I don't think you can lump physics, life sciences, chemistry etc all together. This article mainly pertains to physics, which is notable for being more cathedral like than other fields, where prior work is very foundational. Physics is also a field where mathematical aptitude is particularly important, and there is significant variation between people in this dimension with some clear outliers.
The small elite theory doesn't work in life sciences in my estimation. There are no life science 'geniuses', it isn't that sort of field. The scientists that won the prize for CRISPR have many excellent qualities, but they are not really distinct from millions of other scientists with similar qualities. I have had a bit to do with 2 other life sicence nobel laureates, and I can verify that they are not John von Neumann in a lab coat. The important thing for CRISPR is that Jennifer Doudna and Emmanuelle Charpentier had the opportunity to be studying something like repetitive sequences in bacterial DNA. If you cut science funding, you could easily annihlate fields like this. CRISPR might have taken another 20 years to be discovered, or longer. Charpentier struggled to get funding, living from grant to grant in small labs until she hit upon CRISPR.
Life science research is also conducted in a weird way, where usually the PI mostly does administrative work while grad students / post docs do the actual experiments. It’s weird that they become experts in their field and then put their pipettes down. I feel like in physics the Van Neumann types are always doing research, always at the chalk board, etc.
At least in condensed matter physics I can say that the PIs are not in the lab at all doing research. They lead from behind the desk and the grad students and postdocs are the ones running the experiments and data analysis. The PIs just react to the data presented and offer guidance.
It's a weird system, before they became professors they were running experiments in a lab and were able to get good results to get published and become a professor, clearly there's a skill set there. But the minute they get professor job we take these people out of the lab and put them behind a desk. It's kind of like yanking LeBron James off the court during his prime and making him coach.
Yeah, I saw this in action when I was a grad student at UCSF. In one of the labs I rotated in, the PI (pre-tenure professor) still had a lab bay filled with the most impressive display of reagents I've ever seen (he was a physical chemist). He would go over, look at it sadly, and then disappear into his office to write grants all day. What a waste of talent.
That's more an experimental vs theoretical boundary. It's like hardware vs software, the former just has extra physical constraints that introduce significant overhead that is most efficiently managed in a hierarchy so there are very few practicing PIs.
AFAIK there aren't many (or any) renegade theoretical physicists/mathematicians working on teams at the LHC or ITER or similar sized projects, as an extreme example. Just getting plugged into the ecosystem and getting time to run an experiment is close to a full time job, let alone actually designing a meaningful experiment. Hell, even acquiring lab equipment for small scale experiments is a huge time sink, especially for the nicer stuff that's so expensive it requires labs pooling together and fighting over the allotments.
I'm not so sure that PIs not doing any real work is the real problem though based on my reading of The Structure of Scientific Revolutions and personal experience. Science is about breaking new ground and I'm not convinced that seniority instills ability like it does with mathematics and engineering, where pattern matching builds on itself and its utility increases exponentially. Seniority in science does improve one's ability to navigate the scientific bureaucracy, but the problem is that the PIs accumulate power and the entire system calcifies around them. It was already such a noticeable phenomenon that Plank coined his principle in 1950 ("Science progresses one funeral at a time") when there were far more individual PIs who did their own work. Now we've taken it to the extreme where each subfield has a small group of larger labs creating an incestuous circle of peer reviewers, PIs, and postdocs who end up dominating the subfield just out of sheer numbers and prestige.
> AFAIK there aren't many (or any) renegade theoretical physicists/mathematicians working on teams at the LHC or ITER or similar sized projects, as an extreme example.
the research productivity at LHC and ITER is extremely low. I wonder if these are correlated?
The small elite theory doesn't work in life sciences in my estimation. There are no life science 'geniuses', it isn't that sort of field. The scientists that won the prize for CRISPR have many excellent qualities, but they are not really distinct from millions of other scientists with similar qualities. I have had a bit to do with 2 other life sicence nobel laureates, and I can verify that they are not John von Neumann in a lab coat. The important thing for CRISPR is that Jennifer Doudna and Emmanuelle Charpentier had the opportunity to be studying something like repetitive sequences in bacterial DNA. If you cut science funding, you could easily annihlate fields like this. CRISPR might have taken another 20 years to be discovered, or longer. Charpentier struggled to get funding, living from grant to grant in small labs until she hit upon CRISPR.