The idea (AFAIK) is that both are incremental to a degree, but it’s much more incremental to try things at random than to improve your theoretical understanding, because improved theoretical understanding is relatively likely to generalize to multiple future discoveries while random experiments are more likely to be one-and-done.
I read it as
Biotech startups currently “experiment with things that just might work”, which is [apparently?] incremental thinking.
Instead they should refine definite theories about how the body’s systems operate [which sounds incremental to me, too].
The idea (AFAIK) is that both are incremental to a degree, but it’s much more incremental to try things at random than to improve your theoretical understanding, because improved theoretical understanding is relatively likely to generalize to multiple future discoveries while random experiments are more likely to be one-and-done.