I don't get the "disappointment" with USA scientists in this thread not having yet published a replication. There has only been one weekday since the LK-99 revelation came to light. Have some patience.
The analysis from this article is a publication from Berkeley National Lab that has already come out, and Argonne National Lab has announced they have synthesized the material as of yesterday and will release results from their replication attempt in the next few days.
One thing that has been constant for at least a decade or more is that every time a lab in China publishes even an incremental advance, a legion of internet commenters descends to declare the end of US hegemony because Americans didn't discover it first.
Japanese and South Korean increasing technical prowess didn't, why would Chinese be any different?
China has serious fundamental problems and if you account for its gigantic population - it's seriously underachieving not only compared to US, but even compared to the rest of the developed world.
It's fashionable to talk about China taking over, but it's far from guaranteed. China is pretty much confirmed to be falsifying its economic stats for example [1] [2] but people just take them at face value anyway. I don't understand why.
> It's fashionable to talk about China taking over
It's not so much fashionable as it is literally state propaganda used to try and shoulder its way into the South China Sea and the Pacific by claiming it is so prosperous and populous that it is entitled to increasingly large sphere of influence and direct control.
> it's seriously underachieving not only compared to US, but even compared to the rest of the developed world.
This statement would have been viewed as absolutely obvious and ridiculous 2 decades ago. The fact that it even needs to be said now is indicating how fast they are advancing.
They were underachieving for 2 centuries, ceasing to kill their own citizens by millions and imprisoning anybody who tried to think for themselves is enough to grow if you did so for a long time. But it does not make you a new hegemon.
I highly doubt they can preserve their pace of growth for next few decades without significant changes to the regime and liberalization.
Once they return to the mean - they will very likely slow down. Arguably they already did (3% official growth last year + people accusing them of falsifying data for 1-2 percentage points of growth each year would make them already grow slower than some western countries, including the US).
> I highly doubt they can preserve their pace of growth for next few decades without significant changes to the regime and liberalization.
"Demographics is destiny" also comes to mind.
Note how the era of "Japan Inc." during the 1980s was also the time when the post-war baby boom population were in their 30s-40s (i.e. peak productive worker population). As this cohort has aged and are now in retirement, Japan's economic performance has tailed off.
From the geopolitical standpoint, this may also explain why China and Russia have become more belligerent - they will lack sufficient numbers of young fighting aged men in the next few decades.
The difference with the Japan case is that China is going to have a larger base of young, very well educated people than the US for at least the next century unless something changes dramatically with either birth rates or immigration.
China is facing severe demographic issues for the next 20-30 years due to their terribly short sighted one-child policy and their poor immigration rates. Literally the opposite of what you're saying. Where on earth did you get your facts from?
With the population edge they have, they will still have more young people than us. More of their output will have to go for caring for the elderly, but that will scale with technology.
Why does having a large population matter? If you look at the top 10 countries in terms of population (China, India, US, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Mexico), most of those places are not exactly heavyweights on the international scene.
There are so many factors that contribute more to economic and political success than sheer population. I would even say that a large population is a bad thing in many cases. India would be better off with fewer people. Their infrastructure can't handle their density, they don't have enough jobs for their educated workers, there's a lot of sectarianism conflict between various religions and casts, etc.
High population + low GDP per capita is probably the worst situation a country can be in IMO. So it's not enough for China to have a billion people—those people have to be doing something productive for the economy.
Not Zeihan tier demographic analysis which is a start.
PRC currently generating ~5m STEM per year, aka OECD combined, multiple times more than US has ability to train or brain drain. And relatively proven ability to coordinate talent. Project that out next 20-30s from previous 20-30 years of birth rate and PRC on trend to add 50M-100M STEM to workforce, just STEM, not including other skilled workers, which is the greatest high skill demographic divident in recorded history. Literally no country in the world, at any time in the past or projected future has better demographics for actual global competition than PRC in the next 30 years. Including India who will have more people, but have all the issues you noted that will likely prevent them from actually coordinating human capita enmass successfully in the time it takes their youth demographic divident to expire.
