A big part of my job is to remember failure, and reboot failure, when the timing is right.

Behavioral economics can explain some things, but it's hard to explain a lot of the underlying processes that generate these decisions, much less some of these unconscious things that we don't have a handle on at all.

I worry that we don't have a very good definition of consciousness yet which makes it hard to tackle.

One thing that I've been doing for a long time is to wake up really early. I try to get up around 4 or 5 in the morning, long before most of my lab members are up, which gives me some quiet time to really think without distraction. I think that's important.

Work backward from your goal.

If you could map out a human brain, an open question is, if you simulated it, would it be you? Now, as we discussed earlier, we don't have a great definition or even a good technological handle to know whether something is conscious or not just by looking at it, so there's that aspect that we're not ready to answer, I would argue. But it raises very interesting questions about the nature of identity.

Synthesize new ideas constantly. Never read passively. Annotate, model, think, and synthesize while you read, even when you're reading what you conceive to be introductory stuff. That way, you will always aim towards understanding things at a resolution fine enough for you to be creative.