Don't ask what they think. Ask what they do.
My rule has to do with paradigm shifts—yes, I do believe in them. I've been through a few myself. It is useful if you want to be the first on your block to know that the shift has taken place. I formulated the rule in 1974. I was visiting the Stanford Linear Accelerator Center (SLAC) for a weeks to give a couple of seminars on particle physics. The subject was QCD. It doesn't matter what this stands for. The point is that it was a new theory of sub-nuclear particles and it was absolutely clear that it was the right theory. There was no critical experiment but the place was littered with smoking guns. Anyway, at the end of my first lecture I took a poll of the audience. "What probability would you assign to the proposition 'QCD is the right theory of hadrons.'?" My socks were knocked off by the answers. They ranged from .01 percent to 5 percent. As I said, by this time it was a clear no-brainer. The answer should have been close too 100 percent.
The next day I gave my second seminar and took another poll. "What are you working on?" was the question. Answers: QCD, QCD, QCD, QCD, QCD,........ Everyone was working on QCD. That's when I learned to ask "What are you doing?" instead of "what do you think?"
I saw exactly the same phenomenon more recently when I was working on black holes. This time it was after a string theory seminar, I think in Santa Barbara. I asked the audience to vote whether they agreed with me and Gerard 't Hooft or if they thought Hawkings ideas were correct. This time I got a 50-50 response. By this time I knew what was going on so I wasn't so surprised. Anyway I later asked if anyone was working on Hawking's theory of information loss. Not a single hand went up. Don't ask what they think. Ask what they do.