Monetizing basic research has always been a fools errand. For example going back to1992, research into the cockroach microbiome was still in its early stages. "A key study published that year focused on the contribution of anaerobic protozoa and methanogens to the hindgut metabolism of the American cockroach (Periplaneta americana)." Of course many in Congress would have made fun of this early microbiome research. Thirty years later, built on the shoulders of cockroaches (and other lesser species), human microbiome research is well underway but has yet to lead to significant paradigm shifts in human healthcare. Per your quantum example, there was a substantial lag between quantum physics and smart phones.
An enormous effort needs to go into basic research to build to the point that monetization/societal benefit is achieved (of course along the way a bunch of scientists had fun figuring things out). US government investment needs to keep contributing to this base on which product development can be built. Essentially decide to commit to x% of budget on research and just keep plugging away across a variety of disciplines. No one can know what will pay off, but statistically research will pay off.
Even once the development opportunity is created, the big money goes to the group that solves the "last" problem, not to the first mover. So in my view the most productive ecosystems need to have small risk takers (not invested in the status quo) as well as established players who can buy out and the complete the development of innovative products. Hard to think that any central government (or large corporation) will be sufficient in and of itself to provide a steady stream of innovation. Neither the Chinese or large corporations will fail for lack of trying.
I do have a lingering feeling, though, and correct me if I’m wrong on this but basic research seems incredibly vulnerable to being taking advantaged of in a competitive, geopolitical environment.
Would it not be more sensible for any country, then, to research in regions where theres a large moat? After all, if the US pours 10 billions dollar to figure out that a particular sacred, god-blessed strain of gut bacteria just make depression go away, they’re not getting ahead of China or anyone due to the fact that anyone can now use that knowledge immediately. Meanwhile, China take their own 10 billions to invest in highly specific and monetary research like a specific, hard-to-replicate alloy for EVs or something.
Yep. That's exactly what I was talking about in this paragraph:
(In fact, China’s strategy of stealing breakthrough discoveries and inventions was probably already pushing us toward incremental innovation. Blueprints and experimental results and algorithms can be easily transferred from country to country; in an age when espionage is commonplace and intellectual property isn’t respected, the technologies that have durable moats are the ones that depend on distributed tacit knowledge, like grinding glass, optimizing combustion engines, or operating EUV machines. In other words, if big ideas always get stolen, it makes sense to focus more on the accumulation of a bunch of little tricks of the trade.)
The more China just steals everything the minute it gets discovered or invented, the less incentive there is to discover and invent.
Scientists will still do basic research for intellectual prestige, but governments will have less incentive to give them money to do it.
Basic research is pretty much a global enterprise (even Chinese basic research is primarily in English) so everyone benefits from everyone else's work or in your terms everyone takes advantage of everyone else. But progress is ultimately built from basic research so it is globally important that basic research not go to zero. How much or how little the investment in basic research facilitates the first steps to commercialization is hard to say.
While inventors tend to view all of the value creation in invention, the reality is that most inventions are not valuable and no invention is valuable without development. From a project cost perspective my favorite story is that every good project needs a good idea. If one needs a gross of ideas to find a good idea since ideas are a dime a dozen then the good idea costs $1.20 then we still have all the other costs of the project!
Development is high art all in its own right. In the geopolitical competition the more we elevate that status, importance and compensation associated with development relative to invention, the more competitive we will be. Designing systems to balance the ratio of investment in invention and development is challenging (see the entire history of drug development).
Excellent overview; three quick follow-up questions/comments: (1) in light of involution, how can private Chinese firms appropriate their investments in R&D? (2) in light of consortia sharing, isn't substantial free-riding likely, and if not, are there different legal or cultural constraints at work that prevent free-riding? and (3) what role does competition play, if any, at the R&D stage, and, if present, is it competition among firms, bureaucracies or regions, or all of the above? Certainly the framework for evaluating innovation and incentives (from Schumpeter to Arrow to our recent Nobel prize winners) tends to focus on competition among firms. China's approach suggests either that the economic modeling is incomplete or that their approach will have problems progressing to next-generation products. We shall see.