Contrary to naive PRC demographic pyramid bomb arguments that doesn't address what you correctly note below - quality of human capita. By 2050, PRC is going to add more STEM talent than US is projected to increase population. All the news of PRC climbing up value chains, leading in science and innovation indexes from last few years? Or moving from 1T to 18T economy. Done by growing STEM from ~2M to ~17M STEM exploited via industrial policy that west is now copying. PRC is moving from workforce with 25% skilled talent to 60/70/80 of modern economies with 60% and increasing tertiary education rate that biases towards science will look weak on per capita stats due to huge existing cohort of old / undereducated, but it absolute terms it's a demographic advantage no country currently has conditions to remotely rival. Then note how net population decline will reduce resource dependency and you basically have the most optimal mix of demographic trend for PRC with respect to geopolitical competition.
Are those demographic issues difficult to govern/manage? Yes, but they're also close to ideal conditions for improving comprehensive national power within PRC's constraints. PRC's big population = with big demographic curse but also big demographic divident post 2050s. But they don't cancel each other out. Likelt follow JP/SKR trend where TFR collapsed in the 80s/90s but GDP increased 500% because workforce was net gaining skilled/productive people, while losing unskilled. JP (and later SKR) are/will only entering process of real stagnation when TFR cannot replace level of skilled labour at parity. Which for PRC is a post 2050 problem and even then countries will be competing with a PRC roughly 3-5x larger than now.
On the actual demographic curse of aging, the blessings of huge segment of PRC old getting old before they get rich is there simply isn't going to be high expectations for advanced economy levels of welfare and social support. There's a reason PRC has 90% (96% in rural where poor concentrates) home ownership and very high household savings rate. Old expects to weather most of retirement without substantial state support and increasingly family support since they don't want to burden future gens. If you look at JP, old are basically rotting/dying alone, unceremoniously. In JP it's called Kodokushi, in SKR, it's godoksa, it's happening in HK as well. In households that pressure young for support, you know what the east-asian human response to that is? Being miserable whiel working even harder. Half the reason JP/SKR/TW lead in high end industries they currently dominate is because those societies have resigned to working 100% harder for 10% competitve advantage.
Thank you for writing it out, I think people are sort of in denial and have been for a solid decade now.
Arguments about China’s inevitable collapse are more about comfort than about the an reality, because as you have mentioned if you think at all seriously about the demographics argument it just does not make sense.
I think the biggest risk for China demographically is if their aging population translates to an extremely conservative, risk-averse government/society that harms their technological progress, as has happened with Japan (ie. top multinationals stuck in the 80s with fax machines, paper, etc.)
They do not need that many years of large growth to overtake the US in terms of size. Technologically they are pivoting to AI and the service economy much better than Japan, southeast Asia, etc.
No, it's evidence of how fast they advanced under Deng. It is unclear that the advancement is continuing / has continued into Xi's reign, especially since the beginning of his third term.
That just means China has it harder and is more likely to fail. Access to global markets is a critical factor in growth of Japan, South Korea and China. USA controls that.
>Internal markets in China are becoming increasingly robust
I think this fact is being severely understated, perhaps even denied by almost everyone in the west.
The way the US-China cold war has been playing out, the US and friends keeps closing doors only for China to go "Go right ahead, I don't have to play ball." and just succeed even harder on their own. Adding insult to injury is they then take that success and just wholesale buy the doors the US keeps closing.
A surefire way to lose Pax Americana is to become delusional that Pax Americana is winning when by all accounts it's losing and losing hard. I think this ship can still have its course corrected, because Pax Americana is better for us than Pax China, but not if we just keep up this kabuki theatre.
Japan and South Korea are allies of the US. They are also much too small in population and resources to reach a level of economic power that would allow them to do anything on their own.
The Chinese Communist party has the resources, people and determination to follow a different course. It might underperform relative to its size, and the one-child policy is going to cause a disastrous demographic situation, but even if you consider just coastal, urban China, that's getting pretty close to US in terms of people and economic power. The communist party is developing its military capabilities at a fast clip, and fusing its military and civilian economies to accelerate its technical development. Where almost every other government has been unable to wrestle control of Internet information, the Chinese Communist party has successfully turned its version of the Internet into an effective tool of govt propaganda and social control.
