If you're a scientist, it's best for your career to do small incremental improvements on previous research. That guarantees a steady stream of publications. I

If you're a scientist, it's best for your career to do small incremental improvements on previous research. That guarantees a steady stream of publications. If you go for a big idea and are wrong, then your career is over. Also, peer review is censorship. If you're doing something radically different, it's less likely to pass peer review.