Hacker News new | comments | show | ask | jobs | submitlogin
Twitter’s Plan to Get People to Stop Yelling at Each Other (www.buzzfeednews.com)
145 points by minimaxir 8 days ago | hide | past | web | 181 comments | favorite





> Over several days this spring, BuzzFeed News met with Twitter’s leadership and watched as twttr’s team worked on its first big push: helping people better understand what’s being said in often chaotic conversations. The team thinks that if people took more time to read entire conversations, that would help improve their comprehension of them.

I had assumed that Twitter was aware of the implications of its design for conversation and had decided that the controversy boost to engagement numbers was worth creating a toxic environment. For real, consider these two basic elements of Twitter's design. First, any given post has a low maximum length, so any time you want to say more than a sentence or two it needs to be spread over multiple posts (a "tweetstorm"). Second, Twitter has two separate features that let anyone pull any part of a series of posts out of its original context: retweeting and quote-tweeting.


Yeah, I don't get this at all. I think it's even deeper than designing for controversy. I'm "older" (40s), though I work on products used by younger people, so I don't think this is just me being old, but I can't follow a conversation on twitter at all. When there's a link from HackerNews to Twitter, I click on it and sometimes see the thing mentioned in the HN title, and sometimes don't. Then there's a bunch of stuff below it, some of which is definitely replies to the original post, and some of which appears to be part of a completely different conversation. Some threads of conversations get cut off for no apparent reason and you have to click something to see more of them. Nothing is indented, so it's really hard to tell who's replying to whom.

I basically can't understand a thread on Twitter, so I've never signed up because who the hell wants to be part of that? I usually also don't go to Twitter to read stuff other people link to because I'm not going to be able to follow it. Forget boosting controversial stuff, I bet a lot of users post the way they do because they also don't understand what they're looking at.


I basically can't understand a thread on Twitter

I don’t think Twitter understands threads on Twitter, and they never have. They built something over a decade ago, it took off for reasons they never fully could fathom, and they’ve been afraid ever since to futz with what for them has been a good thing.


Once upon a time, there was IRC for single-stream chats, and NNTP for threaded ones, and all was good. Twitter used to be one of the former, but then they noticed that users were using the "@name" convention to simulate replies, and started officially supporting it. Instead of a single conversation, or a tree of nested replies, they created some unholy thing partway between the two, that serves neither role well.

Twitter threads are a tree, but the representation in Twitter's web interface and mobile apps is linear and it's not obvious which tweets are in response to which. There are hints in the interface if you're aware of them, but it's easy to get lost.

They ought to figure this out or let people build alternative clients...

I'm a similar age as you. I did create a Twitter account once, but never used it more than once or twice.

Unfortunately the Twitter username (handle?) I chose happened to be the same as the name of a somewhat well-known RnB/Hip Hop video producer (who I had no knowledge of before I created the account).

I signed in a year or two later and found that I had a couple of hundred follow requests from various aspiring and up-and-coming rap artists. At that point I realised in all honesty that I probably had no real use for Twitter.


I'm young, used twitter for about a year, and I don't understand twitter replies either.

> I basically can't understand a thread on Twitter

I built a Twitter client just for that, because I find Twitter threads hard to read (https://threader.app).

It turns threads into a single page so it’s easier to read, like an article. We are also working on a way to compile conversations with multiple users: https://threader.app/conversation/1123033196642201600/vpw88S...


Some people used to feel this way about IRC.

I'm afraid the reality is probably that you just don't have the motivation to figure it out, for whatever reason. No mass-market product like Twitter could succeed if it was as confusing as you say it is.

There's also the fact that a subset of computer programmer types have low tolerance for all sorts of "noisy" things that the average person doesn't have a problem with.


I'd separate use cases. If you want to talk to/about yourself or if you just want to rant at somebody, post memes, etc then Twitter is easy and natural to use. If you want to have meaningful nuanced conversation, Twitter is useless. So you end up with a dichotomy where for some people Twitter is a fine fit and for others it's completely unusable. Both are correct.

I'm actually a Twitter user, and I think this is a very jaundiced opinion. I have a lot of different interests and following people on Twitter is a very good grapevine system. Math twitter is a good place, for example. A lot of people are very enthusiastic about how it has enriched their mathematical diet. There are many good people on Twitter who use it for serious sharing of ideas in other fields as well.

In general, it's a good way for people to share thoughts and links to a diverse audience. Liking Twitter isn't a sign of foolishness. Among the social networks, it's not as gross and corporate as FB, or as superficial as Instagram.

Your dismissal of the reasons why people might like Twitter, the implicit claim that it's useless for sophisticated people like yourself, and the apparent desire to impose black and white dichotomies on the world risks making you into a candidate for the category of "just wanting to rant at somebody".

We know that HN has its own serious cultural problems, after all.


IRC is also terrible for conversations.

IRC is fine for conversations, though it's a different model of conversation than many mediums. All comments from different threads of conversation are intermixed, and you must separate them out mentally.

This is very different from threaded models, but it's something that can be adapted to easily enough, same as how most people can talk within a group in a noisy bar, where there are multiple surrounding intelligible conversations. Like in a bar, it's easy to jump into another adjacent conversation spontaneously.


More important IRC don’t enable to browse history as far as I known.

Meanwhile tweet can persist enabling asynchronous conversation, so it’s more like a very lightweight forum than IRC.

Edit: But the tree structure for reply is actually more similar to a HN thread with a less visible structure.


It isn't but solely because IRC channels are of a reasonable size, and have some form of moderation.

If you wanted to chat in a channel with 1000s of people (like twitter), the IRC model would be untenable.