These are all great questions. I don't know enough about the details of the system to answer any of these yet, but I'll keep trying to find out more. I think because we've spent so little time and effort as a society evaluating different research institutions, we don't have a well-rounded picture of how competition and cooperation really work here. And China's system is very new, and we don't have a lot of details about it from the outside. So it could end up being a giant mess, like many of their industrial policy experiments in the 2010s. Or they could iterate and tweak things and improve incentive structures as they observe results!
Great article on an extremely important topic. I had no idea that China was doing this. I will also take a look at Barry Naughton’s article and the other linked articles, which look intriguing.
By the way, you and your readers might be interested in a new series of articles that I am writing on how research and development has changed over the centuries. It goes into more detail on many of the themes that you mention here in this article.
I would be interested in an analysis of how this picture changes in a world dominated by AI in research, product development, planning, coordination and commercialization. Does this have a "leveling" or "democratizing" effect by allowing smaller teams to pursue end-to-end innovation? Isn't the promise of AI that it is the ultimate vertical integrator? It will know what ends we should pursue better than humans, it will do basic research and applied research better and faster than humans, it will know human needs and wants better and so will be better (and faster, and cheaper) at commercialization...
Came here to look for a comment like this. By reducing/eliminating the barriers between key institutions (universities, labs, companies, etc), they're creating their own intranet. Sort of like the idea that the ultimate point of the internet was to connect all information to enable AI, China might be doing a better job of connecting their institutions so that AI can use it for whatever ends them deem worthwhile. Makes me think that while the US may be ahead (for now) in development of AI, China may actually utilize AI better because there are fewer barriers between information pools - at least within the sectors that their government mandates.
I am not sure yet if research is a "talent" problem or an "organization" problem. There have been frontier labs with the highest concentration of talent that have consistently NOT "won". Xerox Parc, Bell Labs come to mind.
A future lab filled with brilliant AI researchers will still require state capacity to organize.
First, this was a fantastic writeup of something I've been trying to get my head around for a while.
One big gap in my understanding is how competition between regional governments figures into the whole thing. As I understand it, the way China got 200 startup EV companies, now starting to get widdled down to just a few, was from provincial governments sort of acting like venture capital firms funding local champions in a competition to turn them into national champions (and maybe get technology hubs to be located in their territory). At least, that's how I think it works.
It seems like despite the system being largely nationally coordinated with a lot of central planning, there's still a huge element of competition between regions and individual start-ups that helps keep the system more dynamic.
Thanks, I'm glad you liked it! Yep, they're probably relying on regional competition to keep competition in the system. But also A) competition between labs for prestige, B) competition between companies to take the lead on each project, and C) competition between institutions with overlapping mandates.
Fascinating essay re the CAS, I had no idea. I love a good experiment, and it will be interesting to see how this Chinese one pans out!
My first impression is that nationalism as a science motivator definitely works (think Manhattan and Radar Lab projects) for specific tasks. Maybe that is enough to accomplish great things, even if I don't think it will drive 'basic' research very far.
I think nationalism is a motivator, but researchers can't subsist on nationalism alone. They will need intellectual prestige, credit, and individual recognition.
The Nobel Prize is dumb because science is a team effort, but there's a reason we have it -- dreaming of getting a big gold medal that says "you're smarter than the other smart guys" is a huge motivator for people who have spent their lives basing their self-image on being "the smart guy". Japan's policy of having managers steal credit for the inventions of researchers and engineers has probably materially hampered its innovative capacity.
Meanwhile, involution is a huge problem on the downstream side. Companies will participate in this system because they have to, but they may not end up making much money off of it, because every other company is getting the same technologies and getting paid by the government to make the same products. So companies may free-ride and expend less of their own in-house innovative efforts, and in the end China may just end up with CAS labs and universities trying to do all the work on their own to make up for lackadaisical unmotivated corporate labs.
I'm not saying these things *will* happen, but I don't think nationalism will be enough to overcome these problems all on its own!
"Companies will participate in this system because they have to, but they may not end up making much money off of it"
Companies as a whole may not make much money, but you can bet individuals will game the system substantially for their own personal gain. "If I can't get rich with my shares/options I'll just pad my expense account and bonuses, etc."
That will surely hamper the collective effort due to vicious, selfish infighting over who gets what perks and favors. There will of course be periodic purges for "corruption" but the system has built-in incentives to encourage it. If you can't get rich as a team then you can enjoy the facade of being rich as an individual, if you're clever enough to hide it well.
