The whole "Facebook manipulating emotions" fiasco is one of the most ridiculous things I've seen happening in the field of social sciences in front of my own eyes. And it illustrates a very dangerous thing about science reporting - hyperbolizing. Almost nobody reads the primary sources (like actual papers), so when a journalist starts using strong words like "manipulated", it colors people's interpretations of the issue, leading them to severly overestimating the meaning of the issue - lo and behold, you can create a shitstorm out of thin air.
The way press was describing Facebook's "manipulating users' emotions" is akin to describing sneezing as an act of biological warfare. Scope matters.
Anyway, I see how people can have problems with Facebook's actions here (though I personally don't, and I'd actually encourage more of such studies - it's an unique vector that could lead to actually meaningful results in social sciences). But the point is, it was totally overblown in media reports, and it still is being totally overblown by commenters on Hacker News. Facebook could not drive you to suicide on purpose by this, so please everyone, stop saying as if they could.
[1]: https://www.google.com/search?q=facebook+study+manipulating+...