Whenver I see something about a black box algo in healthcare it makes me think of this study: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.

Whenver I see something about a black box algo in healthcare it makes me think of this study: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0141357 https://journals.plos.org/plosone/article?id=10.1371/journal... of pigeons trained to read radiology images. Obviously if a doctor was to tell you that the pigeons thought you had cancer you would think it is ridiculous, but if an algorithm said the same thing and was just as likely to be right it comes off as much more credible, yet it might not be any more understandable or effective.