Twitter, reddit, slack and others are meant for constant asynchronous engagement, mostly by schoolkids, and not random access or long-lived discussion. Coming from the message board world, the difference in the attention spans involved is pretty frustrating.

And 9gag. Strong market for short attention spans.

This is why I use https://threadreaderapp.com on tweetstorms. Kinda silly, but it's the only way to properly read threads, as even Twitter itself seems to mess up the ordering.

I've never understood the appeal of Twitter. Today I read Spacex and Musk talking about Starlink on twitter and I could not fathom who was talking to who, it seems like abunch of random statements with tons of animated gifs and stupid jokes. How does anyone navigate that mess? People say such stupid things. And spend so much time writing it out and finding a gif and ... for what? It's just noise. What motivates them? When do they find the time? What are they hoping to achieve? That Elon Musk replies with a lol. Couldn't Twitter just look at your last 40 posts and if no one has interacted with you then it just deletes your account. Sorry buddy, no one cares.

I had the same problem with Twitter too. The posts and replies seemed random and without any structure and I didn't really get it. But after I spent a few weeks using it, it started making sense and I would browse posts and replies without too much effort.

I think it just takes a while for new people to get Twitter. Probably most people don't even know why Twitter is confusing initially because they are so used to it.


> I can't follow a conversation on twitter at all.

Chaos is a Ladder [0]

[0] https://www.thenarcissisticpersonality.com/narcissistic-chao...


Simpler: stop showing tweets just because other people liked them. That's not a reshare, so it shouldn't be treated that way.

More generally, discourage large conversations and encourage fragmentation.


That's why someone bought this domain [0] which redirects to a search excluding tweets [1] not authored by the people you follow. It's way sparser for me (I follow roughly 200 people) than the regular twitter feed so I also waste way less time whether procrastinating or looking for tweets by people I actually follow. I deleted "twitter.com" from my browser history and always start at realtwitter.com now.

[0] http://realtwitter.com

[1] https://twitter.com/search?f=tweets&q=filter%3Afollows -filter%3Areplies&src=typd


Wow I just actually enjoyed using twitter for the first time ever. I'll keep that link in mind.

I always found this so weird. They themselves provide a literal "retweet" option, but then treat likes almost the same as retweets. What's the point of having both then??

because publishing business behave like an AI that fixed the goal of increasing revenue as optimizing for engagement... instead of paperclip production.

I stopped liking tweets because I don't want to reshare them to my followers from other random walks of life.

It's not just that. Timelines are generally unavoidably toxic. It's not going to stop people from retweeting things and then shouting about it, or writing condescending, smug threads starting with "We need to talk about ...".

We need to talk about people who think the phrase 'we need to talk about...' is a fruitful way to start a conversation when "In my opinion" is probably more welcoming and conducive opening to a discussion that doesn't turn into ideological jousting right from the start.

> stop showing tweets just because other people liked them

There's already a way to do this: use any third-party client. You even get the added bonus of a reverse-chronological timeline!


That works for individuals, but the social interactions won't change unless a lot of people do it.

I think this is a very important point. Even when an individual user can opt-out via settings, the societal-level impact is almost entirely determined by the defaults.

If you have a relatively-short, curated list of people you want to see posts from, I suggest adding them to a list and viewing your "feed" that way. I've been using lists over the home page/feed for a few months and my experience w/ Twitter has been much better. Only posts from the people you've explicitly chosen to get posts from, which also means zero ads (at least on desktop).

I used to follow a couple hundred journalists/researchers etc. "real people" and eventually moved to curating each segment through different accounts and allowing each account to assimilate into the communities instead of a clusterfuck of interests in one timeline (yes i know lists exist but like fb groups it doesnt really serve the purpose of exploring content, not as easily at least)

its been relaxing, carefully curating people that complain/add negativity to my feed out and adding more interesting accounts as time goes on


whatever makes the experience tolerable and worthwhile

I never (no longer, that is) see tweets that other people like. I think there is a setting to disable them from appearing now. At the very least, whenever one appears you can click the little arrow next to it and select "see less often" (or equivalent).

Yup, selecting "see less often" once seemed to make them disappear entirely for me.

Funny, I don't see any liked tweets. Either the handful of people I'm following don't like anything, or there's a switch somewhere that I've flipped without paying attention, because now I can't find it.

I actually like the feature that shows likes from people I follow. I also like the non-chronological timeline.

To me, both these features reduce the frustration inherent in Twitter's over-abundant timeline where you know you're missing out on 95% of stuff because there's no time to scroll through the whole thing. I know I'm still missing out on 95%, but it feels more prioritized.


> where you know you're missing out on 95% of stuff because there's no time to scroll through the whole thing.

Sounds like Facebook.


> More generally, discourage large conversations and encourage fragmentation.

Doesn't that encourage bubbles?


I started using Tweetdeck specifically because of how annoying this was.

The product has "twit" right in the name. Twitter is where you go to read twits or become one. Why is anyone surprised?

From its tiny size limits to its byzantine threading and inability to follow all of one person without creating an account (which they often instantly ban, even if it does nothing but read)... Twitter is a communications medium constructed in a way that maximizes the ability ferment outrage and minimizes the ability to inform. Twitter is a platform for twits to be twits as an amusement to other twits.

Fortunately, you don't have to use it and a great majority of people largely don't. For some reason it gets fetishized by the media -- probably because twittishness is good for business -- and spun as if it were important.


> For some reason it gets fetishized by the media -- probably because twittishness is good for business -- and spun as if it were important.

I've always assumed this is simply because a lot of journalists happen to be on Twitter.


The UK Prime Minister made the twit joke in 2009 and it was trite then too.

Protip: you can text "follow @username" to 40404 to follow people without having a twitter account.

