A few points:
Yes, this study did you AI agents. That can cast some doubt on the results as AIs aren't humans, but that does mitigate some of the ethical issues these kind of studies usually entail (yes, I'm looking at YOU U of M.)
The study is trying to foster "non-partisanship" and "democracy", which...is not what a lot of people use social media *for*?
The study also doesn't factor in minorities or baseline societal bigotry. Another reason many turn to social media instead.
