[jorleif]

Yes, but you say that as if it was a bad thing. For me it was a really hard experience when doing science, that I could not "be smart about it", and do things "in smart ways", because that would require me to be able to predict which experiments would work out etc. Any high uncertainty environments are "throwing shit on the wall". You can't be smart about the throwing part, only what you do when something sticks. This is very hard for smart people to accept. See also: http://www.ted.com/talks/uri_alon_why_truly_innovative_scien...