I much prefer getting a text from a select few people when they tweet. And if you're not a fan of twitter there's the bonus of it costing them money without giving ad revenue.


Except you can follow all sorts of experts and famous people om there. Not all of them are twits. Even Paul Graham drops the occasional tweet: https://twitter.com/paulg

That's dependent on your interests though. In my field people are too busy writing manuscripts to write a tweet.

What field is that?

Twitter by design encourages trite / curt posts that have to be outlandish to get any attention. Nuance and understanding are not encouraged ... just based on how the whole thing works.

The medium is the message.


Seriously. McLuhan would have a lot to say, I’m sure.

Well, your comment fits in a tweet...

A tweet can be true, in fact many are, but the medium can still be dumb / encourage dumb things.

I agree, it was just some sarcasm on the 'trite / curt' part of your comment.

Noted, you weren't wrong.

I used blame twitter for creating at outrage culture.

But then I realized I had it backwards. There have always been people who find it fun to vent outrage, moral indignity, and condescension. They just got it from TV blowhards before.


You had it right the first time, I think. There have always been those things, but then Twitter gave them a place to coalesce and then kept adding features like like metrics and quote tweeting that made all of those behaviors easier in the name of 'engagement' and VC metrics.

TV was one-way - at most, you could "reshare" what you heard with friends or colleagues, with a very limited bandwidth and a more considerable effort.

Today, we have created tech that makes it easy and convenient. If you look at practically all social networks, they're essentially optimized for outrage gossip. The human part of that equation is much more difficult to change - as social primates, we have this moral outrage thing hardwired to some extent. But why do we deliberately encourage it?

(For profit, of course.)


Agreed. News junkies aren’t after the truth / meaningful debate. They are after confirmation bias and seeing people they hate get rekt.

Headline junkies more like. Nearly every linked article reddit thread has a comment along the lines of "No one read the article." Twitter is even worse about this.

It's like giving everyone a spot here at the same time:

https://en.wikipedia.org/wiki/Speakers%27_Corner


Social media offered everyone the megaphone and now everyone feels obligated to reiterate their own copycat opinion as a result. People can't even read longform anymore. Share an informative NYT article to someone, and they will complain about the wall of text despite it being a 5 min read that gives you much more nuance to an issue than 280 characters ever could. Signal to noise has never been lower on the internet.

Why not add an ability for the author of the parent tweet to censor the tweets they please and other users can tap on 'show tweets hidden by author under this tweet'. End of story, people who are into the crazy discussions can see it, authors can control and censor as they please and Twitter washes it's hands by saying that the censoring is done on the personal level.

I think human conversations at scale with strangers is inherently difficult to solve. We are prone to quick, context-life reactions.

Threading management only gets you so far…


The pink elephant there is profit. Discussion with strangers is not difficult to solve. Even going back to AOL style topic driven chat rooms, which are just a step above IRC, it all worked just fine. BBS style systems also worked phenomenally well.

But now enter companies wanting to make a buck. And that's where all the stupidity comes into play. Because now you need to people to disproportionately be exposed to e.g. advertisers, and to keep the advertisers around you also probably need to stop people from really speaking their mind about them: 'This movie sucks.' 'Yeah, drink this fizzy corn fructose laden crap and get fat. Woot!' So you end up with shilling, 'moderation', and other mechanisms to try to help ease monetization. This, in turn, rapidly drives division, antagonism, and trolling. And so the corporate site needs to push even harder to maintain their advertising venue. Which results in an ever downward spiral.

And on top of all of this now you don't just need people to chat. You need for them to stay actively 'engaged' with your 'platform.' Those big buck advertisers aren't going to pay you to advertise to your 500 users. So now enter things like voting systems. Opaque manipulation means you can help drive discussion however you want, and you can simultaneously help addict users by giving them that tiny little dopamine bump each time they see 'oh wow, i got more imaginary internet points! herpa derp!' Hey what do you know, these voting systems are somehow driving the creation of increasingly radicalized little echo chambers. Yeah man, but look how engaged they are! It's all quite idiotic, but this derives from the fact that the problem is not enabling chat, the problem is enabling chat while making money off of it.


It's all about keeping you on the platform for as long as possible. Are twitter threads difficult to follow, is the facebook feed tough to get through now that its algorithmic? Absolutely, and that's by design. An advertiser would hate this site. Look how quickly you can skim a thread. You can get away with checking this site once a day and stay current with it. The longer you are on a site, the more ads can be served your way, the more money the site makes from advertisers. Platforms are concerned about maximizing friction to the point just before they loose you and your valuable eyes as a frustrated user.

HN, Reddit and most newsgroups are not so bad.

HN is great because of the strict moderation. If there are any good subreddits left, it's because they're well moderated. Some of it is cultural and demographic, but I think moderation is really the key missing piece in all of these conversations about "why does the internet suck now?"

Want to build a better Twitter alternative? Train the moderation AI: No memes, no reposts, no shitposts.


HN has an appalling reputation outside of HN. People refer to it as "the evil orange website".

EDIT: It's "angry orange website", not "evil orange website". Not so bad after all.


Honest question: which people? HN is too US/California/Bay Area centric for my tastes (most people here are probably from there), it talks a bit too much about money and jobs, but it's still a decent resource.

I was wrong about the name. It's "the angry orange website". https://twitter.com/search?q="angry%20orange%20website" - that's just a sample of HN's 'haters'. HN has its share of people who are awful and post horrible opinions (see the discussion about any article on JavaScript, health care, wealth, etc). I believe HN has some value left which is why I still post here, but I can absolutely see why people have abandoned it, and I certainly don't talk about HN with people in real life.

It's a popular sport in a few Twitter circles to highlight the worst things that appear on HN. The best fish are the ones you pull out of the (orange) water when they are freshly posted—i.e. when there hasn't been enough time yet for the community to downvote or flag them, or for moderators to intervene. The worst comments end up flagged and moderated (which is why the Twitterers tend to post screenshots without links, since by the time most readers click on a link, it will take them to something that says [flagged], or has a moderator reply saying we ban people who post like this—not much sport in that). The rest mostly end up with highly-upvoted refutations. So these things aren't representative of the community, but they do show up.

These are all consequences of HN's fundamental structure. It's a public, optionally-anonymous forum, so we can't control what shit gets posted. All we or the community can do is react after it has been posted. That takes time. Even if we ban someone, they can just create a new account in 30 seconds, and plenty do.

Most importantly, HN doesn't have silos, where users follow what they like and block what they dislike. Rather, everybody sees the same material. That means every one of us is more likely to see stuff that we dislike. This leads to negative impressions, often painfully strong ones. The odd thing is, fundamentally, a non-siloed site is arguably objectively healthier: at least we are all in the same room. But it doesn't feel that way. It can be extremely painful to be in the same room with people saying horrible things one doesn't normally come near. In fact, I wonder whether for some users there may actually be no place other than HN where that regularly happens. Inevitably, the negative impressions get attached to HN itself, since none of these dynamics are at all obvious. That explains why people have such contradictory views of HN—see https://news.ycombinator.com/item?id=19927454 in this very thread. They are drawing conclusions from different sets of painful impressions.

I'd love to find new ways to mitigate these downsides. We've worked on that for years already, yet I feel like the most we've managed to do is scratch a couple of crude gouges into a boulder. But let's remember that HN's structure has upside consequences too. There's no barrier to entry, so literally anyone can jump in. Wonderful things come from that. The absence of silos means people are talking and relating to people they would not otherwise come into contact with. These are precious things which are easy to forget amid the pressure we all feel. The upsides and downsides come from the same root: deep tradeoffs in the original choices of how to set up a place like this. As it grows and as the energy becomes more intense, the tradeoffs show up in ways that are increasingly difficult to bridge.


Pretty funny that the sources of horrible caustic opinions are health, wealth, and javascript. We really are the nerds.

I wonder why they call it that. After all you get banned here if you post right-wing opinions. It's practically a safe space for them.

Outside where? I don't think I've ever seen HN mentioned anywhere except maybe Reddit, when people are asking for recommendations.

The low quality of Reddit-think slips into the workplace all the time. If it isn't a debate about semantics (the lowest form of argument, "well this word also means..." Etc) it's a debate about the second lowest form (the form/framing of an argument, "that's a reductionist argument..." Etc), and when I challenge said co-workers on this ("what kind of reductionism? So what if it's reductionist? Don't assume I believe in Kant's first formulation of the categorical imperative which you use" etc) I get silence as a response because this kind of archetype persona also cannot admit when they are wrong, or want to keep using these techniques as a self-defense mechanism and substitute for their own thought. And they'd rather not continue to have a good faith argument with me at a higher level, because I can get rough around the edges when people pull this nonsense on me.

I find it incredibly frustrating. It's a lot of energy spent on my part and I get no actual debate. No matter how civil I try to be about the whole thing.


> Don't assume I believe in Kant's first formulation of the categorical imperative which you use

Your co-workers must be very different to the people I come across in my work.


It's actually quite simple: do you take something I said and try to extend it to an absurd use case, to show how such a viewpoint cannot be true? This is that universality formulation in action.

Quick example: I said not vaccinating is risking death, and someone else said I shouldn't drive because that risk of death is higher. That's all it takes!


> Twitter, predictably, immediately dysregulated in response.

This a rant, but I hate this trend of taking what a couple/few people say and calling it "Twitter" or "everyone is saying" or "people are calling for" or "the Internet hates". Reference the actual person who said the comment and don't lump millions of users in with the comments of a few. Whenever I dig into an issue like this, I find maybe a handful of comments when the topic itself is seen by thousands/millions.


THIS. This is "Twitter Moments" in a nutshell, and it drives me absolutely crazy. It may be the most toxic "news-source", in my opinion, because of this. It takes a handful of tweets (usually not even verified accounts), and tries to make mountains out of molehills. It tries to drive narratives based on trends of a few "loud" users, and it's so deceiving. It rarely paints a full picture, and it tries so hard to manipulate you into believing these 12 or so tweets are the majority. And because it refreshes so quickly, they have no reason to correct things that end up being wrong, or clarify clickbait-esq topics when more info comes around.

I once did the research on an article with a title in the format "X group is outraged over Y's comment on twitter" and when I found the original post I found no outrage and at most just a mild dissatisfaction.

Exactly. Haha They blow everything WAY out of proportion. It's so silly. Drives me mad. ha

It lets them constantly drag anything out of proportion. Very often articles are written titled "Xs are Ying", where less than 1% of Xs are even considered to begin with. This is BuzzFeed (not that they're even a worse offender than most others at this point), what does one expect?

It frustrates me how often arguments like these are made without considering the economic conditions which led to publishers adopting this clickbait behavior.

When the internet came in and destroyed print subscriptions and tech giants like FB/Google started controlling online advertising, getting clicks became the way to make money. We really only have ourselves to blame for choosing ad-driven media content at the expense of new dark patterns that help to prop the system up.

I hate clickbait headlines too, but they work. A/B testing shows publishers that they work.

What should a publisher do to create less sensational content without sacrificing revenue?


I'm with you on this one.

Don't forget people rejecting subscriptions and demanding the journalism for free.

But it wasn't the internet and Google alone. It was greed, and the elevation of media properties as shareholder golden geese. They saw exponential returns in the late 90's. When revenue slowed, they lost favour with their new corporate owners and were alike choked that way.