Basic research is consistently extremely valuable, but also produces very little exploitable short-term profit. It seems mainly to thrive in places where some motivation other than "efficiency" is driving the effort. Doesn't really matter if that's nationalism, misplaced pride, or a telco monopoly burning profits to get R&D tax credits. People are curious and if you fund smart ones to do what interests them, (eventually) useful things will happen. The main question is whether the "inefficiency" is politically/financially sustainable.
In the 1980’s ,when we had Japan envy, MITI was seen as a differentiator. How is CAS different? As a capitalist I have learned that governments are not particularly good at picking technologies or companies. Maybe you get some wins but there is a lot of wasted effort and money that could go to the people.
MITI is really just on the production side. China has similar institutions now that are actually even stronger, and they cooperate with CSTC, MOST, and CAS.
This might end up being a case of "too many cooks". And of course there's always the possibility that central planning just hurts here, period. People still debate how much MITI hurt or helped in Japan.
I generally agree with your skepticism. Large government created and/or directed operations tend to collapse or become counterproductive eventually under the weight of bureaucratic inefficiency and sclerosis. OTOH, it does seem like their current innovation system still incorporates a good amount of meaningful competition. It's not just the government picking winners and losers.
I don't fully understand how it all works, but EVs are providing a good example of this competition. I think provincial governments provided seed funding (maybe acting like VCs?) for hundreds of EV startups. Each province was in a competition for national prestige and to get selected by the national government to have a technology hub located in their region.
This created the involution problem, but it also created a competitive process where mostly market forces pick the winners and losers as those hundreds of firms now get whittled down to just a handful.
I think it's possible that people (governments and think tanks) will learn and get better at devising ways to stimulate innovation. Ideas about this in the 1950s were pretty simplistic (build PARC). In the 1980s Japan was trying MITI. China's systems seem much more sophisticated. Even if people don't "figure this out," whatever "this" is, they may get better and better, and the theory that "government can't pick winners and losers" might have been totally correct before, but in the coming decade, it might not be so black-and-white.
really interesting article, with perhaps two lacunae, respectfully. first, such a system is ripe for internal corruption, policing of potential corruption, over the long, would seem to me very likely, secondly, to increase buraurcratic resistance to exactly the type of innovation this system is
supposed to cultivate thus one can foresee a short term period of productive but eventually the system fails, as everyone in it is either trying to grab turf or money, or (perhaps also) looking over their shoulders at internal bureaucratic competitors the well-known Chinese sci-fi trilogy, Cixin Liu's Three Body Problem seems to describe that, although in the context of the cultural revolution if the 60's
Massive room for potential corruption, and a decade ago bribery to get to the front of the line for talent programs / government grants was common. The Chinese government has tackled this seriously, the number of audits / checks and balances are insane. Doesn't mean it's solved, but environment largely works.
Thanks for this deep dive into the China innovation model. Although you do point out that the piecemeal pipeline story is an oversimplification, it may be that the story does not represent how innovation worked in the west for most of the scientific age. I am not sure how you would fit in that framework the story of how Max Planck discovered the basics of quantum theory trying to solve the problem of designing a better light bulb (in competition with the US!) (https://www.ornl.gov/blog/quantum-science-exploring-little-particles-big-promise). Or that one of the main projects of the newly minted Royal Society in London in the 17th century was to solve the Longitude problem (https://mathshistory.st-andrews.ac.uk/HistTopics/Longitude2/), leading to fundamental advances in astronomy. As an oceanographer, I can tell you that the oldest oceanographic institutions such as Woods Hole and Monaco were founded and funded initially in the late 19th century to solve fishery problems, but to do that the scientists had to ask fundamental questions about how oceans work. Finally, there is the Louis Pasteur quote "There is no such thing as a special category of science called applied science; there is science and there are its applications, which are related to one another as the fruit is related to the tree that has borne it”. Maybe China rediscovered an older (pre WW II), more vertically integrated, model of innovation.
That's a good point...I have a pretty clear idea of how America's research institutions work *now*, but a much sketchier idea of how Europe's scientific institutions worked a century ago (to say nothing of *three* centuries ago!).
