June 18, 2012
In R&D, the rate-limiting step (to use a term from my chemical background) is usually not the number of people working on a problem. Not after a certain point, at any rate. Automation and miniaturization have been changing that, as in so many other industries. We can test more compounds and generate even bigger server-choking piles of data faster than ever before. The problem is figuring out what all of those numbers mean and what they’re telling us to do next. The failure rate for new drug candidates going into human trials is well over 90 percent. That’s my industry’s problem, right there, and throwing more people at it won’t help much. What we’re short of is great ideas that will help us stop doing things this way.
And those ideas, needless to say, are probably not going to be supplied by Plotz’s Army of Mediocrity. I think that many other areas of science have the same problem. What worries me the most about the future of R&D in this country is whether we’re still attracting the smartest and most capable people to do it. I keep having dark thoughts about the next Claude Shannon—one of the leading mathematicians of the 20th century— refining risk models for derivatives trading, rather than changing the world.
I was talking with someone the other day who advanced the proposition that there are probably only 50 really first-rate scientific minds produced in the United States every year. And then came the question: Does the current system of training and funding scientists encourage those 50 to stay in the game, or to find something else to do?