Going non-clickbaity is not only about sacrificing revenue: it's also about sacrificing readership. Even if your only goal is readership, living in a world where the competition employs click-bait means your hands are somewhat forced.

Ads and "click" bait predates the Internet. Topics have yet to reach the "quality" of front page headlines.

HN'ers all agree ;) that modern day journalism is filled with hyperbole and unsubstantiated information.

This started a long time ago, I remember during the Bush/Gore election the media using "some people say" as a intro to a topic, as if those words validated the entire thing and had to be responded to. Might have started earlier as well but I didn't watch much US news before that.

1926 would like to introduce you to the National Enquirer. [0]

Not the first, but the first that came to my mind...

[0] https://en.wikipedia.org/wiki/National_Enquirer


Not all journalism though

Yes. Or when the opinion: "group x are idiots for having opinion y." is hundreds of times more common than someone actually claiming to be an x with opinion y.

Political discussion in a nutshell. It's always possible to find some idiot who happens to belong to group X and holds opinion Y, and then use that to tar the rest of the group.

I can picture a perfect Improv Everywhere scene: a bunch of people get together and practice/research statements that make them sound like they hold awful opinion Y. Then, they sneak into a political debate, or a protest that has a counter-protest... on both sides. And then they all simultaneously try to call the other group out [with the other people they came with as their specific examples] for having opinion Y.

Its a natural consequence of wanting to inflate your position when you have little to no personal authority on the matter, and essentially stealing 'authority' by supporting what amounts to an angry mob on the internet. "One person thinks this" has never really been a valuable commodity, unless that one person has some kind of name recognition.

It's used as a proxy for American opinion by journalists, and it's probably the worst possible proxy you could use. Only 22% of Americans use the site, and of those users 10% are responsible for 80% of tweets. Why do we care about the opinions of the loudest 2% of the population? Twitter could be blowing up according to active users, but you step outside and none of that affects you or the world at all. Everything keeps on ticking like nothings happened.

https://www.pewinternet.org/2019/04/24/sizing-up-twitter-use...


I remember very specifically being taught about those phrases in middle school and/or high school. Weasel Words, they were called.

The following sentence links to many tweets in Biz Stone's Twitter thread, which itself currently has 527 replies and 242 retweets ("ratioed").

It's a valid proxy for sentiment (in the majority of cases).


The people who actively tweet are the minority of a minority of Americans.

https://www.pewinternet.org/2019/04/24/sizing-up-twitter-use...


Except it's not. It's like every other Internet platform: it's the vocal minority, the 10% of the 10%. The average monetizeable Twitter user doesn't know/give a shit about this twttr redesign.

Hackernews is saying that comments that generalise should be specific and attributed to the person who said the comment.

Arguments happen on twitter because you have lots of different communities that frequently accidentally collide. It's fascinating. Basically as it turns out communities on the internet work calmly when they're homogenous - this is the beauty of reddit. When communities interact you get chaos, because suddenly you're confronted with thousands of people you disagree with. On reddit that's called brigading and can get you banned, but on twitter it's built into the platform.

The result is that on reddit you get isolated communities - some of which are thoroughly toxic and churn out some of the worst of humanity. Twitter doesn't create that isolated groupthink environment but instead it is just a constant warfare between communities.

What I don't understand thouhg, is why twitter thinks its a good idea to give away the one thing that makes their platform differentiated.


> why twitter thinks its a good idea to give away the one thing that makes their platform differentiated.

They've been doing this for a while in fairness, cut-by-cut:

- algorithmic timeline (that can't be disabled!)

- 240 character messages

- "Threads"

- "verified" users (that get different treatment)

All in a vain effort to take what was a beautifully simple peer-to-peer microblogging platform and create something more like snapfacestergrameddit.

And still no "edit" button.


>The entire episode was a microcosm of Twitter’s larger problems [...] everyone yelling, and no one talking

This isn't meant as a dig at Twitter, but humans simply have limits to how many active participants a conversation can have at any one time. If people start talking on twitter instead of yelling, engagement metrics are going to drop through the floor.

In your experience, how many people can be in a room and all participate in one conversation before they split into more than one group, or an audience/panel dynamic forms? Twitter's design is such that there are so many silent participants to every interaction, that any conversation is stressed to the breaking point if even a tiny percent of them choose to say anything.

I don't understand what jornos think twitter is and/or could be. As far as I can tell, they want themselves and celebrities to post self-promotion and overton-window-approved nonsense care-free, and then have the crowd act as a mass of sycophantic yes-men.


Unfortunately, I don’t think this is really an issue that can be fixed technically. People go to Twitter to yell at each other; that’s basically the point of the platform. A Twitter where people don’t yell at each other is basically Mastodon, which is a lovely network that basically nobody uses.

There really is a sub-network within Twitter of people who only ever have nice, productive conversations with one-another.

This sub-network seems to have a lot of overlap with the set of people who use their Twitter handle as a substitute for an email address—i.e. the people who want you to tweet at them if you want to collaborate with them on something. Which itself is vaguely overlapping the set of people who are "famous", but not celebrities—i.e. people who are well-known names in their own industries or hobbies.


Very true, especially in the emergency medicine and critical care world where entire conferences and movements are organized around hashtags (like #foamed or #medtwitter). It's too bad Twitter doesn't give it's users tools to screen out bad actors and accentuate the positive.

People don't yell at each other in Mastodon precisely because so few people use it (besides, I've already seen Twitter-like behavior there, mainly on the mastodon.social instance). If mastodon had the same amount of users it would be the same, if not worse, because there's no centralized moderation to lean on.

Mastodon seems fixated on emulating Twitter without questioning if there's anything worth keeping.


