TiB 125: The tragedy of lost genius; why scientists shouldn't be creative; interests and ideology in US-China policy; and more...

Welcome new readers! Forwarded this email? Subscribe here. Enjoy this email? Forward it to a friend.

I’ve said before that I see macro-talent allocation as one of the world’s most important challenges - and one of my primary motivations for co-founding Entrepreneur First (I talked about this at greater length on Invest Like the Best back in January). So I was excited to come across this amazing paper by Ruchir Agarwal and Patrick Gaulé. It asks how important talent is for knowledge production and how much the world misses out on when exceptional people don’t have the opportunity to fulfil their potential.

The answer to both questions, at least within mathematics, appears to be “a lot”. The authors look at participants in the International Mathematical Olympiad (IMO), in which exceptionally gifted high schools students compete to solve maths problems. They find that differences in IMO score predict large differences in future research output and future exceptional achievement, even though they’re looking at a group where everyone is very strong to begin with.

They also find that, conditional on the same IMO score, being from a poor country predicts lower research output, presumably because of less access to opportunity and infrastructure. It’s significant: the total long-term research output of the group would be 17% higher if everyone had output at the same level as participants from rich countries. Perhaps mathematics is unusual, but these findings point to the enormous value of removing geography as a barrier to ambition - and seeking out and supporting the “lost geniuses” on whom the world is missing out. 

A recurring TiB theme is how to make science better. We’ve talked before about the replication crisis and the fact that most published research findings are false. Why does this happen? In a fascinating thread (or here, as a single page), David Chapman suggests it’s because the stories we tell about science are wrong. There’s a popular idea of the scientist as a “creative contrarian” who’s right when others are wrong. Chapman says we have this backwards: good science is rarely surprising and, done well, it’s careful rather than creative - but the incentives pull in the opposite direction.

There are lots of provocative ideas in the thread, including thoughts on why it may matter that the dominant figures in Silicon Valley no longer have PhDs, why scenius matters, and why Silicon Valley may be poorly suited to innovate on new types of research institution. It’s worth reading this in conjunction with my friend Ben Reinhardt’s new post on the classic Bloom, Jones, Van Reenen and Webb paper, “Are ideas getting harder to find?”, which we’ve cited several times in TiB (while you’re there, you should also check out Ben’s tour de force on Why Does Darpa Work?) 

Ben notes that the problem we face is that “exponential progress [which is what we’re used to] is going to require exponential inputs and specifically exponential numbers of researchers”. What we need, then, is not institutional tweaks, but to “break the researcher-idea production coupling”. As I’ve said before, I’m optimistic about the potential of machine learning to give scientists superpowers, but there’s still a long way to go. 

There’s a great line in the John Garnaut speech on China that I linked to last week that sums up the challenge that the West faces in engaging in China: 

Whether to think of China as an economic partner or a security threat is the theme of a long and rich review essay by Adam Tooze in this week’s LRB, entitled “Whose century?”.

We talked last week about the neglected importance of CCP ideology in understanding China, but Tooze makes the same point in the other direction: it was the economic ideology of the Washington Consensus that drove the US’s relationship with China in the 1990s and 2000s. The belief that all that mattered was integrating China into the global economy and that political “normalisation” would follow now looks horribly wrong.

Why did the US adopt the policies it did in the 90s? Tooze says the answer lies in US domestic politics. He argues these policies reflected the interests of American corporate elites, but not ordinary Americans. Trump may have betrayed his core constituency, but his rhetoric correctly diagnosed this divergence in interests between elites and the median voter. We’re all China hawks now, but though sensible strategies exist, if Tooze is right, a coherent US policy towards China will be hard without a new domestic political settlement. 

Thanks for reading Thoughts in Between. If you enjoyed this, I’d love you to help me grow the readership. Just forward it to a friend or - even more effective - share it on Twitter or LinkedIn.

Lots of newsletters get stuck in Gmail’s Promotions tab. If you find it in there, please help train the algorithm by dragging it to Primary.

I’m always happy to hear from you. Just hit reply or message me on Twitter.

Until next week,

Matt Clifford