Comment by caddemon

To contribute to fundamentally new knowledge in software or CS also can require quite a lot of background learning though. I think the bigger culprit is that biomedical studies are fucking expensive. There needs to be a big upfront commitment, so there is naturally going to be more reliance on credentialism/gatekeeping. If 99% of random self-driven software projects turn out useless to the rest of the world, that's no big deal, and then we just hear about the 1%. But it's not possible in biology.