Are we making more mistakes now? I don't think so. Science is a high-risk activity. And when you do science—this is very important incidentally for the general public, and for policy makers—if you are not wasting some of your money, you are not doing good science. It's a funny way to say this. You've got to back high-risk opportunities. And high-risk opportunities means some fraction of them are going to fail. And I think in any science funding scenario, you've got to say, 10, 20, maybe even 30% of your funds are going to be invested in failures.

Comments on need for failure in scientific research.