But I'm pretty sure they had nothing even remotely similar to the planning and coordination capabilities of the modern Chinese party-state. Probably no one else ever has. It's just an open question of A) how much this is reinventing the wheel, and B) how effective any of this will be.
Thanks for your reply. I completely agree that nothing in the past in Europe compares with what the Chinese are doing, and my view is that we certainly don't want to do the same. However, I would maintain that recapturing some of the spirit of pre-war innovation and incorporating that into science institutions would help accelerate impactful innovation here. It is already happening. That is what (I think) Focused Research Organizations are trying to do. I have heard that they don't want to deal with universities because the focus of universities is too far removed from what they see as needed to efficiently solve big technological problems for society. In Nova Scotia where I live, the FRO model has been adapted to deal with social problems. People are trying to change the pipeline.
This was a good overview and primer to the ecosystem. Noah does a nice job of capturing most, if not all, pieces of the innovation "pipeline" in China.
I think it does risk missing the relative "weight" of each piece, however. For example, this top down planned economy vision of innovation only applies to certain critical industries where a) strategic risks for the country exists (so China needs them), and yet B) the risks / capital required are way too large for private capital to take the lead: the "big airplane" project of C919, nuclear power, SEM replacement to ASML, etc.
But much of the other innovation that Noah / the rest of the world is enthralled by comes from a different application of the innovation pipeline. I'm talking about AI models, robotics, EVs, etc. It's still the same pipeline, often starting with a Five Year Plan calling attention to the opportunity, with core R&D advancing in national labs... but the most critical piece is the competitive nature of what Lin Keyu called the "mayor economy". Mayors are measured by commercial output, and they essentially act as VCs - they are the most powerful muscle in the entire innovation pipeline. They put together the focused capital that transforms patents / knowledge into revenue, products, and jobs. They cut red tape, they find talent, they do whatever it takes. When you speak to one of these mayors as an inventor / entrepreneur, the question you'll most frequently hear from them is: "how can I service you to help you succeed?"
IMO, this aspect of the innovation pipeline is what most distinguishes China from other nations, and is most responsible for the recent success in the industries most of you care about.
Great article. I do think that there is a fundamental difference in the world view of leaders in China and here in the US (the broligarchy). Xi Jinping and much of the CCP leadership are intent on reversing the "100 years of humilation." In other words, they are nationalists. Not unlike Churchill's and Roosevelt's leadership in WWII.
This is in contrast to US leaders like Thiel , Andressesn and Musk who envision a world shaped like Neo-stalinist Ayn Rand's fevered dreams in Atlas Shrugged. They have their dual citizenships, and if their national program goes south, its, "Oh well..." (cf. Musk and Doge and Thiel's and Andressen's puerile essays).
Hopefully, Noah will write an essay on China's use of open-source AI versus the US approach. Scott Galloway had an interesting take on that in terms of a marketing strategy and as a possible danger at the heart of the US economy ( 80% of value for 10% of cost leads to market capture and collapse for the Open AI and other US companies that constitute the bulk of the US stock averages)
It may be worth noting that the US has long had an abundance of national labs well beyond DARPA. Surely names like Lawrence Livermore or Jet Propulsion Lab ring a bell. There are many more. There are also Federally Funded Research and Development Centers. Other great names include Lincoln Labs, Fermi Lab, Argonne and more.
For sure, but it's also noticeable how very few of the IP created in these national labs are directly meaningful in the companies that dominate the market / economy today. (Yes, YES, we wouldn't have Uber without DARPA-funded GPS and Internet... but that kind of indirect "tree" effect isn't the direct, high efficiency innovation pipeline Noah is talking about today.)
This is very interesting. It reminds me of what the US did during the manhattan project and the space race.
China (and most of east Asia) have a huge competitive advantage over the US to go in this direction. They culturally already have a collective mindset. The only way the US will do something like this is if they believe there is an existential threat to our country. Otherwise, everyone is out for themselves.
Monetizing basic research has always been a fools errand. For example going back to1992, research into the cockroach microbiome was still in its early stages. "A key study published that year focused on the contribution of anaerobic protozoa and methanogens to the hindgut metabolism of the American cockroach (Periplaneta americana)." Of course many in Congress would have made fun of this early microbiome research. Thirty years later, built on the shoulders of cockroaches (and other lesser species), human microbiome research is well underway but has yet to lead to significant paradigm shifts in human healthcare. Per your quantum example, there was a substantial lag between quantum physics and smart phones.
