> If AI solved cancer, disease, anything that lowers quality of life, but you have no idea how it did it: is that good enough? A lot of medicine is already lik

> If AI solved cancer, disease, anything that lowers quality of life, but you have no idea how it did it: is that good enough? A lot of medicine is already like this. Shown to work in clinical trials, no complete end to end mechanism understood. They still get approved if the empirical results are strong.