This reminds me of the incident that led to me quitting Facebook: There was an article[1] about Facebook manipulation of users news feeds, and I had just joine

This reminds me of the incident that led to me quitting Facebook: There was an article[1] about Facebook manipulation of users news feeds, and I had just joined the company. So, I looked into it, and found without much effort the dataset and analysis of that experiment. One of the fields was "CauseOfDeath" and one of the analyses was that feeding a user all negative news created some whole-digit percentage increase in the suicide rate of users. This was presented matter-of-factly, and none of my coworkers seemed to see a problem with this. Many of Facebook's experiments would never pass a IRB. Actually, that's not true at all: modern IRBs have been stacked with the same kind of monsters. Facebook would never pass a sane and functional IRB. [1] https://www.google.com/amp/s/amp.theguardian.com/technology/2014/jun/29/facebook-users-emotions-news-feeds https://www.google.com/amp/s/amp.theguardian.com/technology/... is the one I found with a quick search, though there were a few dozen in quick succession over that time period; I think by timing it was probably one of the articles written in response to their apology that actually came to my attention first.