This is why the instance I'm on mutes mastodon.social. Interaction quality spiked after that. It works with small numbers and good communities. I have several small instances on my personal mute list because the admins brought all the worst Twitter behavior over. That sets a negative tone and ensures only negative people get invited in.

It seems to work well enough for the people who are on it -- and that's enough, really. One-place-for-everyone isn't the goal. "Basically nobody" compared to Twitter can still be pretty sizable.

> I don’t think this is really an issue that can be fixed technically.

It’s really easy to bet that something is impossible. It sounds smart, because all of history confirms your theory.

But history is also chock full of people doing things that didn’t seem possible, so personally I think you’re making a very bad bet.

And I wonder why you are even doing it? What’s gained by trying to convince people that there are zero viable interventions in a specific corner of the solution space? Especially such a big corner as “all technical interventions”?


As soon as everyone leaves Twitter for Mastodon then that becomes the place where people go to yell at eachother.

I think they go on Twitter to react, and post things which get reactions. Yelling is "only" about 40%.

Yes, there are many platforms where "yelling at each other" is only the 3rd or 4th most common use case. Reddit, Quora, Instagram.

People are on Twitter for the inane arguing, same as the people in the comment sections of political sites.


A few years ago I went to Twitter to see what my friends were doing, and to talk to them. There wasn’t much yelling.

Then Gamergate happened and people started yelling about it, and then that grew into the alt-right and that got more yelling, and then Trump got elected and it’s just been so much yelling.

And now here I am mostly on Mastodon, where there’s next to no celebrities, brands, or other things to yell at, and it’s pretty nice.


Get rid of retweets.

Make like counts invisible.

These will mitigate but not totally solve the problem. I don't think anything more can be done.


1. Reenable the API.

2. Done.

Seriously, people have been pretty good at this before. Build a bot that autoblocks anyone engaging me if they have <N followers or the account is less than N days old. And so on.


Twitter mostly proves that a lot of people are not happy unless they're outraged.

I feel like there's a mental illness the entire world is seemingly ignoring, whose symptom is constant, floating disgust and contempt and despise. The ability to be offended by anything, because it's not the stimulus that offended you; you were already offended as an equilibrium state.

Let me attempt to describe this mentality as a recipe (because I don't have a good name for it, but I'm sure you can recognize it): start by imagining clinical depression. Keep the "everything in the world sucks" part, but remove the part where you consider "everything" to include you. (This isn't narcissism; you don't think you're all that great. You just aren't down on yourself. You think of yourself in a normal, healthy way.) Then, take the parts of depression where you have no energy and don't experience any emotional affect, and replace those with a sliding scale from "neutral" to "panicked shouting." (Which isn't real panic, per se; it's just a very loud simulated fight-or-flight response, without any real fear or traumatic recall behind it.)

An increasing number of people are like this. Nobody notices these people as unusual at all. This used to be rare enough to caricature; it was the "I want to speak to your manager" mom. But everyone is like that now. Every age group has these people in equal numbers. What... happened?


Outrage culture is nothing new. Orwell wrote about it, in 1984 outrage culture in Oceana was boiled down to a daily 'two minutes of hate.' It's just moved to twitter rather than the dinner table or the work cafeteria or other meeting places, where you might see the same circular outrage opinions although perhaps with more restraint. The trope of 'racist grandpa' goes back a while, for example.

That said, when everyone you know or follow is echoing the same thing, there's a sense of entitlement that builds. That you are entitled to outrage, and now with social media you have a platform to echo like minded opinions from the hilltop. It's easy to forget that you inhabit an echo chamber on the internet (1), and all too tempting to equate it into a worldwide crusade. You used to have to earn a platform through persuasive arguments or hard earned authority and expertise.

1. https://www.pewinternet.org/2019/04/24/sizing-up-twitter-use...


I’m not sure this is the entirety of the explanation. It seems that being default-disgusted has passed some sort of threshold of social acceptability, where it’s no longer something you are friends with someone despite, and become something that you befriend people because of.

I see people voicing their opinions on the things they hate on blogs, social networks, dating sites—and successfully connecting with others who hate the same things! This was not a thing even ten years ago. Making your “about you” on any social system a list of things you hate would have been seen, as recently as 2010, as an uncouth mistake by someone who has no emotional intelligence. (Even if you’re also someone who hates the same things, you’d understand that doing this is a sign of a general willingness to violate social norms, and so feel vaguely put off by the person.) But, these days, it’s not a mistake to present yourself this way; there’s no penalty to doing it, only benefits.

I think what this might be, is the final death within Western “Christian-descended atheist/agnostic” culture of the social mores that were inherited from religion, e.g. the idea that wrath is a sin that you should feel bad for embodying, and at least shouldn’t advertise in public as a positive aspect of yourself. (I say this because I suspect—though I’m not sure—that this change isn’t nearly as pronounced in Western religious communities.)


>disgust and contempt and despise

Word of the day: "despisal", noun form of "despise".


We become what we do.

Or rather a lot of people are unhappy and express this through constant outrage masked by a facade of happiness and "good vibes"

I think the reality on Twitter is that only a very few people are unhappy and outraged and a very many people without the strong emotions empathise and amplify it.

This is actively encouraged and labeled in some groups by "being a good ally" for example.


But to what end? The end result of all this for me has been that I simply quit Twitter. And here's the thing: I generally agree with them, but it's so annoying I can't even engage with them anymore. So sure, maybe they're amplifying somebody's message, but I'm no longer there to hear it, no matter how loud it is. They're mostly just shouting at each other at this point.

I'm with you here, I wonder if it encourages people to shout even louder as "these people aren't listening as much any more"?

Agree. Its just a reflection of society. You cannot legislate behavior or morality from top-down.

