#1. Agglomeration Effects and Emergent Clusters
Why do clusters of geniuses form in certain areas and not in others?
Why is it that clusters of highly skilled people form in certain areas and not in others? Today, for technology, it’s Silicon Valley. In the early 20th century, it was Hungary for mathematics. 19th Century Impressionism was founded in France. Florence was the hub of art during the Renaissance, and the most famous philosophers all lived in Ancient Greece. Why these areas and not others? Why did they come about and collapse when they did? And why did they specialise in these subjects rather than others?
This is the introductory article in a series that aims to understand the conditions that allowed genius to flourish and clusters to form, with the hope that these conditions can be replicated in the future.1
It is inspired by the work of both James W. Phillips, who found common themes in some of the best research labs history has ever seen, and Patrick Collison and Tyler Cowen’s 2019 article, ‘We Need a New Science of Progress’. Both of which I would recommend reading if you haven’t already.
In recent years, many articles have made the point that genius has been collectivised and that the days of the solitary thinker, the lone creative genius, the great man or woman, have passed.23
Perhaps this is true.4 However, it’s worthwhile pointing out that the solitary researcher rarely exists nowadays.5 Instead, research teams—for whatever reason—are growing in personnel. Perhaps this is because teams are more impactful in terms of citations than individuals.6 Perhaps this is because ideas are getting harder to find and so require more people to work together.7 Or, perhaps not.
What we do know is that whilst “large teams develop science and technology, small teams disrupt it”. So, could this be the reason behind the many, “science is stagnating” and, “transformative scientific breakthroughs are declining” claims?8 That is, teams are less disruptive than they used to be because they’re bigger in size. Where are today’s small, disruptive clusters?
And, if this is the case, in keeping with Collison and Cowen’s article, shouldn’t we be looking at the past to help us understand how these clusters came about so that we can develop new clusters in the future?
It’s undoubtedly true that intelligence is largely genetic and clusters are reliant on this genetic genius, however, we also need to understand the cultures and social conditions that enable successful clusters to come about.
Whilst Max Perutz praised the creativity and serendipity within the Laboratory of Molecular Biology (LMB), the opposite was the case for John von Neumann (and perhaps other ‘Martians’), in early 20th Century Hungary. How can it be that the creativity of the LMB incentivised the best molecular biologists to move there, but the external social pressure in Hungary produced the best mathematicians?
Is Hungary unique in its ability to create clusters through social pressure? Was von Neumann wrong? Do other clusters like this exist? And, can we create a unifying theory to explain both types of cluster? These are all questions this series hopes to explore.
Max Perutz: “Creativity in science, as in the arts, cannot be organised. It arises spontaneously from individual talent. Well-run laboratories can foster it, but hierarchical organisation, inflexible, bureaucratic rules, and mounds of futile paperwork can kill it. Discoveries cannot be planned; they pop up, like Puck, in unexpected corners.”
John von Neumann: "[Hungary’s exceptional mathematical ability] was a coincidence of cultural factors: an external pressure on the whole society of this part of Central Europe, a feeling of extreme insecurity in the individuals, and the necessity to produce the unusual, or else face extinction.”
Why is this important to understand?
When we refer to individuals with unrealised intellectual potential, we might colloquially refer to them as 'lost Einsteins'. And, whilst it’s true that we are losing these Einsteins through unequal opportunities, the phrase neglects a much bigger problem: we are also losing the agglomeration effects between Einsteins as well.
We are losing Einsteins. But, we are also losing Renaissances, Manhattan Projects, Bloomsbury Sets, and Bell Labs by not trying to understand or replicate the conditions that allow these clusters to flourish.
Science is an endless frontier.9 So on top of this, we are also losing the offshoots these agglomerations spurt and the new fields they discover. Without the LMB, there would be no vaccines. Without the Manhattan Project, World War Two might have been even more deadly. And without the Renaissance, we wouldn’t have the Teenage Mutant Ninja Turtles!
Paul Graham sums this up when he makes the case for understanding clusters from a social perspective in his article Taste for Makers:
“Something was happening in Florence in the fifteenth century. And it can't have been heredity, because it isn't happening now. You have to assume that whatever inborn ability Leonardo and Michelangelo had, there were people born in Milan with just as much. What happened to the Milanese Leonardo?
