Discover has an interview with genius physicist Murray Gell-Mann. Gell-Mann on Heisenberg:
But by the time I knew him, although he was not extremely old, he was more or less a crank.
On Feynman:
[He] was pretty good, although not as good as he thought he was.
On Fermi:
[He] was good, but again with limitations—every now and then he was wrong. I didn’t know anybody without some limitations in my field of theoretical physics.

It's got to be irritating to be asked about Feynman every interview he gives, but hey - the dynamic between Gell-Mann and Feynman was legendary, and a lot of us can't hear enough anecdotes about it.
The most intriguing quote in the interview is this:

Along with several other people around the world, I’m looking to see if there might be alternate ways to mathematically characterize entropy, the measure of disorder of a system. It might be useful to employ alternate formulas for looking at different circumstances such as financial markets or social interactions. Maybe this will turn out to be an extremely flexible tool for handling all kinds of situations. That’s what people hope. Other people think it’s nuts.

I can see (in a very vague, non-specific, crankish way) how this could potentially be extremely useful for modeling biological systems.