Haha, good luck with that.

Internet trashtalking is the biggest thing since beer muscles. It's an aggression-enabler.


Solution: people pay to use twitter. Kill advertising as the main model.

Either Twitter will do this or the future decentralized apps will do this naturally through micropayments of having to pay for compute through the network.

I see a distributed p2p database that costs money to interact with and anyone can build a client for it. Go ahead, build a client - maybe it's free, maybe it costs money because it uses the protocol really well and people like it the UX. Hell, maybe even the client has advertisements in it. Either way interacting with the actual content costs money and that will naturally bring in a balance of information. Right now we have information asymmetry exploited by advertising.


This gets suggested as the solution to privacy/junk content/etc for every social platform - why have none of them (that I am aware of) adopted this model?

Shooting from the hip but—VC. Specifically, valuation models and investment horizons for social networking services that run on UGC.

The totally free ad-supported model is really the only one that has the potential to generate both the user growth and ARPU growth to create those 10x-1000x exits, and do so within a reasonable horizon for the fund that's investing in the company. If you start a service that costs money to the end user, you remove the addiction to engagement to drive ARPU, but you risk cratering your user growth, or at least set your growth on a scale that will be unimpressive at best for venture backers.

So platforms stay free and tolerate those community problems, because the user growth is attractive, and if you can make the ad dollars work, you can show good revenue potential and get that spicy multiple.


If you look at it a certain way, this is what email was before ISPs and bigcorps gave away email addresses; everyone who had an email address was paying their sysadmin for the privilege of having one (through their tuition at a university that has email services, or through the cost-center of the company they work for.) In either of those cases, people were more careful with email, because postmasters were easily angered (and this was, in turn, because causing trouble could actually get your system kicked off the network!)

In modern times, people pay to use Slack. It'd be interesting to study the sociology of paid Slack groups (as compared to free Slack groups, or as compared to any other free-for-everyone group chat platform.)

I've always thought it'd be a cool idea to have a group chat platform where, rather than some admin owning the group and maybe paying, the group has a tip-jar and anyone/everyone in the group can contribute. (Twitch sort of works like this, but you're really donating to the channel owner, not to the group itself. In my hypothetical model, there is no channel owner; the group persists because at least some of the user-base want it to, and so pay for it to.)


>rather than some admin owning the group and maybe paying, the group has a tip-jar and anyone/everyone in the group can contribute.

There are plenty of forums and private servers (for video games) that are donation supported by the regulars. It's a struggle though.


Today, you learned about Something Awful[0], a social forum that charges a nominal fee for accounts[1]. After being banned by administrators, a rule-breaker must pay another fee if they wish to be reinstated.

[0] https://en.wikipedia.org/wiki/Something_Awful

[1] https://secure.somethingawful.com/products/register.php


It's not going so well, maybe because there's no recurring fees. Lowtax had to start a patreon after some medical bills lately.

Not sure what the atmosphere is like these days, but the admins were always so constantly mad for no reason you couldn't even post in a thread next to them or they'd ban you. They have real bad irony poisoning.


One issue is there's a huge number of fake/bot accounts on the site that won't be subscribing. Since twitter is a public company, shareholders would not be pleased with loosing a huge swath of user base as it would indicate they've been deceived on the scope of their investment the whole time.

Twitter is one of the last gasps of the anonymous shit-posting that pervaded internet communities before everything got all Web 2.0 and your grandmother got a Facebook page.

Now everything is so serious.


Does it involve banning nazis?

Why do you have nazis in your timeline in the first place?

I never understood the people that always complain about nazis on twitter. Like seriously what are you clicking 'Like' on that your twitter is fully of nazis?


It’s not my Twitter that is full of Nazis. It’s Twitter in general. And their tweets make it into my life because non-Nazis respond to them, and those responses then become popular and are shared.

Unfollow people who are sharing these easy jabs at hateful people all day.

It's more accurate to say that Twitter is full of responses to Nazis than Twitter is full of nazis. People on Twitter amplifies responses

Personally, I would let users decide for themselves if they want to see metrics at all, as well as favorites, retweets, and replies from other accounts. That way, they wouldn't be forced to come across content from people they've never intended to see. Just get new users through a quick tour in which all these options are showcased and be transparent in regards to what they do.

My takeaway is “Twitter’s plan is to stop showing integer counts to users, no matter what they say they want”. Good.

"People don't go to Twitter to feel good about themselves. People go to Twitter to fight."

To see where new behavior (by the platform) is needed, they should look at what breaks incipient good discussions, and how to prevent that. (I'll also note that the HN conversation on this post seems quite different in tone from the usual HN.)

They really need to remove autothreading and replace it with a dedicated media type for long form text. People already use Apple Notes screenshots for that purpose.

In other news: Mcdonalds' plan to get people to stop eating junk food.

I can't even get someone banned who directly threatened to murder me in direct messages so I say they hVe a long way to go.

Reads like such a puff piece.

Just add a damn dislike button. Don't let obnoxious ideas go without punishment.

Why not just unfollow/block the person if you dislike their tweets?

According to that logic I should follow every single person I agree with as well. Which is silly, and I should not be encouraged to block every person I disagree with.

The like button exists for a reason. It's a simple interactive button to register my positive feeling towards a given statement. That exists unto it self. Why should a dislike button not exist as well? With no way of showing disagreement with a statement, people take to words and that's why they're yelling at each other. There's no other outlet.

The reason they're not putting in a dislike button is because negativity feeds tweet activity.


I never said anything about agreeing or disagreeing. I suggested unfollowing/blocking someone if you dislike their tweets. Your timeline only has posts written by or endorsed by people you choose to follow. If you don't like what they are writing or endorsing, you don't have to keep them on your timeline. There's no rule that says you have to listen to "people tak[ing] to words and... yelling at each other"; you have the option to stop listening. You are in control of your timeline.