There are roughly a thousand times as many people alive in the US right now as lived in Florence during the fifteenth century. A thousand Leonardos and a thousand Michelangelos walk among us. If DNA ruled, we should be greeted daily by artistic marvels. We aren't, and the reason is that to make Leonardo you need more than his innate ability. You also need Florence in 1450.
Nothing is more powerful than a community of talented people working on related problems. Genes count for little by comparison: being a genetic Leonardo was not enough to compensate for having been born near Milan instead of Florence. Today we move around more, but great work still comes disproportionately from a few hotspots: the Bauhaus, the Manhattan Project, the New Yorker, Lockheed's Skunk Works, Xerox Parc.”
Questions this series hopes to investigate:
How do clusters scale? Why do they die? And do they have a carrying capacity?
Does what Sydney Brenner calls, ‘the slavery of graduate students’, reduce the likelihood of clusters, as young, ambitious scientists are precluded from pioneering research?10
“Today the Americans have developed a new culture in science based on the slavery of graduate students. Now graduate students of American institutions are afraid. He just performs. He’s got to perform. The post-doc is an indentured labourer. We now have labs that don’t work in the same way as the early labs where people were independent, where they could have their own ideas and could pursue them.” Sydney Brenner, Dzeng interview in Kings Review 2014
Do clusters favour academia rather than the private sector because universities increasingly focus on basic science and the private sector on applied research?
Can we understand clusters by understanding agglomeration?
Can we understand clusters by understanding emergence?
Just as Geoffrey West found similarities between the scaling laws of cities and animals, can we learn anything from cities, venture capital, and start-ups and apply it to clusters?
Update: I have since written about this here.
What is a better way to create clusters: spontaneous order or mission-focused planning?
Does academia incentivise large research teams rather than small disruptive clusters through perverse incentives?11
If research teams are increasing in size, will we see a resurgence of small, disruptive clusters with the development of Large Language Models? And, will Large Language Models open up new frontiers, or help researchers expand or deepen their burden of knowledge?
Can science learn anything from looking at clusters from other disciplines (such as the Renaissance)?
Here is Max Perutz talking about this:
“Every now and then I receive visits from earnest men and women armed with questionnaires and tape recorders who want to find out what made the Laboratory of Molecular Biology in Cambridge (where I work) so remarkably creative. They come from the social sciences and seek their Holy Grail in interdisciplinary organisation. I feel tempted to draw their attention to 15th-century Florence with a population of less than 50,000, from which emerged Leonardo, Michelangelo, Raphael, Ghiberti, Brunelleschi, Alberti, and other great artists.
Had my questioners investigated whether the rulers of Florence had created an interdisciplinary organisation of painters, sculptors, architects, and poets to bring to life this flowering of great art? Or […] how the 19th-century municipality of Paris had planned Impressionism, so as to produce Renoir, Cézanne, Degas, Monet, Manet, Toulouse-Lautrec, and Seurat?”
To what extent has research been outsourced to external organisations and researchers, rather than labs and companies conducting their own basic/applied research? If this is the case, what is the impact?
“Large corporate labs […] are unlikely to regain the importance they once enjoyed. Research in corporations is difficult to manage profitably. Research projects have long horizons and few intermediate milestones that are meaningful to non-experts. As a result, research inside companies can only survive if insulated from the short-term performance requirements of business divisions. However, insulating research from business also has perils. Managers, haunted by the spectre of Xerox PARC and DuPont’s “Purity Hall”, fear creating research organizations disconnected from the main business of the company. Walking this tightrope has been extremely difficult. Greater product market competition, shorter technology life cycles, and more demanding investors have added to this challenge. Companies have increasingly concluded that they can do better by sourcing knowledge from outside, rather than betting on making game-changing discoveries in-house.” — Eric Gilliam
Further Reading:
Innovation (mostly) Gets Harder by Matt Clancy
Are Ideas Getting Harder to Find Because of the Burden of Knowledge? by Matt Clancy
Lovelace Vision Document: ARIA's proposed twin by James W. Phillips
Why We Stopped Making Einsteins by Erik Hoel
Where’s Today’s Beethoven by Holden Karnofsky
The Burden of Knowledge and the “Death of the Renaissance Man”: Is Innovation Getting Harder? by Benjamin Jones
How academia and publishing are destroying scientific innovation: a conversation with Sydney Brenner by Elizabeth Dzeng
Midjourney prompt: ‘Science as an endless frontier’
This is something Graham Allison and Niall Ferguson have called “applied history.”
