Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't understand why Twitter hasn't implemented some kind of reputation gradient that would allow users to filter by levels.

Yes, I realize they are tricky to implement and subject to abuse but imagine you got points for using your real name, real photo, demerits for insults etc, then users could filter out hordes of trolls and gutter nonsense and instead focus on the meaningful conversations that do actually occur there -- the verified checkmark is a proto-version of this.

Instead they rely solely on blacklisting, which is a never-ending game of whack-a-mole. Even HN has a more sophisticated system, where users can be shadow banned or have their comments appear lower than normal.

Not sure how useful this macro-scale philosophizing by Dorsey on Twitter itself is for delivering new, useful features to users. Serious question: has Twitter released any new, major features in the last 2-3 years?



> Even HN has a more sophisticated system

HN has such low volume that it can be centrally curated by a few people who enforce policies like "no flame bait -- and I'll know it when I see it." Or, for example, when someone writes a quick attack on HN, they are encouraged to flesh out their point with substance which is something that Twitter fundamentally discourages.

It's hard for me to see any way forward for Twitter. By design, it amplifies the worst characteristics of humans from tribalism to outrage fetishism.

For instance, even if you configured Twitter to never see posts from the opposing cultural faction or trolls (usually used as synonyms), the people you follow will retweet/quote them and bring them back into your feed.


Your point about Twitter's scale is fair, but I personally don't think the situation is hopeless, as you seem to suggest.

If you have troll engagers in your feed, unfollow them. But right now I have no way of filtering replies...it's frustrating. No system is perfect but let's at least TRY to improve the status quo.

If Twitter is consistently beta-testing new features to improve the quality of the UX/tweet replies, I personally am not seeing it.


But people like the drama even if reluctantly.

It's like how you can /ignore someone on IRC and phpbb, but nobody actually does. They only threaten it. Hell yes they want to be clued in when you open your mouth so they can bark back! Humans love it.

Look how many people waste their time responding to Donald Trump's tweets with things like "RESIGN!" They could unfollow Trump. But they choose to spend their precious time on earth bickering about Trump because it's pure entertainment.

What also happens though is that we kinda know this. The more astute/introspective people will realize Twitter isn't bringing out the best in us. Sam Harris lamented this a while back and was able to quit Twitter for a couple days before being driven back to the crack.

I find myself doing this on HN too. Ever read a headline like "Javascript sucks", know the arguments you'll find in the HN comments, and click in to respond to one? The whole reason I'm here is to socialize, so I don't think filtering can be very effective. I'm not going to filter out the very things that draw me like the opportunity to call someone out on the internet, which is the same reason why nobody uses the ignore feature on forums.

Nobody would use the block feature on Twitter if it didn't tell the other person you blocked them.

There are some bugs in basic human psychology that I'm not so sure technology is ready to solve. Maybe this is the Great Filter that no civilization can surpass in the Fermi paradox.


> Nobody would use the block feature on Twitter if it didn't tell the other person you blocked them.

Well, there is a "mute" feature which acts a lot like a "silent" block, and plenty of people use that.


The block function makes it inconvenient for trolls to read your tweets and impossible for them to RT them for a pile on. It's a "last resort" before moving to a protected account (which can't be RTd at all)


Personally I don't consider real name or real photo to be much of an asset of any of the circles I was in on Twitter. In fact, in my experience the ones with real name/real photo (or what seemed that way) were generally the least interesting posters. They were never willing to put out their own opinion, especially on controversial topics. In fact, in the kind of groups I was in (let's say political activism), having your real details on there is anything but an asset.

Not having a real name on Twitter isn't even a good heuristic for anything.


In a lot of circles, a Twitter profile with a "real photo" as the profile picture usually means you're looking at a spambot with a profile picture scraped from either Facebook or a dating site.


It's not a good heuristic for unique opinions -- sure (see: LinkedIn, the most horrifyingly boring social network feed). But it is if you want to filter out hate speech, which a lot of people do.

But my point is not real photos/names as the be-all, end-all. How about many heuristics, and let users filter accordingly? All this data available and none of it is used to improve the user experience, at least as far as I can tell.


Less than 5% of the accounts I follow have a checkmark. Most of the most interesting ones don't use their real name or profile pic.

The real problem is that reputation can't be absolute but is relative to a context. Hence the blockbot partial solution.


As I mention in another reply -- I'm not saying this filtering would result in the most interesting feed, but it's frustrating you can't filter your feed at all.

If some people want more interesting opinions, great. If others want more bland but less abuse, also great. Right now you can't filter for anything and most of tweet replies are a giant waste of time.


I'd like that as well. Make the process so anybody can be verified then let me configure my client to only show stuff from verified users.

I'd like to see them try other things too. One of the nice features in G+ was that the owner of a thread (the originator) could delete replies. I think that could work nicely in Twitter too.


> One of the nice features in G+ was that the owner of a thread (the originator) could delete replies. I think that could work nicely in Twitter too.

Interesting; I think adding this feature without changing anything else would severely exacerbate the already intolerable echo-chamber effect. Sure, trolling and abuse would lose visibility, but so would even slightly contrary opinions. Thought bubbles would shrink even further, and the platform would become even less enjoyable. (Context: I'm a light Twitter user, reading ~10 min/day and sending 1-2 tweets/month)


Maybe. But does Twitter even experiment to see?

Take Maggie Haberman, NYT reporter. Her Twitter replies are always a cesspool...what if she could lock down her tweets so only a whitelisted group of users (say people she follows) could comment? You'd still (hopefully) get contrary opinions but none of the dreck that dominates her feed now.

I don't know if that's the answer. But I don't see Twitter even trying, just Dorsey waxing philosophical about it.


What I call "Twitter Communities" is my answer to that problem. Invite only groups. The world can read but only group members can post to the group. Otherwise exactly like Twitter currently is. You can create a community and invite people to it and then when you post you either post to a community or global, everything still appears in your feed but you can filter down to just a certain community posts.


She can moderate her replies, but what if we want to see all the replies? What if we can choose our own moderators?


That sounds too complicated.

If it's her thread, she should be able to curate it in anyway she wants.


It's worth trying. If it makes Twitter worse they can always undo the change.


> Serious question: has Twitter released any new, major features in the last 2-3 years?

Double-length tweets is the most major thing that comes to mind for me.


Yes. You're right. But even that -- how innovative is that as a feature?

I guess what I mean is: what innovative features have they introduced?


I think the features themselves aren't ground breaking on their own, but I'm willing to believe they do a lot of large scale semantic text analysis (eg for abuse prevention) which is innovative at least from an infrastructure standpoint and their size


wouldn't it be incredibly easy for this kind of system to be abused by opposing cliques? erroneous mob reporting is already common enough, giving aggrieved parties the ability to do cumulative damage to someone seems like it would only amplify that behavior.

>Instead they rely solely on blacklisting

twitter has implemented very heavy handed throttling as well in order limit the reach of users deemed unsavory. whether this is a good thing or not is up for debate. in some sense i wonder if these networks have simply become too generalized to be moderated effectively in a manner that satisfies all of the different facets of the userbase.


If it's an improvement, it's almost certain that Twitter won't implement it.


I'll tell you why it's not based on that.

Becuase you might not like - or they might not want you to see things that are popular.

How much would you complain if your Twitter feed was filled with Trump tweets? Better yet, how much would they not like that!? - Sorry though it's based on reputation levels and his is up there relative to non-famous users.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: