Thiel argued for enabling riskier research grant-making via institutions such as the NIH, as well as abandoning the scientific staple of the double-blind trial and encouraging the U.S. FDA to further accelerate its regulatory evaluations. He said that these deficiencies are inhibiting the ability of scientists to make major advances, despite the current environment that is flooded with capital and research talent.
“There’s a story we can tell about what happened historically in how processes became bureaucratized. Early science funding was very informal – DARPA’s a little bit different – but in the 1950s and 1960s, it was very generative,” said Thiel. “You just had one person [who] knew the 20 top scientists and gave them grants – there was no up-front application process. Then gradually, as things scaled, they became formalized.
He then cited the success of major scientific programs – such as the development of the atomic bomb in the Manhattan Project, the Apollo space program and Watson and Crick’s discovery of DNA – that hinged on having “preexisting, idiosyncratic, quirky, decentralized scientific culture[s]” and were accelerated rapidly by a major infusion of cash.
That was from the article, “Thiel calls for improving research grant, regulatory processes to enhance scientific innovation.”
I think there is a fair amount of ambiguity in the term “science.” The Manhattan Project, ballistic missile projects, and the Apollo space program were largely engineering problems. Most of the “science,” for example, in the Manhattan Project was performed up through Otto Hahn discovering fission in uranium in 1938-1939. After that, most of the theoretical work had been complete. The Chicago Pile largely proved out the theory and then it became an engineering matter of separating the U-235 isotopes and developing a trigger/gun.
I tend to agree with Thiel, however, that the formalization of the process stifles innovation. It seemed that in the past, more investment decisions were made on the basis of an individual’s anticipations about what is to be fruitful. There was less reliance on an action being grounded in “facts” for it to have merit. This created a bias toward action.
Today, there is a bias toward paper studies. Researchers must generate evidence that there is an articulated plan with a high likelihood of achieving intended outcomes. If the only projects able to get funded are those where the outcomes are well-defined and there is a definite path to success, then only conventional projects can be selected.
Thiel advises we select projects based on the people. Moreover, we should select people with ideas in between the conventional and the crazy. I think the point he was trying to make — in John Boyd’s terms — is that we shouldn’t select people who reason from deductive logic alone (they are good at making sure some scientific idea is consistent with reality), or inductive logic alone (they are great at coming up with exciting new scientific concepts), but you need someone whose creative synthesizing of a new concept is tempered by their rigorous analysis of consistency with observed facts.
Leave a Reply