Post #3533735

every week of work with biologists brings another “guess we doin circles now” moment. there’s simply so many things that were never measured, measuring them requires knowing what are you doing, proper experimental design and interpretation of results, that are affected by myriad factors and dozens of feedback loops. none of that is in training data of course, and even if it was it would help very little because humans measuring that would make better use of these results first. but don’t worry about that, machine god will learn relativity from three video frames and will make peter thiel immortal right away