In other words, there are lots and lots of very concrete reasons to believe that the post-development trajectory of China can be very different from Japan's or South Korea's.
Geopolitical competition is driven by absolute advantage not relative. At PRC scale, activing fraction of population to high skill is enough to compete with US+co, and PRC's fraction of that high skill talent is set to increase multiple times over coming decades. There's a reason why PRC moving from ~2M STEM to ~20M STEM grew economy from 1T to 19T and is at close to parity / at parity / even leading in full spectrum of industrial sectors whereas "small" countries like JP and SKR has to pick and choose where to compete. With PRC population they can do everything, and in fact can't not considering they're graduating ~5M STEM per year and will add more STEM (50-100M) than US will add population in next 30 years. PRC is not JP and SKR, and TW, it's on trend to be 4-6x JP+SKR+TW without the geopolitical constraints that prevent US partners from full competition against US in critical strategic industries.
>comfirmed falsifying
Here's a more rigours NBER light study on PRC GDP being underestimated by much more accomplished economists aka not retarded like Martinez study.
Here's a more comprehensive GDP resconstruction by Rhodium, leading economic search group that focus on PRC that also reports underestimation, ~10% by 2014. Including work by David Dollar, one of the more competent PRC economic analyst at Brookings that also led discussion of your second link, noting it had 2 models of over estimation and went with the much higher overestimation model vs the lower (basically marginal amount) one that was probably more correct.
No one takes PRC GDP reporting at face value, but anyone with half a brain can see PRC climbing western science and innovation indexes (controlled for quality), moving up global value chain, displacing other advanced economies in increasing amount of intermediate and final goods and doing all that rapidly in last 10 years that US basically needs to go full spectrum containment beyond what they did for USSR while acknowledge PRC is greater challenge then USSR ever was. Anyone with half a brain can also napkin math future skilled workforce between US @35M STEM + ~600K per year, vs PRC @20M STEM ~5M (M as in million) per year and understand PRC will have more STEM workforce than US in a few years, and possibly 2-3x more STEM by 2050 even factoring in PRC demographics. Cut that in half and PRC's future potential is still outrageous.
1.2 billion _old_ people. While quantity has a quality all its own, in this case the demographics are not what you want to see if you are predicting long-term Chinese growth just based on population.
Depend on them for what, the cost of a bowl of rice? What is it exactly about 80 year olds that will collapse a country with a brutal government like China has?
You think that a country that has a strong cultural history of venerating their elders is just going to pivot on a dime and decide to put grandma and grandpa onto an iceflow and cut them loose? Those elderly out in the hinterlands are also the people who raised a large fraction of the youth population while their parents were off in Shenzhen pulling shifts at Foxconn.
The other thing to look at in those population distributions, besides the upside-down age distribution, is the absolutely pathological gender imbalance. The number of surplus males under 30 should be keeping Xi Jinping up at night because if there is anything that is going to trigger another revolution it is tens of millions of disaffected young males migrating across the country who eventually get fed up with what they perceive as a dead-end future.
It's a mistake to think in terms of the US only. Because unlike China the US doesn't think that way. And if you include the core part of Team America, it's even steven.
Who is the core part of team America? Because the anglosphere doesn’t even it out I don’t think and Western Europe is not unequivocally on our side in US-China industry competition.
The whole edifice the US, Britain, Japan and Germany created during the cold war to protect societies from the Soviet Union and the Chinese Communists.
The core part defined by long standing military and economic ties is about 1.5 billion people.
The only argument I'm making is that it's insane to draw sweeping geopolitical conclusions from the fact that Chinese scientists posted their attempts at reproducing a Korean lab's result on Twitter before the Americans did.
I agree. Not to discount the work the Chinese academies are doing, but I don't regard twitter activity as indicative of what U.S. labs are doing, and I'd strongly urge others not to do so either.
As you noted, there's labs working on this, they've said they're working on this, and I take the lack of minute-by-minute social media updates to indicate that it's serious and they want to get it right.