An enormous effort needs to go into basic research to build to the point that monetization/societal benefit is achieved (of course along the way a bunch of scientists had fun figuring things out). US government investment needs to keep contributing to this base on which product development can be built. Essentially decide to commit to x% of budget on research and just keep plugging away across a variety of disciplines. No one can know what will pay off, but statistically research will pay off.
Even once the development opportunity is created, the big money goes to the group that solves the "last" problem, not to the first mover. So in my view the most productive ecosystems need to have small risk takers (not invested in the status quo) as well as established players who can buy out and the complete the development of innovative products. Hard to think that any central government (or large corporation) will be sufficient in and of itself to provide a steady stream of innovation. Neither the Chinese or large corporations will fail for lack of trying.
I do have a lingering feeling, though, and correct me if I’m wrong on this but basic research seems incredibly vulnerable to being taking advantaged of in a competitive, geopolitical environment.
Would it not be more sensible for any country, then, to research in regions where theres a large moat? After all, if the US pours 10 billions dollar to figure out that a particular sacred, god-blessed strain of gut bacteria just make depression go away, they’re not getting ahead of China or anyone due to the fact that anyone can now use that knowledge immediately. Meanwhile, China take their own 10 billions to invest in highly specific and monetary research like a specific, hard-to-replicate alloy for EVs or something.
Yep. That's exactly what I was talking about in this paragraph:
(In fact, China’s strategy of stealing breakthrough discoveries and inventions was probably already pushing us toward incremental innovation. Blueprints and experimental results and algorithms can be easily transferred from country to country; in an age when espionage is commonplace and intellectual property isn’t respected, the technologies that have durable moats are the ones that depend on distributed tacit knowledge, like grinding glass, optimizing combustion engines, or operating EUV machines. In other words, if big ideas always get stolen, it makes sense to focus more on the accumulation of a bunch of little tricks of the trade.)
The more China just steals everything the minute it gets discovered or invented, the less incentive there is to discover and invent.
Scientists will still do basic research for intellectual prestige, but governments will have less incentive to give them money to do it.
Basic research is pretty much a global enterprise (even Chinese basic research is primarily in English) so everyone benefits from everyone else's work or in your terms everyone takes advantage of everyone else. But progress is ultimately built from basic research so it is globally important that basic research not go to zero. How much or how little the investment in basic research facilitates the first steps to commercialization is hard to say.
While inventors tend to view all of the value creation in invention, the reality is that most inventions are not valuable and no invention is valuable without development. From a project cost perspective my favorite story is that every good project needs a good idea. If one needs a gross of ideas to find a good idea since ideas are a dime a dozen then the good idea costs $1.20 then we still have all the other costs of the project!
Development is high art all in its own right. In the geopolitical competition the more we elevate that status, importance and compensation associated with development relative to invention, the more competitive we will be. Designing systems to balance the ratio of investment in invention and development is challenging (see the entire history of drug development).
Excellent overview; three quick follow-up questions/comments: (1) in light of involution, how can private Chinese firms appropriate their investments in R&D? (2) in light of consortia sharing, isn't substantial free-riding likely, and if not, are there different legal or cultural constraints at work that prevent free-riding? and (3) what role does competition play, if any, at the R&D stage, and, if present, is it competition among firms, bureaucracies or regions, or all of the above? Certainly the framework for evaluating innovation and incentives (from Schumpeter to Arrow to our recent Nobel prize winners) tends to focus on competition among firms. China's approach suggests either that the economic modeling is incomplete or that their approach will have problems progressing to next-generation products. We shall see.
These are all great questions. I don't know enough about the details of the system to answer any of these yet, but I'll keep trying to find out more. I think because we've spent so little time and effort as a society evaluating different research institutions, we don't have a well-rounded picture of how competition and cooperation really work here. And China's system is very new, and we don't have a lot of details about it from the outside. So it could end up being a giant mess, like many of their industrial policy experiments in the 2010s. Or they could iterate and tweak things and improve incentive structures as they observe results!
I bumped this comment to the main post!
Great article on an extremely important topic. I had no idea that China was doing this. I will also take a look at Barry Naughton’s article and the other linked articles, which look intriguing.
