> we saw people with Fleets talk more on Twitter. Those new to Twitter found Fleets to be an easier way to share what's on their mind. Because they disappear from view after a day, Fleets helped people feel more comfortable sharing personal and casual thoughts, opinions, and feelings.
It is clear that new user engagement metrics are still the only thing that really matters at a social media company. I recently deleted my Twitter account because it is a completely toxic place now. Until there is a clear metric for that, it will always be terrible.
Anecdotes of course, but the last I pulled up my personal Twitter feed 1 in every 5 tweets was Promoted (and 2 of three of those were toxic engagement bait). 2 in every 5 comments were "so-and-so liked this engagement bait" from an algorithm somewhere, leaving just 6 in 15, before I got angry and bailed, "curate-able" tweets from actual people I followed.
Admittedly I switched to Mastodon months back and some weeds are to be expected as I have intentionally avoided Twitter, but it was pretty well curated when I was a heavy Twitter user, and I absolutely can blame Twitter directly for the low value of Promoted content and algorithm generated content I never directly followed and cannot directly curate.
I found that over the past year, people who didn't talk about politics or who had well thought out posts have changed into hot-takes machines or constantly ridiculing bad replies. I have purged follows several times over the past few years, but I find that the same people that I learned new things from a few months ago changed and got more toxic over time. I recently realized that I would read Twitter for a half hour and at the end couldn't think of a single useful thing I learned but my brain felt more chaotic and stressed. Every once in a while I find a new person to follow and learn a new perspective, but if you step back and ask how often that really happens, it is becoming more rare. It has just gotten too hard to filter the noise myself. Twitter inherently sends you superficial, inflammatory ideas faster than I can figure out if they are BS or not. People who I thought were reliable a few months ago change and become unreliable as they get recruited into some tribal war in their own information bubble.
Sorry if this question is off base (I haven't been on Twitter in a couple of years now), but I thought Twitter had gotten rid of the reverse chronological timeline and instead were surfacing tweets their algorithm thought you'd like. Is that not the case?
In the old days, we wrote blogs, and got replies days, weeks, months, even years later. The feedback cycle consisted of your readers, who you also read, writing replies... a modern version of letter writing, but in public. It worked well, in fact it still works well, but it's not twitchy enough for modern tastes.
The downside of the modern system is that the attention economics encourage hot takes, and the advertising economics encourage pushing only content likely to promote engagement, at the cost of reverse chronological order, and a full feed of everything you thought you signed up to read.
Because people are encouraged to act rashly, history in those types of environment is a mine for seeking out poorly thought out statements. It's built into the system.
The technical fix offered here is to erase history, which makes everyone effectively anonymous, as they can say what the want, causing real world effects, and the evidence disappears after X amount of time. Do we really want to make twitter and facebook even more like Lord of the Flies?
There are other ways to deal with this... limiting the audience of one's words to only the friends you have at the current time, and making that stick for all time, is the way to go, in my opinion. If someone isn't in your audience at the time you express a thought, they never will be. It's fairly easy to implement this with the addition of starting and ending times for access.... if I add Bob on November 17, 2020... the default might be to give access starting a Month earlier... and never ending... until it is revoked.
Here at HN, wiser people than me set up a system that seems remarkable resistant to "first post" pressure, and more rewarding of reflective contemplation that will stand the test of time.
Twitter is going to see this new "feature" heavily used to game a lot of things, once people catch on.
I’m guessing the same reason as browser extensions get sold - remote code execution on a bunch of people’s computers for something they wouldn’t volunteer for. Displaying ads, exploiting data leaks to plugins that might exist in VSCode, waiting for a zero day in electron to drop without a patch, something miserable.
> But some of you tell us that Tweeting is uncomfortable because it feels so public, so permanent, and like there's so much pressure to rack up Retweets and Likes. That’s why, unfortunately, there are so many Tweets left in drafts!
really? I thought the whole point of tweeting was that it was easy and no pressure?
I think we lost that when public figures started having to answer for their old tweets. Interestingly, that only happened because enough people bought into the original idea for it to become a permanent thing.