China has more people, more money to spend on research, more equipment, more access to raw materials and chemicals, more manufacturing base, more STEM graduates, more everything, and all of that by huge margins. USA scientists will eventually put something out, but 10x more slowly than China.
I'm not looking at official numbers here, but in my field (academic biotech research) I still don't see much innovation coming out of China.
All of what you say is true, yet in academic research China is still far behind the USA in nearly every field. They haven't yet been able to build institutions with the staff, structure, and culture needed. That will likely change in time, but at present the best Chinese scientists are still coming to the USA to work, and staying here. Despite the USA having a lot more research output, a substantial fraction of the top scientists in the USA are from China.
My Chinese colleagues tell me that the bureaucracy and authoritarianism in Chinese institutions puts a lot of hurdles in place when trying to do research. Simply buying equipment, hiring staff, etc. is a nightmare, and results in "evaporative cooling" where the top researchers with the option to work anywhere don't tolerate this and leave.
Edit: I will also add that in some sense US academia is supported by overcharging Chinese students for tuition. If and when Chinese institutions get to the point where the best students want to stay there, there will be a huge crisis in US universities, possibly being unable to support tenured staff.
US academia isn't really funded by the foreign grad student tuitions. It's not even really funded by tuition, as tuition+fees is <30% of total US academia funding. Of course every organization wants to have more money and so they work on those revenue streams, but if all the money from the 5% students which are foreign suddenly disappeared (even ignoring the foreign students whose tuition gets funded by being TA/RA), that wouldn't be a crisis, it would just reduce the profits of the universities and have some administrators be extremely sad about less bonuses.
Varies by field but last few years PRC been reaching parity to topping various science and innovation indexes, PRC institutions also breaking global top100 and general trend of moving up rankings. TLDR is PRC science exploded after mid 2010s - lag effect from academic reform in 00s. Biotech/bioeconomy a outlier though - just got elevated to strategic sector with 500b usd investment in last year's 5 year plan, so expect PRC to start being competitive in 5-10 years.
> at present the best Chinese scientists are still coming to the USA to work
IMO not true anymore. This isn't pre 00s where PRC send best abroad as part of state strategy and best have some english fluency because it's needed for science. More and more PRC best aren't English fluent since there's sufficiently large and growing chinese science ecosystem and best also have good access to resources in tier1 labs. Hence US cracking down on 1000 talents program where PRC entice scientists to work in PRC due to unparallel resources. Or acadmeic exchanges with PRC in general. Also see stats of record amount of scientists returning to PRC this yearm
>best students want to stay there,
Best PRC students go to PRC C9 (Ivy equivalent) now. Top US institutions captures the occasional talent with english proficiency and desire / resources to go abroad. But PRC best have been largely staying in PRC last decade. TBH most Chinese students in west now are those who couldnt hack PRC gaokao/teritary selection system but have rich enough parents to send them abroad. They're B tier talent. Still adequate to work in western labs but you simply don't hear much about PRC students in west who tested top of their districts anymore like in 00s. They're no sending their best and haven't been for while. But generally their ok is good enough. Medium term expect even less PRC students, partly due to geopoltics but also PRC likely phasing out mandetory English as core subject which will make brain draining next gen more difficult. There's always India who seem systemically incapable of preventing brain drain.
I’ll take what you are saying at face value, but point out that academic career pipelines are slow- it’s often about 15 years to go from college freshman to starting out as a “young” PI in the USA. If what you are saying is true, it will still be another 5+ years before we start seeing top students that stayed in China publishing as PIs. Edit: Which lines up with what you said at the top of your post.
That's a reasonable timeline. PRCs have been developing pharma/bio talent since 2015s. It’s part of medicine&devices category for MIC2025. Focus was on devices, pharma/bio got extra resource/political attention in the last few years due to covid. Like most PRC ventures it's going to be rocky/messy until it's not. Broad point is bio is one somewhat neglected up until covid sector that PRC wants to become a global player with industrial policies that has worked well to very well in other MIC2025 categories..
