"My real bottom-line hypothesis is that nobody has a sweet clue what they’re doing. Therefore you better be trying stuff at an insanely rapid pace. You want to be screwing around with nearly everything. Relentless experimentation was probably important in the 1970s-now it’s do or die."