It is very easy for us humans to identify troll accounts. Why is it so hard for Twitter to take some precautions? I know the complexity can increase exponential

It is very easy for us humans to identify troll accounts. Why is it so hard for Twitter to take some precautions? I know the complexity can increase exponentially (relevant xkcd: https://xkcd.com/1425/ https://xkcd.com/1425/ ) but at least some basic measures like monitoring ip addresses and cross retweeting/liking could help filter out most of the basic troll accounts. HN does it AFAIK, why can't Twitter? Edit: From the downvotes, I think there is a misunderstanding. When I say troll-accounts, I mean automated or semi-automated accounts that are created for the sole purpose of creating a hostile atmosphere for people in the other camp. I don't call the people I disagree with as "trolls".