China spends nowhere near as much as the US on basic research. The US spends $100 billion on basic research annually, compared to $25 billion in China.
However, total R&D is quite a bit closer, with the US spending $660b to China’s $556 billion. [1]
There's PPP considertaions. PRC R&D funding also bias towards experimental/applied vs basic research. Hence not surprising they're hammering these replication efforts.
This may be a hard pill to swallow, but science isn't a "more is better" game.
It ultimately requires being embedded in a culture that, to quote Popper, "seeks truths that are difficult and interesting". From that view, the problem for US scientific efforts is entirely home made, but it's also a problem that's much more pronounced in China.
All the smart people in the US work in tech and finance.
If you have a big ole' brain, why would you take $80k to work in some crumby lab when you can get paid $350k to maintain a login screen from the comfort of your mountain side home.
As an academic mathematician-turned-software engineer, I can assure you my reasons for leaving academia were not a “monomaniacal focus on financial gain”. The money and job security didn’t hurt, but the real turning point for me was solving the two-body problem.
Choosing the $350K job isn't necessarily monomaniacally focussed on money, there's plenty of combinations of priorities that would favor that.
But when you don't assume all smart developers are monomaniacally focussed on money, there’s also plenty of combinations of priorities that might favor an $80K lab job over the $350K login screen job.
When you work in a lab you have more ability to work on what you want to. Well at least once you get to the top of the lab - many spend a lifetime working on the interesting problems someone else is interested in but never reach the top where they are in charge. Meanwhile many people who work in tech are able to work on problems they find interesting enough (maybe not the most interesting, but still interesting) and they get paid well for it. Still if you want to work on some problems a lab is the best change to get there.
Sure if you have a huge drive to work on science directly.
But the fact of that matter is that many people would rather work on something they find interesting enough, while also being able to comfortably afford a good life for themselves and their family.
It's a saying referring to having pride in the nation you live in. Make your country's status in the world just a tiny bit higher. Win a Nobel Prize for your country. Win the Olympics for your country. The US is not big on collectivism and that's okay. Other countries do things differently.
Maybe Venture Capital should spend less money on Juciero, automated pizza ovens, and gig-economy bs and more money on materials science and fusion research.
Even if you end up lighting the money on fire, at least it's a more societally productive fire than, say, SoftBank's portfolio.
There isn't a good likelihood on strong return on those type of investments. You need real engineering talent and leaders. You can glue together a delivery or dating app with a bunch of 25 year olds who just got out of a ruby on rails bootcamp, raise a few million with charismatic/well connected leaders/founders, and maybe gain enough users to be acquired or something. Anything in materials science is going to require some deep expertise, labs/machines/composites/fabrication/manufacturing setups, inside connections at the DoD, trial and error that costs millions.
Yeah, I agree, its way cooler, but way more risky for VC.
Timelines matter. The BS will either return the millions in a couple years or it won't. Basic research may return much more, but your best case is still many years before you gets results. The risk is actually higher with basic research than for BS as well - the pet rock earned money and there are plenty of other examples of stupid things working well quickly. Basic research in fusion hasn't returned anything yet even though the physics has said it works (though with recent reports maybe fusion is just around thee corner - or maybe it is still 50 years out)
See, the thing is, you're making an argument about how VC works, and I'm making an argument about allocation of money towards VC vs. other socially beneficial projects.
I understand how VC works. I'm annoyed that - let's even take the Saudis as an example, since they fund half of Silicon Valley anyway - would rather flush their money down the drain on stupid shit than do long-term projects that may help secure their existence post-oil. Even with every motivation to focus on the long-term, still it gets blown on short-term complete garbage.
Saudis don't fund half of Silicon Valley. Aramco is worth 2 trillion, apple alone is worth 3 trillion.
Saudis have built entire top-equipment universities out of nothing (KAUST), and hired top foreign professors, and offered generous scholarships to foreign students to fill it up.
You rant about short termism, but don't seem to like doing research yourself.
The analysis from this article is a publication from Berkeley National Lab that has already come out, and Argonne National Lab has announced they have synthesized the material as of yesterday and will release results from their replication attempt in the next few days.