By the way, you and your readers might be interested in a new series of articles that I am writing on how research and development has changed over the centuries. It goes into more detail on many of the themes that you mention here in this article.
Here is the introductory article to the series:
https://frompovertytoprogress.substack.com/p/r-and-d-organizational-profiles-of
I would be interested in an analysis of how this picture changes in a world dominated by AI in research, product development, planning, coordination and commercialization. Does this have a "leveling" or "democratizing" effect by allowing smaller teams to pursue end-to-end innovation? Isn't the promise of AI that it is the ultimate vertical integrator? It will know what ends we should pursue better than humans, it will do basic research and applied research better and faster than humans, it will know human needs and wants better and so will be better (and faster, and cheaper) at commercialization...
I think this is the quadrillion-dollar question, but no one knows the answer.
Came here to look for a comment like this. By reducing/eliminating the barriers between key institutions (universities, labs, companies, etc), they're creating their own intranet. Sort of like the idea that the ultimate point of the internet was to connect all information to enable AI, China might be doing a better job of connecting their institutions so that AI can use it for whatever ends them deem worthwhile. Makes me think that while the US may be ahead (for now) in development of AI, China may actually utilize AI better because there are fewer barriers between information pools - at least within the sectors that their government mandates.
I am not sure yet if research is a "talent" problem or an "organization" problem. There have been frontier labs with the highest concentration of talent that have consistently NOT "won". Xerox Parc, Bell Labs come to mind.
A future lab filled with brilliant AI researchers will still require state capacity to organize.
First, this was a fantastic writeup of something I've been trying to get my head around for a while.
One big gap in my understanding is how competition between regional governments figures into the whole thing. As I understand it, the way China got 200 startup EV companies, now starting to get widdled down to just a few, was from provincial governments sort of acting like venture capital firms funding local champions in a competition to turn them into national champions (and maybe get technology hubs to be located in their territory). At least, that's how I think it works.
It seems like despite the system being largely nationally coordinated with a lot of central planning, there's still a huge element of competition between regions and individual start-ups that helps keep the system more dynamic.
Thanks, I'm glad you liked it! Yep, they're probably relying on regional competition to keep competition in the system. But also A) competition between labs for prestige, B) competition between companies to take the lead on each project, and C) competition between institutions with overlapping mandates.
Who knows how much any of that will work out...
Bumped this comment to the main post!
Fascinating essay re the CAS, I had no idea. I love a good experiment, and it will be interesting to see how this Chinese one pans out!
My first impression is that nationalism as a science motivator definitely works (think Manhattan and Radar Lab projects) for specific tasks. Maybe that is enough to accomplish great things, even if I don't think it will drive 'basic' research very far.
I think nationalism is a motivator, but researchers can't subsist on nationalism alone. They will need intellectual prestige, credit, and individual recognition.
The Nobel Prize is dumb because science is a team effort, but there's a reason we have it -- dreaming of getting a big gold medal that says "you're smarter than the other smart guys" is a huge motivator for people who have spent their lives basing their self-image on being "the smart guy". Japan's policy of having managers steal credit for the inventions of researchers and engineers has probably materially hampered its innovative capacity.
Meanwhile, involution is a huge problem on the downstream side. Companies will participate in this system because they have to, but they may not end up making much money off of it, because every other company is getting the same technologies and getting paid by the government to make the same products. So companies may free-ride and expend less of their own in-house innovative efforts, and in the end China may just end up with CAS labs and universities trying to do all the work on their own to make up for lackadaisical unmotivated corporate labs.
I'm not saying these things *will* happen, but I don't think nationalism will be enough to overcome these problems all on its own!
"Companies will participate in this system because they have to, but they may not end up making much money off of it"
Companies as a whole may not make much money, but you can bet individuals will game the system substantially for their own personal gain. "If I can't get rich with my shares/options I'll just pad my expense account and bonuses, etc."
That will surely hamper the collective effort due to vicious, selfish infighting over who gets what perks and favors. There will of course be periodic purges for "corruption" but the system has built-in incentives to encourage it. If you can't get rich as a team then you can enjoy the facade of being rich as an individual, if you're clever enough to hide it well.
