Thread by @Carnage4Life

There are two fundamental problems with social media and toxic behavior that boil down to human nature which we haven't solved.

1. Confirmation bias: People follow topics/people that they agree with & validate them. This is susceptible to forming in-groups that demonize others

2. Group dynamics and toxic behavior: Any online forum where you can form groups will attract groups of people with toxic behavior. This can lead to gamification of toxic behavior and one-upmanship leading to real world violence (swatting, mass shooters, Capitol Hill riots, etc)

Discussion in the U.S. usually misses these issues and focuses on ascribing malice to social media services. In Europe, just like with GDPR they're ahead and have proposed legislation on how content moderation and algorithmic disclosure should work. en.wikipedia.org/wiki/Digital_S…

I expect the DSA will help with group dynamics & toxic behavior by setting standards on content that needs to be moderated but this likely will then conflict with other EU laws around privacy. This conflict between privacy and obligations on big tech will be an ongoing EU issue.

Confirmation bias will be trickier even with disclosure on how algorithms work. People like content that makes them feel special and better than others. Sometimes it's harmless like these tweets, other times insidious.

Algo transparency won't solve racism or misogyny.