I actually don't use Twitter much because of this. I'll lurk for sure, but I'm more active on Snapchat where I know all my thoughts and photos will be gone. You ever see those folks get fired because of what they said years ago? Pass.
> You ever see those folks get fired because of what they said years ago?
Yep, and I'm totally fine with people being accountable for their own words and actions.
If the concern is people being fired over something that's not actually that bad because of malicious actors making a mountain out of a molehill, then there's better ways to address that than making systematic changes to the way we communicate to allow people to suffer no consequences from proliferating hate speech.
Lol, I joined Twitter around 2013 because it seemed like everyone in tech was there and it'd be a good way to keep up.
Then I saw cross talk between radical Left Twitter and people I followed in tech. Uh oh. I got into exactly one political argument and realised that with my ADHD impulse control, my career, and my belief that homo sapiens is sexually dimorphic, I should not be on that site of I valued eating.
I deleted the account. Now I read Twitter, but I will not log in. Ever. I also deleted my Facebook account around that time.
Nowadays, I only use my real identity with people I know in private groups on self hosted servers. Even commenting here with a pseudonym gives me pause; every opinionated post is a liability if you get doxxed.
I have strong convictions and I cannot imagine being so concerned about my livelihood that I effectively self-censor.
I guess that's why I've never used my real name.
I do also remember when Twitter wasn't so political. But all that changed when more information became available and people started to realize how serious our issues are. If you have concerns you shouldn't be afraid to speak up because that's how a Democracy works.
As far as I can tell, this is a new tab, that only works on the app, not on the website or on third-party clients. So now my followers can only see see my Fleets if they're using the official app?
The idea is not a terrible one, but the execution just seems to be incredibly confusing and poor. It seems like this is a a completely separate product that doesn't integrate well into the main app.
I think it almost has to be that way, no? If it was available on the website or third party clients, it would be trivially easy for someone to write an app that archives everyone's "ephemeral" tweets and stores them permanently, defeating the entire purpose.
I mean, I foresee that happening anyway, but my guess is that this is an attempt to make it more difficult/less widespread.
But with Facebook and Instagram, the "ephemeral" nature of it seems to be a minor feature, not the main point of it (whereas with Fleets that is the main point), so they may care less if people do scrape the stories and archive them.
A screenshot is not verifiable. You can take a screenshot of my Tweet but that is not proof that I tweeted it, unless it's captured and verified by a screenshotting service that has a solid reputation for integrity.
I completely agree, but screenshots of Tweets are always floating around places like Reddit and even Twitter itself. Not enough people have been knowingly caught out for them to be immediately considered potentially fake yet I guess.
Everything twitter bought in it killed off and everything its tried to do since the first offering has failed, they even managed to make a poor job of rolling out their new Web client.
I admire twitters engineering, they provided lots of excellent open source in Scala, but their product teams... Eesh
It just feels like each rewrite of their web UI is a downgrade, both in terms of UX design and in terms of overall engineering quality. I recently broke the new web app doing nothing but a sequence of actions they clearly didn't anticipate.
I mainly use Tweetdeck on desktop because at least it doesn't suck.
The major downside of these “fleets” is that they will be able to be used to harass others but subsequently hide the evidence of that automatically after 24h.
This seems tone-deaf to me. While the world is screaming about either censorship or misinformation issues with Twitter, they're launching a service to make ephemeral tweets ala Snapchat or Instagram stories. I guess they've got to keep developing something, but it feels like a weird time to launch this.
> While the world is screaming about either censorship or misinformation issues with Twitter
The world isn't screaming either of those things, just some individuals living in a political bubble. Most of Twitters millions of users think Twitter is mostly fine.
Absolutely, they are completely out of touch with the masses who are generally fine with how twitter works, they don't even think about it. There is a small bubble that obsesses over celebrity politics on twitter and cares about who got banned and who should have been, but that is a relatively tiny niche. Twitter is an entertainment website for most people, just check the top 100 on twitter - 98% of it is music, tv, sports, etc
And that is specifically why it's "tone deaf". The misinformation issues around social networks are neither limited to USA nor to people who follow politics, they are far more widespread.