“Heroic lone-wolf entrepreneurs may be the preferred heroes of narratives spun by the media, but history has shown us that teams—and the networks that come from them—are the true engines behind innovation in Silicon Valley and far beyond. No one understood this better than Bob Taylor.” Obituary of ARPA/PARC pioneer Bob Taylor
"Every man works better when he has companions working in the same line, and yielding to the stimulus of suggestion, comparison, emulation. Great things have of course been done by solitary workers; but they have usually been done with double the pains they would have cost if they had been produced in more genial circumstances." - Henry James, Italian Hours
Rohit Krishnan makes the point that the definition of genius has changed over time: from philosophers in the pre-modern era, to military geniuses, scientists, and now, business leaders. Could it be that we just don’t recognise geniuses as geniuses?
I think there is some weight to this argument. Taste is subjective and sometimes genius is only appreciated in a specific time and place. Shakespeare, for example, probably wouldn’t be appreciated as much today as he was in the 16th/17th Century because theatre isn’t seen as high-status. Likewise, Galileo’s acceptance of heliocentrism was deemed heretical by the Church at the time but isn’t today, and Katalin Karikó was demoted by her University before she won the 2023 Nobel Prize for her work on mRNA vaccines.
NB: this is an exaggeration.
Bhattacharya and Packalen (2020) suggest the incentives created by citation in academia have increasingly led scientists to focus on incremental science, rather than potential (risky) breakthroughs.
“New problems require new knowledge to solve, but using new knowledge requires understanding (at least some) of the earlier, more basic knowledge. Over time, the total amount of knowledge needed to solve problems keeps rising. Since knowledge can only be used when it’s inside someone’s head, we end up needing more researchers.” See: Are Ideas Getting Harder to Find Because of the Burden of Knowledge.
“Over the past century, we’ve vastly increased the time and money invested in science, but in scientists’ own judgement, we’re producing the most important breakthroughs at a near-constant rate. On a per-dollar or per-person basis, this suggests that science is becoming far less efficient.” - Patrick Collison and Michael Nielsen
As is culture. See: Science— The Ending and Endless Frontier(s)
“I strongly believe that the only way to encourage innovation is to give it to the young. The young have a great advantage in that they are ignorant. Because I think ignorance in science is very important. If you’re like me and you know too much you can’t try new things. I always work in fields of which I’m totally ignorant……Today [we] have developed a new culture in science based on the slavery of graduate students…….The most important thing today is for young people to take responsibility, to actually know how to formulate an idea and how to work on it. Not to buy into the so-called apprenticeship.” - Sydney Brenner
Perhaps misguided incentives are driving scientists to pursue incremental gains rather than attempting risky scientific breakthroughs. Perhaps academia is creating a hypercompetitive environment favouring careerists rather than pioneering scientists. Perhaps perverse scientific funding incentivises unambitious research.
For a look into life as an artist in Renaissance Italy I recommend "The Autobiography of Benvenuto Cellini" maker of the famous silver and gold ware on the Pope's table, soldier, and man-about-town.
Also, I want to point to the effects of the Internet on software development as a modern confluence of very energetic thought. The co-operation of Open Source, the widespread availability of tools and libraries, the forums, the ability to work remotely, are all parts of this story. You used to be able to get an engineering job without a degree (like me), just by coding (and working very hard on your own). Possibly this is waning due to academic requirements, student loan baggage, maybe a plateau of solved problems. And AI - what will it do to software engineering? Nobody knows.
Thanks for reading- I’ll definitely add The Autobiography of Benvenuto Cellini to my reading list!
LLMs are going to make the job market very interesting in the future. They lower barriers to entry for all people/professions. So, formal education is going to lose its comparative advantage (NB: Jobs most exposed to LLMs require under/postgraduate qualifications)