Basic research is consistently extremely valuable, but also produces very little exploitable short-term profit. It seems mainly to thrive in places where some motivation other than "efficiency" is driving the effort. Doesn't really matter if that's nationalism, misplaced pride, or a telco monopoly burning profits to get R&D tax credits. People are curious and if you fund smart ones to do what interests them, (eventually) useful things will happen. The main question is whether the "inefficiency" is politically/financially sustainable.
In the 1980’s ,when we had Japan envy, MITI was seen as a differentiator. How is CAS different? As a capitalist I have learned that governments are not particularly good at picking technologies or companies. Maybe you get some wins but there is a lot of wasted effort and money that could go to the people.
MITI is really just on the production side. China has similar institutions now that are actually even stronger, and they cooperate with CSTC, MOST, and CAS.
This might end up being a case of "too many cooks". And of course there's always the possibility that central planning just hurts here, period. People still debate how much MITI hurt or helped in Japan.
I generally agree with your skepticism. Large government created and/or directed operations tend to collapse or become counterproductive eventually under the weight of bureaucratic inefficiency and sclerosis. OTOH, it does seem like their current innovation system still incorporates a good amount of meaningful competition. It's not just the government picking winners and losers.
I don't fully understand how it all works, but EVs are providing a good example of this competition. I think provincial governments provided seed funding (maybe acting like VCs?) for hundreds of EV startups. Each province was in a competition for national prestige and to get selected by the national government to have a technology hub located in their region.
This created the involution problem, but it also created a competitive process where mostly market forces pick the winners and losers as those hundreds of firms now get whittled down to just a handful.
I think it's possible that people (governments and think tanks) will learn and get better at devising ways to stimulate innovation. Ideas about this in the 1950s were pretty simplistic (build PARC). In the 1980s Japan was trying MITI. China's systems seem much more sophisticated. Even if people don't "figure this out," whatever "this" is, they may get better and better, and the theory that "government can't pick winners and losers" might have been totally correct before, but in the coming decade, it might not be so black-and-white.
really interesting article, with perhaps two lacunae, respectfully. first, such a system is ripe for internal corruption, policing of potential corruption, over the long, would seem to me very likely, secondly, to increase buraurcratic resistance to exactly the type of innovation this system is
supposed to cultivate thus one can foresee a short term period of productive but eventually the system fails, as everyone in it is either trying to grab turf or money, or (perhaps also) looking over their shoulders at internal bureaucratic competitors the well-known Chinese sci-fi trilogy, Cixin Liu's Three Body Problem seems to describe that, although in the context of the cultural revolution if the 60's
Massive room for potential corruption, and a decade ago bribery to get to the front of the line for talent programs / government grants was common. The Chinese government has tackled this seriously, the number of audits / checks and balances are insane. Doesn't mean it's solved, but environment largely works.
Super interesting post, thank you!
Thanks for this deep dive into the China innovation model. Although you do point out that the piecemeal pipeline story is an oversimplification, it may be that the story does not represent how innovation worked in the west for most of the scientific age. I am not sure how you would fit in that framework the story of how Max Planck discovered the basics of quantum theory trying to solve the problem of designing a better light bulb (in competition with the US!) (https://www.ornl.gov/blog/quantum-science-exploring-little-particles-big-promise). Or that one of the main projects of the newly minted Royal Society in London in the 17th century was to solve the Longitude problem (https://mathshistory.st-andrews.ac.uk/HistTopics/Longitude2/), leading to fundamental advances in astronomy. As an oceanographer, I can tell you that the oldest oceanographic institutions such as Woods Hole and Monaco were founded and funded initially in the late 19th century to solve fishery problems, but to do that the scientists had to ask fundamental questions about how oceans work. Finally, there is the Louis Pasteur quote "There is no such thing as a special category of science called applied science; there is science and there are its applications, which are related to one another as the fruit is related to the tree that has borne it”. Maybe China rediscovered an older (pre WW II), more vertically integrated, model of innovation.
That's a good point...I have a pretty clear idea of how America's research institutions work *now*, but a much sketchier idea of how Europe's scientific institutions worked a century ago (to say nothing of *three* centuries ago!).
But I'm pretty sure they had nothing even remotely similar to the planning and coordination capabilities of the modern Chinese party-state. Probably no one else ever has. It's just an open question of A) how much this is reinventing the wheel, and B) how effective any of this will be.