Imagine there's a Fire Department in your city and most people are happy with it as they put out the fires, but it turns out a small subset of the houses are kept getting broken into by firefighters on duty. Wouldn’t people be concerned why the hell is Fire Department not getting act together and looking into the matter? Soon they might be the next ones that get broken into.
I would agree, but the feature itself does seem interesting, I am actually looking forward to trying it - to be honest I do feel kind of a weird social pressure on Twitter given that the tweets are somewhat of a public record; it's why I still prefer IRC for instance.
I wouldn't have made the connection between misinformation problems and ephemeral tweets were it not for your comment.
Just remember that for all intents and purposes, fleets will be just as public.
I.e. I predict politicians, activists, celebrities will stumble over them just the same.
As a thought experiment: do you think if Corbyn had fleeted his comment about that graffito, would it have really protected him? I find that unlikely.
Edit: and after typing this out, I'm pretty sure I actually think that last bit is a good thing, independent of my stance on the substance of the matter. But don't mistake "I can't see it anymore" for "nobody else will remember."
Fleets aren't to protect the Leader of the Opposition -- they're to protect the random insurance adjuster who doesn't Tweet anything except sports because he's afraid of getting fired for it. Fleets might let him open up a bit.
Sorry, in my mind even that is closing your eyes and expect nobody to be able to see you.
I'm pretty sure insurance companies, three letter agencies, and anyone else with a vested interest will make sure to have a copy of as many of these as they can get their hands on. And you won't be able to rely on yours not being in any of these stashes.
Nobody will have any guarantees, but controlling the probabilities lowers the number of people affected. I know that there were multiple women in my college being serially abused by a guy who had unconsensually shared consensually recorded videos of them having sex with him, and the police were not helpful because the laws were not in place at that time. He would keep uploading to a single website with their real names. The takedown of that one website dropped to 0 their constant harassment, even if there likely are some copies of those videos still out there somewhere.
For what it's worth, I don't mean to make a link between fleets and misinformation. Kind of the opposite - it is so totally unrelated that it feels tone deaf
This barely solves anything. All public people get screenshoted and once this rolls out, I imagine someone will find a way to screenshot all tweets.
HN is a huge community and I imagine full of ex-employees of Twitter, what exactly does the organization believe the businesses core focus/value is? Like what is the framework they are growing towards?
I've never worked for Twitter, my guess is their business idea is "Well, Snapchat has it, Instagram/WhatsApp/Facebook have copied it, we should also offer it.".
I returned to browsing Twitter more after the election, and IMO it hasn't changed, it's just a place to get your dopamine hit when you see a tweet you agree with get traction, or for some adrenaline when you read something supremely idiotic. I thought about replying to 1 or 2 tweets, but geez, will anyone read it, will it change anyone's mind?
Would a screenshot of a deleted tweet hold up as evidence, either for an investigative journalist or in court? Given how easy it is to produce fake tweet screenshots, it seems it would be difficult-to-impossible to 100% verify authenticity, unless you had Twitter’s cooperation.
I agree. The memes with the Twitter fact checks are hilarious, but they're also tragic. I have no hope of any meaningful Section 230 reform anytime the way things are going and applauding big media companies for their censorship.
I feel like all the big media platforms need to be required by law to not editorialize on anyone else's content:
There is a real cognitive disconnect in our legal system with how businesses can go after something like youtube-dl for for potentially facilitating infringement while companies just straight copy features from their competitors. Why should Disney have this protection of their IP innovation but Snap doesn’t?
Copyright protects the expression of an idea in a tangible medium. All of these story features are arguably expressions of the same idea, but they're not the same expression, so it's not a copyright issue.
There's an amazing variety of creative products that are based on the same idea; often the better ones aren't the first ones.
Copyright protects more than just the specific expression, they also protect collections of ideas. If I publish a novel about a magical wizard who can control things with his mind, flies around the galaxy in a spaceship shaped like an X, carries a laser sword, and battles his cyborg father who wears all black, I bet I would get a call from Disney's lawyers. That is because my work could be deemed substantially similar to Disney's IP even if it is a different expression.