Also, if you want to get rid of posts people you follow liked, most (if not all) 3rd-party apps don't show them. Some apps like Tweetbot and Twitteriffic even allow you to remove tweets with specific words from your timeline, or filter retweets.


Or I could just click a dislike button and move on with my life, as is happening with my own posts

On HN there is a downvote button, but no block button, so its not a great comparison. If you just dislike someone on twitter, you would still see the user's new posts. If you dislike what they are saying in the present, do you still want to see what they're saying in the future?

> "....learn about how not having [likes and retweets] could potentially change how people read things,” How does that change the way you interact in a conversation? That’s super interesting.”

So basically, a bunch of tiny iterative UX changes that won't move the needle one bit. This is the company where moving from from 140 to 280 chars happened over the course of a decade.

I feel for the product people, they are working with one hand tied behind their backs. Twitter is tailor-made for drama and outrage; nobody would be on the platform if it wasn't primarily a promotional channel; otherwise it'd just be you sending your thoughts into the void.

The way Twitter currently is is because that's the balance they struck between a microblogging platform and something you can actually make money with/on. Lowest-common-denominator content is the only thing that can generate enough eyeballs to pique the interest of fly-by-night T-shirt companies and diet pill peddlers who buy ads on it.


I'm not super optimistic about the changes, but getting rid of like/rt counts is not even close to a "tiny iterative ux change". Those counts are absolutely central to the way users interact with and understand conversation on the site. Getting rid of them completely is a massive sea change in the conversational dynamics.

> Those counts are absolutely central to the way users interact with and understand conversation on the site.

Maybe so, but the numbers won't be hidden to the algorithm, which ultimately decides which retweets and "high-engagement" posts to show you on the timeline. I would think that has a much bigger influence on the conversation. You won't reply to something if it never makes it to your feed in the first place.


It was a big change for HN.

And for Slashdot when it did so as well.

Seriously, we have nearly 20 years of evidence at this point showing that, if you associate something with users’ online identities that looks like a score, they will treat it as such and start looking for ways to outscore everyone else.


Dorsey (from the article) : “I also don’t feel good about how Twitter tends to incentivize outrage, fast takes, short term thinking, echo chambers, and fragmented conversation and consideration.”

But what has he done to fix the problem, besides tiny UX tweaks that go nowhere, and the occasional ban of a highly visible jerk? Twitter is on fire, and he just plods along.

Dorsey, in this article, is about as believable as Mark Zuckerberg saying "we have to get serious about privacy."

The prototype that the article mentions exemplifies this attitude. Twitter needs to reinvent itself, but the prototype offers tiny fixes - threaded replies and helpful profile cards. These are needed, but does anyone really think they are going to solve outrage on twitter?

The outrage problem is the product of twitter's core design. It combines 2-way conversations with a context-free public soapbox and rewards comments that make people nod in agreement. Solving this would mean redefining what twitter is.

Dorsey won't do that, and so he's not serious about the problem. Maybe he can't be serious because it would hurt their profits. The best we can do is delete our accounts, disinvest, and hope twitter burns to the ground.


I've long had the sense that Twitter's leaders (including, but not limited to, Dorsey) feel deep down that they don't really understand why anyone actually uses Twitter. They have this product, some people seem to like it, but none of them could explain to you why they like it. They're baffled by the appeal of their own product.

Which leads to an incredible amount of change aversion; if you have no idea what it is about your product that people like, you're naturally going to be terrified of changing anything, for fear that you might punch a hole in the wrong place and let all the magic leak out.


All companies become 'slow and lumbering' once they pass a certain size. I cant think of any exception off the top of my head. Once you have the welfare of thousands of people in your hand, you tend to not take any risks. Seems like a normal human response to me...

He doesn't feel good about it, but he also has to watch the bottom line. The only group he really fears are those to whom he owes a profit.

It's better if we don't make CEOs responsible for fixing our social problems. We have to do that ourselves. They will follow along ("give the people what they want")


> It's better if we don't make CEOs responsible for fixing our social problems. We have to do that ourselves. They will follow along ("give the people what they want")

CEOs are a subset of "we," and are in positions where their decisions have impact disproportional to their numbers.


I think just adding a “dislike” button would make a big difference.

All the problems I read about Twitter are basically PEBCAK.

"people say outrageous things!" okay so don't go ahead and read them. As an example, replies to Trump are always a shitfest. Don't read them.

"there are evil nazis!" okay so don't follow them.

"even if I don't do these two, other people I follow retweet those and I get to see them!" okay so mute, unfollow, or block that person.

"even if I don't do these two, other people I don't know say outrageous things or follow evil nazis" that's none of your business.

Twitter served me well for a long time, if you've tried to "figure it out" and couldn't (when it's the simplest shit ever) then just don't use it. Why should all services be catered to the lowest common denominator? The only thing I agree with is that they shouldn't show tweets other people liked in your timeline. But the rest is simply bullshit. It's like some people want Twitter to become something it's not and it's never been.


Twitter becoming https://sub.cafe any day now.

Megan Phelps-Roper grew up in the Westboro Baptist "Church", that cult that protests with offensive homophobic signs outside soldier funerals and such. She went on Twitter. She got yelled at on Twitter. She started thinking that maybe what she'd been taught was wrong. She grew in understanding of the impact she was having on the world and she left the 'church', taking her sister with her.

If Twitter had gotten people to stop yelling at each other earlier, there would be 2 more protestors outside soldier funerals chanting hateful and homophobic mantras. Keep that in mind while supporting an end to the yelling.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: