current crisis we have and say, well, here's one thing that, you know what, I would have supported some sort of government intervention, both on the front end, in helping develop the knowledge that we need to develop the vaccines and the tests.

But also to S to, to at least backstop some of the commercial risk. Yeah., without it, nobody will make investments. There was a very nice article that I read about the, either the Ebola or the SARS epidemic, where some people made significant investments in, in scaling up capacity for, for vaccines and or for PPE.

And then we're left, hanging. Because if the government doesn't, doesn't intervene [00:41:00] actively to buy some, you know, to provide some significant quantity of demand is just not viable. People. People are not going to pay for tests for themselves, or at least the right people will not pay for tests for themselves.

They will pay for tests when they think they're sick. You want the test when they don't think they're sick. Exactly. And most people won't pay for that. You do need, this is a classic case where you do need the government to come in and say, well, we'll pay for it. And within in India, for example, smallpox are more or less wiped out because the government subsidized not only the acquisition of the vaccines, but also the large scale deployment.

Yeah, we also need government to help perhaps deploy this, but given our fragmented insurance and medical delivery system, we need, we need a lot of help both on the regulatory side and on the dollar side to help that. And so while I share your biases, I think there are definitely cases where [00:42:00] we need the government to be more thoughtful and more, more progressive.

Yeah, I think that the tricky thing, so I agree and I think. The tricky thing that I, I don't know how to do is how to decide like where, where where's the line., and that's, I think beyond, beyond the scope of this, but, , it's, it's an interesting, , question that I don't think that we have a good conversation about yeah.

I mean, w w a different way to do it is we don't have to draw the line, , which we can do it on a, on a, as is basis. Right? I mean, part of the thing that I asked. Over time. I've come to the view that we want to have principles, policies, or principle ways of, of, of, of approaching these kinds of decisions, which is fine.

And it's a, it's a, it's an instinct and an impulse that I understand, but sometimes that stands in the way, so, so we can draw the line. Fine. Yeah. [00:43:00] It's okay. Let's do it here. Then when the next one comes along, we can figure out whether the government should do something like this or not, or whether the government and the reality is the government does get involved.

Right. I mean, yeah. So all the nice lines we draw don't seem to stop anything in any case. Yeah. And actually speaking of bare, mid and shifting a little bit, I, you, you sort of outline., the antitrust trends over time. And,, like in, in the paper, , you make the argument that antitrust was pressure to, for corporations to do more internal research so that they could expand markets.

and there's examples. Both of this., but then at the same time, there are also examples of, , sort of the, the technology. Like another reason why people have really liked corporate labs is because [00:44:00] technology also like escapes them, right? Like you see, , the transistor coming out of bell labs and, and to my understanding, they government antitrust also.

Sometimes would prevent, , big companies from going into new markets, even if they were to invent things. So,, can you, can you sort of like walk me through, , your, your argument of. The value of trust. Yeah. So that's a really good point and I should have, I should have thought about that. Thank you for reminding me.

here's a, here's an interesting piece. Just going back to DuPont in 1911, DuPont had a monopoly on smokeless powder. They were the monopoly suppliers of smokeless powder to the us Navy in 1911 in a famous antitrust case. The government went in and broke up DuPont into three parts. So there was DuPont Hercules and Atlas three companies to, to provide, , to provide smokeless powder.

And this is quite remarkable. If you think about right, they took a dominant one form that was doing well. And they said, you know, we're going to break you up [00:45:00] into three producers. Now, those Atlas soon March with Hercules, I believe. And then Hercules' remained as an independent company for many years in similar spaces to DuPont and they had reasonably friendly relationships.

But I think that episode was a singular episode for DuPont because they understood that that ability to grow was going to be constraint that they could not do what companies do now, which is just buy up competitors and, you know, look at T-Mobile and sprint. Right. I mean, it's, we've gone from four to three and I'm not an antitrust experts.

I don't understand these things, but there's a sense in which. If growth is, is an imperative for companies. And if they're going to be constrained from growing, by buying up, , you know, by expanding by, by, through, through, through sort of what we call inorganic acquisitions, there's a sense in which, , they will [00:46:00] have to find new products and these new products could come from startups.

Although there's an antitrust scrutiny there as well. , but it's much easier if they come internally. And that's essentially how DuPont went from being a producer of large, largely undifferentiated chemical products, like explosives and fertilizers into new materials. And they eventually went almost all the way into getting into textiles and then they stopped and they, you know, when they got into nylon and polyester, they said, how far do we go?

Do we want to be producing our own cloth? And they said, no, no, no. We want to stop here. We're going to demand an industrial company. , ATMT while they could not get into merchant semiconductors, they use semiconductors internally. It was very important for them that they needed those to solve their own internal problems and to grow the telephony business.

Yeah. Did, did they,, produce, did they make their own semiconductors or did they purchase them from, [00:47:00] I believe. Hi, I'd been interested. I believe they did for, for quite some time and an IBM similarly. Right. I don't think they were a merchant supplier of semiconductors to others, but they've certainly produced any conductors for themselves.

