> I’m not under the impression that the replication crisis in the social sciences looms as large for others as it does for me Not just social sciences. I decid

> I’m not under the impression that the replication crisis in the social sciences looms as large for others as it does for me Not just social sciences. I decided to get out of biology when I helped an adjacent lab with some machine learning, discovered that the entire effect they were looking at was due to the old gotcha about groups of correlated files getting split across the train/test barrier, gently communicated the bad news to the person who had asked for my help... and a year or so later saw the uncorrected result on the front page of Nature. Brutal competition against brilliant peers for shit pay is one thing, but this? This was a step too far. I couldn't compete with this. I hit the eject button and didn't look back.