Thanks for your reply. I completely agree that nothing in the past in Europe compares with what the Chinese are doing, and my view is that we certainly don't want to do the same. However, I would maintain that recapturing some of the spirit of pre-war innovation and incorporating that into science institutions would help accelerate impactful innovation here. It is already happening. That is what (I think) Focused Research Organizations are trying to do. I have heard that they don't want to deal with universities because the focus of universities is too far removed from what they see as needed to efficiently solve big technological problems for society. In Nova Scotia where I live, the FRO model has been adapted to deal with social problems. People are trying to change the pipeline.
I hang around with Tom Kalil enough to have been fully FRO-pilled!
This was a good overview and primer to the ecosystem. Noah does a nice job of capturing most, if not all, pieces of the innovation "pipeline" in China.
I think it does risk missing the relative "weight" of each piece, however. For example, this top down planned economy vision of innovation only applies to certain critical industries where a) strategic risks for the country exists (so China needs them), and yet B) the risks / capital required are way too large for private capital to take the lead: the "big airplane" project of C919, nuclear power, SEM replacement to ASML, etc.
But much of the other innovation that Noah / the rest of the world is enthralled by comes from a different application of the innovation pipeline. I'm talking about AI models, robotics, EVs, etc. It's still the same pipeline, often starting with a Five Year Plan calling attention to the opportunity, with core R&D advancing in national labs... but the most critical piece is the competitive nature of what Lin Keyu called the "mayor economy". Mayors are measured by commercial output, and they essentially act as VCs - they are the most powerful muscle in the entire innovation pipeline. They put together the focused capital that transforms patents / knowledge into revenue, products, and jobs. They cut red tape, they find talent, they do whatever it takes. When you speak to one of these mayors as an inventor / entrepreneur, the question you'll most frequently hear from them is: "how can I service you to help you succeed?"
IMO, this aspect of the innovation pipeline is what most distinguishes China from other nations, and is most responsible for the recent success in the industries most of you care about.
It's not often that I read a post about a topic I had never thought about before. Super useful.
Great article. I do think that there is a fundamental difference in the world view of leaders in China and here in the US (the broligarchy). Xi Jinping and much of the CCP leadership are intent on reversing the "100 years of humilation." In other words, they are nationalists. Not unlike Churchill's and Roosevelt's leadership in WWII.
This is in contrast to US leaders like Thiel , Andressesn and Musk who envision a world shaped like Neo-stalinist Ayn Rand's fevered dreams in Atlas Shrugged. They have their dual citizenships, and if their national program goes south, its, "Oh well..." (cf. Musk and Doge and Thiel's and Andressen's puerile essays).
Hopefully, Noah will write an essay on China's use of open-source AI versus the US approach. Scott Galloway had an interesting take on that in terms of a marketing strategy and as a possible danger at the heart of the US economy ( 80% of value for 10% of cost leads to market capture and collapse for the Open AI and other US companies that constitute the bulk of the US stock averages)
It may be worth noting that the US has long had an abundance of national labs well beyond DARPA. Surely names like Lawrence Livermore or Jet Propulsion Lab ring a bell. There are many more. There are also Federally Funded Research and Development Centers. Other great names include Lincoln Labs, Fermi Lab, Argonne and more.
For sure, but it's also noticeable how very few of the IP created in these national labs are directly meaningful in the companies that dominate the market / economy today. (Yes, YES, we wouldn't have Uber without DARPA-funded GPS and Internet... but that kind of indirect "tree" effect isn't the direct, high efficiency innovation pipeline Noah is talking about today.)
extreme ultraviolet lithograph, the technology used to make the most advanced semiconductor devices, was largely developed through the national labs (https://newscenter.lbl.gov/2022/06/03/extreme-uv-light-tiny-microchips/)
Excellent post. Where are the footnotes? I can't find them.
At the bottom!
Not on my screen. Is it just me?
This is very interesting. It reminds me of what the US did during the manhattan project and the space race.
China (and most of east Asia) have a huge competitive advantage over the US to go in this direction. They culturally already have a collective mindset. The only way the US will do something like this is if they believe there is an existential threat to our country. Otherwise, everyone is out for themselves.