Yeah. And so this is going to be a naive question, but like, why, so, I mean, I would like, I guess the question is like, why doesn't Bowie, why don't we see a lot of amazing research coming out of Bowie? Because I sort of think of them as. Basically a monopoly on, I mean, it's like, there's like Boeing and Airbus and they're the only, , companies that can produce like giant airplane, like giant complex airplanes.

so if, , the, if like the monopoly profits are what enables like longterm thinking and,, really great corporate labs, like why, why doesn't Boeing have an amazing corporate lab? That's a great point., I don't know, but remember boy was allowed to buy McDonald Douglas,, [00:48:00] and, and words was not stopped there.

And really, , that, I don't know. I don't know the answer, but I think, I think part of it is at some, they, they never either never developed the capability to do this kind of fundamental research. Are, or they do. And somehow it's all tied up in various kinds of government contracts that we don't see, you know, because they had a large government contractor as well.

And maybe the government is not demanding from them. The same sorts of innovative products that it might have demanded of its contractors in an earlier era. Yeah. So, but it's a, it's a great question, Ben. I don't know. I never thought about it. Like why don't they have,, a large corporate lab? No.

Remember one reason we will never go back to these large corporate labs is because they are incredibly difficult to manage insight, a publicly traded company. Yeah. And this was the big [00:49:00] point of disagreement between my, my, my coauthors and me. My view is research inside of a large public company is always a strange animal.

It's always a strange animal., and the only way it's happened is if, if corporate headquarters protects it and nurtures it as the shiny example, here is Microsoft. Microsoft is, had set up under bill Gates, , and more of a world of fantastic research lab, which, you know, they, they nurtured, but, and this is a big part.

If, if you ask yourself, what did Microsoft shareholders get out of that lab? That, that would be an interesting quality. I don't know that it's been systematically studied. Yeah. But looking at it from the outside, if you look at what Gates as successors have done, I think I'm going to get in trouble for saying [00:50:00] this, but I think.

What we are now seeing since probably 2015, 2016 is the beginning of the end of Microsoft research. Wow. That's and I say this because you're that the person was heading, Microsoft research was replaced by two people. Oh yeah. And part of that was, was taken away. And part of what was, was moved into sort of more applied.

so. You know, I that's, why both, I think we'd never go back to the corporate labs, , in a, in a serious way. Yeah. Because it's incredibly hard to manage and it's incredibly hard to justify, you know, the billion dollar, roughly expense that Microsoft research, you know, cost Microsoft. If you think about huge, huge public good.

, but it's, it's, it's no longer justified given that Microsoft now faces significant competition. Yeah. In terms of growth goes back to the externalities piece [00:51:00] again. Yeah. So there's certainly large externalities., you know, if you think about IBM when IBM had to take a potential crisis or even a little bit before that IBM's Watson research lab, essentially God reoriented and pushed more and more into being managed by the individual divisions and the businesses inside.

IBM. And those divisions have quarterly reporting responsibilities, right? You have to justify the capital that you get from, from, from the parent corporation. And it's really hard to say what something you've, you know, investing in now may or may not work. If it works. We might see the results in six years, you know, in terms of the technical and scientific findings and the commercial benefits might be even further down the line.

It's, it's hard to manage that. I, I don't blame it on, on. On on short term as though I don't believe that argument, but I certainly do think that it's this, this kind of bundling between a main activity of a corporation, let's say [00:52:00] IBM, which is to sell computers, to produce, and then to do all this other cool research, which could be relevant, but perhaps not.

But, you know, perhaps someday it's really hard to manage and do that. And I think this is the relentless logic of, of. You know, golf capitalism is we have to unbundle things. We have to specialize. So they go back to, to kind of Smith and say, and as I said, this was the point of contention between the research team that's working on this problem.

So you're just hearing one side of it. They think I'm,, I'm overtaking the customer so to speak. I, well, I mean, I certainly buy your argument, so. I forgot what that's worth., I'll, I'll tell them, , so the last thing I always like to ask guests is,, what is something that people are not thinking enough about that they, they should be thinking more about?

That's a tough one, [00:53:00] not very imaginative sort of guy.,, I would, I would disagree, but, well,,

Okay. No, I don't think I can give a, give a clever enough to justify the advertisements. That's that's fair. That's fair. Well,, I really appreciate you doing this. , this is,, and I really appreciate you being willing to, to sort of like, Almost like play with these ideas., I think that, that people don't do that enough.

And,, I, my, my hope is that by I sort of playing with them, , we can, we can figure out new ways to make awesome things happen. I really hope that, you know, I would really like to learn because none of these ideas are set in stone and it's. I, I thank you for the opportunity to talk to you. And I hope we'll, we'll, we'll learn some more and people will, will come up with, with new and better ways of thinking about this problem.