Disney's lawyers are famously litigious. I'm not clear that they would win a case here. (Although they might have success with this HN comment as evidence)
The Adventures of Willy the Wizard: Livid Land (1987) has a similar high-level of comparison to Harry Potter and the Goblet of Fire (2000), and that case was thrown out.
Except that book didn't have "a similar high-level of comparison to Harry Potter and the Goblet of Fire". To quote the judge: "The contrast between the total concept and feel of the works is so stark that any serious comparison of the two strains credulity"[1]. There is a sort of "know it when I see it" nature that is hard to define, but copyright certainly protects works that are substantially similar even if not straight copies, at least in the US.
"Both Willy [a wizard] and Harry are required to work out the exact nature of the main task of the contest which they both achieve in a bathroom assisted by clues from helpers, in order to discover how to rescue human hostages imprisoned by a community of half-human, half-animal fantasy creatures."
is very comparable to your completely hypothetical example above.
So, if a paragraph of cherry-picked, high-level description are the only identified similarities, there is no case in copyright law.
The only real similarity between those two stories that isn't easily attributed to a trope of the genre is the fact that scene took place in a bathroom. Everything else is unspecific and common. Also there is a difference when one scene shows similarity rather than the overall arc of the story.
Not defending the youtube-dl take down, but you can write a story about a boy living under a closet with an abusing adopted family who finds out he is a wizard and there is nothing anybody with the rights to Harry Potter can do about it. Copyright protects concrete versions, not abstract ideas (very much by design, and very much deliberate).
That isn't true. Copyright does not just protect "concrete versions". For example, here is a famous court case[1] in which one of the criteria was whether a "ordinary reasonable person" would find there was substantial similarity between the two works. Some of the similarities can be as vague as "McDonald's character Mayor McCheese and Sid and Marty Krofft's character H.R. Pufnstuf both are fictional mayors that possess disproportionately large round heads."
A copyright happens automatically and is a much more powerful tool to stop infringement while a patent is a pain to acquire, involves public disclosure of a process you often want to keep secret, and is often defined very narrowly. The two aren't really comparable in the level of protection that they provide for your IP.
I would generally be on the side of weakening copyright over strengthen patents, but the disconnect between the two just seems strange when they theoretically are supposed to serve the exact same purpose.
> Presumably Twitter staff can still access expired Fleets. It's not an issue with Instagram stories.
Twitter already has problems with harassment and tweets that are deleted before the report is reviewed. I would bet that harassment will be a problem for Fleets as well.
Fleets actually make it easier for Twitter to censor more content without being detected. By the time you realize your Fleet was shadowbanned, it would have been gone anyway.
It'll be interesting to see whether they make this stuff available to paying customers via the API. Currently they say they have no plans to do so [0], but what if it takes off?
When you are an API customer you are supposed to honor all the actions that a user takes on their tweets (delete them from your in-house datastore if it's deleted from Twitter, for instance) but it's weakly/not enforced by Twitter.
If they start putting stuff that's inherently ephemeral into the API feed then how will they deal with making customers not just persist it anyway? Or maybe analytics for Fleets will only be provided by Twitter, which would be a bit of a slap to their API customers but at least Twitter can enforce the rules around the content.
What a terrible name: fleets. To me, it means multiple sets of several ships. I can't fathom how this would relate to what the product/functionality actually is.
Edit: I guess you can argue that "fleeting" is similar to "evanescent", but... No, I still think about the ships.
Anecdotal evidence only: it's been in testing here in India for a while, and 90% of the "fleets" I see are from media accounts. Hardly any of the people I follow post fleets.
Yes. It's familiar to the majority because Twitter copied every other platform out there. It's like how every white girl loves Starbucks and Instagram selfies.
This feature should be implemented into hackernews. When users really disagree with someone, they tend to pull up their post history and downvote them on multiple comments to multiply the damage. I would like to have no visible history.
It is clear that new user engagement metrics are still the only thing that really matters at a social media company. I recently deleted my Twitter account because it is a completely toxic place now. Until there is a clear metric for that, it will always be terrible.