What Twitter needs is algorithmically enforced moderation. So when a twitter reply calls for extremism, the user will then be prone to see more calls for peace in reply to that comment. Or when somebody asks for dox on a subject, the user then sees tweets mentioning all the previous times doxxing went wrong and innocent people were hurt. The user would not know this was taking place, but they would avoid radicalization and echo chambers currently happening on Twitter. Also instead of showing the opposition to break their thought bubble, which probably wouldn’t be helpful, the user sees their own side, just more moderate and centrist. If no tweets showing moderation exist, twitter could use a bot that appears to be a real human to make these moderating tweets. Most of these twitter conversations are bots anyways.