[anigbrowl]

Left to themselves,a significant number of people are awful. You know, between natural language processing, sentiment analysis, clustering and so on, IDing things like hateful tweets or persistently hateful tweeters seems well within the capability of today's AI. Nor do I think it would be too hard to find funding or a revenue stream for a reliable Asshole Detector. I mean, companies don't want to hire assholes, and browsing the news would be so much more pleasant if comments from assholes were pre-emptively labeled as such.

[djsumdog]

Twitter does censor stuff, and that's not a good thing. I mean no, you shouldn't be a bully/asshole/send death treats and that's a horrible thing to do; but the censoring and removing posts of people you don't like starts to create an echo chamber. People with non-hateful grievances can have their posts swept up in the same net.

When people are anonymous, they tend to be more candid about what's really deep down inside. The solution isn't to create A.I.s to censor everything. The tweets are a symptom of frustration in society. The real problem is much more difficult to fix.

[drdeca]

Well, surely you don't mean that the people receiving the bad messages should have to read them, because of course you agree that they don't have to look at twitter at all. But, why would it be important to require them to see these in order to see other posts on twitter, if they only want to see the other ones.

So, I don't see what the problem would be in providing the users with an opt-in filter option. If they are concerned that it could put them in a filter bubble, they could just not enable it, and if they are not concerned about the possibility of a filter bubble, and find the bad messages unpleasant, they could enable it. This would all just be following user choice.

Yes, if the features you give people for filtering messages are set up in certain ways, they may make it a bit easier than would be best for users to create filter bubbles for themselves, but it would at least be the user's own choice. As long as someone who wants to not get in a filter bubble can reasonably avoid it, then it seems like that's good enough on that front to me.

edit: Ideally, I think, the user would be able to fine tune the filters that they choose to use, and the simplest options would be such that they do not cause much of a filter bubble. But I think to really respect user freedom, allowing the user to create a filter bubble for themself (perhaps using 3rd party tools) is maybe necessary?

[dom0]

> When people are anonymous, they tend to be more candid about what's really deep down inside.

Actually no, there was a study about hateposting against immigrants conducted on German social media posts, the result was IIRC that most hateposts were actually released under a real name, and not pseudo- or anonymous.

[vacri]

> When people are anonymous, they tend to be more candid about what's really deep down inside.

And I, 'vacri' can respond as such to you, 'djsumdog'. However, if either of us started sending rape or death threats here on HN, we'd be shadowbanned at a minimum.

There's a difference between being candid, and being an arsehole because you're anonymous. Just because someone is frustrated doesn't mean they can take it out on random passers-by.

[anigbrowl]

The censoring and removing posts of people you don't like starts to create an echo chamber

Yes, I'm very well aware of this. The thing is, I really don't care any more. When I was younger I used to think that if you would just keep peeling away the layers of awful from a vicious person eventually you'd find some root cause that would be worth fixing and so on. As I've gotten older I've come to the conclusion that some people are just assholes, are never going to change, and that studying them further is not going to result in any great revelation which will heal the world.

For myself, I don't worry about ending up in an echo chamber. I devote a certain amount of time to studying people I don't agree with - I've monitored neo-nazi forums and related pools of internet extremism for nearly 15 years as I have an interest in knowing what those people are up to and how they think and so on. So I'm not worried about my political instincts atrophying from only being exposed to opinions that I agree with.

On the other hand, when I'm not actively monitoring such people and just want to pass some pleasant time for my own enjoyment, I would really like if there was a way to just screen out bullshit and offensive chatter without sacrificing all social content in the process.

[mwfunk]

It is a problem and it is difficult to fix, but the short term solution is not to do nothing, which is all anyone seems to propose (if they propose anything at all) when this topic comes up.

In general I disagree that toxic/hateful tweets are a symptom of frustration in society- they're mostly just a symptom of many humans being hateful jerks, plain and simple. I don't like echo chambers either, but I really don't think that there's any crucial intellectual gold hiding amongst the masses of toxic garbage out there. To a lot of people, posting on the Internet is akin to writing random garbage on a bathroom wall.

[ianai]

You can label them without deleting them. It'd be no different than a up/down vote at that point.