I've always felt that the human-centered approach to computer science leads to more interesting, more exotic, more wild, and more heroic adventures than the machine-supremacy approach, where information is the highest goal.

America's Facebook generation shows a submission to standardization that I haven't seen before. The American adventure has always been about people forgetting their former selves - Samuel Clemens became Mark Twain, Jack Kerouac went on the road. If they had a Facebook page, they wouldn't have been able to forget their former selves.

A remarkable thing about the Silicon Valley culture is that its status structure is so based on technical accomplishment and prowess.

Human beings either function as individuals or as members of a pack. There's a switch inside us, deep in our spirit, that you can turn one way or the other. It's almost always the case that our worst behaviour comes out when we're switched to the mob setting. The problem with a lot of software designs is that they switch us to that setting.

I love science fiction stuff - I'm a bit of a dweeb like that.

My complaint is that there are more books and news articles than there are primary scientific papers. I am probably the biggest critic of the hypesters, because it's dangerous when fields get overhyped.

With any good projects, I feel like the off-screen chemistry factors on-screen. It's great when you don't have to force it, but when it's not there, you better focus on getting there because, as we live with these characters, we spend more time with one another than we do our families at home.

You can't argue with physics, mate.