3 ms·
>Simply do not follow these people if you don't want to see "abuse". I don't think that's the solution of abuse, or even the source - trolls on Twitters keep o
by a_bonobo 11y ago
>Simply do not follow these people if you don't want to see "abuse".
I don't think that's the solution of abuse, or even the source - trolls on Twitters keep on making new accounts to @mention their "targets" with vitriol/threats of violence/etc. The attacked can keep on blocking these new accounts, but that's it. By default the Notifications tab has everyone mentioning you - you can set that to "People you follow", but then you can't interact with your audience, which is the major feature of Twitter.
If I do, let's say, science communication on Twitter, I want to be able to interact with every "stranger".
- x5n1 11y agoneed ai or Bayesian filter to detect such attacks and block them from most users. But leave the user and feed alone, just make sure the target of the attack and the general public never see the messages. Much like a shadow ban.
- majewsky 11y agoI'm skeptical of any form of filtering or censorship. How can you know which rules the algorithm abides by? If it's some deep-learning-based AI, how would even Twitter know exactly which rules the AI abides by? A more transparent solution could be a prominent warning mark on tweets from people who have more people blocking them than following them.