Twitter accidentally suspends its own CEO's account
money.cnn.comThe interview was pretty interesting. Sacca's description of how they make it hard to follow what's going on, even though they know first, sums up why I don't use Twitter. Twitter always feels like I'm walking into the middle of someone else's conversation or inside joke. I rarely have enough context to engage, but people who like to just be loud, or talk to themselves, or don't have a clue can go to town.
Maybe I'm doing it wrong.
I use Twitter a lot, but mostly only as a content consumption platform. I follow people who make things I like and who say things on Twitter that I enjoy reading. The idea of using it as a conversation platform is insane to me. Sometimes I have trouble fitting a single sentence into a tweet.
> fitting ... into a tweet.
I enjoy the constraint and view it as a minor new art form, similar to the constraints of a Haiku. Seeing that '0' when I'm done is generally mildly pleasurable for me.
I almost exclusively follow stand up comedians on Twitter, and the character limit is perfect for a concise one line joke. I will actually avoid any "twitlonger" links or images of text, because that's not what I'm there to see. Likewise, I've unfollowed quite a few people on Instagram when they start regularly posting pictures of text. Maybe I'm just a format purist snob?
Images of text on Instagram is definitely wrong. Get a blog.
There's one comedian I know who deleted his twitter account (I think in part because of harassment he was getting, and also in an attempt to obscure his less savory posts from his non-famous days). Then he took to Instagram and started posting images of text with comments disabled. At that point, why even bother? Does the number next to the heart button mean that much to you?
Maybe to preven plagiarism? If people share the same image around, you can trace it back to the original author with a watermark, etc.
You'd be surprised by how little compunction people have with just copying it (via screenshot) and sharing it anyway.
See also: https://theawl.com/the-triumphant-rise-of-the-shitpic-e25d8e...
I should've been clearer -- he's not even posting jokes or anything creative, just random social commentary mostly.
Especially with Instagram, you can write as long of a comment as you want. Post a picture of some random thing and then write your long-ass comment below it.
Great, but how many people online read, write or enjoy writing Haiku? If writing a Tweet requires as much time and (intellectual) energy as a Haiku, than I'm not surprised that the growth of Twitter is tanking.
I enjoy reading things that have more time and intellectual energy put into them.
This is not true of a lot of twitter ... but the short constraint I think is the only thing that really moves the needle in that direction on there.
Taking it the extreme, one needs Reply All's "Yes, Yes, No" segment (where the podcast hosts try to break down inside-joke tweets) in order to understand what I'd assume is the cleverest content.
News search & discovery is all I use the service for and I'd like to see Twitter focus on that. Also, the narrow, single-file, non-pagnated way of scrolling & searching is not appealing at all to me. I'd prefer a more horizontal and populated set of tweets with a simple X in the corner of each one in order to teach Twitter what content I'm interested in (think: Youtube's home page while signed in). The way the site looks is obviously in tune with the 140 char limit but I've never much liked that either. In short, I'm not an ideal user.
Back and forth discussions are not the primary use of Twitter. I had a hard time with that for years too before I finally "figured it out." You just need to follow a lot of people who share your interests and then your timeline will become more interesting.
It does... until everyone stops talking about those interesting things, and starts ranting non-stop about politics. I may even share their political views... but I'd rather be able to mute that garbage and only read about the stuff I have interest in.
Yeah, I follow mostly nerds and comedians on twitter. Up until this election season, most of my feed was nerds talking about things they built and comedians telling jokes.
Now my feed is literally 60% nerds talking about politics and comedians talking about politics.
I'm also something like this. Luckily, Americans are a minority in my feed, so while the previously topical tweets of those who I unfollowed will be missed, it was not a big change for me.
I've spent a lot of time curating who I follow, and I generally enjoy the people I have on there. But Tosh and the Mitch Hedberg bot don't put out enough to make up for the on-slaught of garbage, ads, and political opinions. I might find it more interesting if I was into Justin Beiber, who's trending ever time I look. Maybe I just need some suggestions on interesting feeds.
Mitch's low twitter output could come from a variety of factors. I'm banking on either a) heavy touring schedule or b) his death in 2005.
> You just need to follow a lot of people
Or a few people :) I follow less than 50, and a fair number of those are inactive.
Silicone Valley companies wield too much power when it comes to these sorts of things. This is just another example of this, if they don't self regulate there will be laws that regulate them.
Google was banning people's email and their entire digital life because of buying multiple pixel phones. This is an unrelated issue but they took it upon themselves to just use their unlimited power to screw people. It's like your electricity company shutting off your heat because you didn't pay for some stove, if they sold both things.
These companies would so themseves a lot of justice by setting up a this party group that manages hear abuses and is impartial.
"Silicone Valley" is actually in Southern California—land of the implants.
“You keep using that word. I do not think it means what you think it means.” Inigo Montoya
Silicon != silicone :D. Silicon is a metalloid commonly used to produce silicon wafers for integrated circuits. (Hence, Silicon Valley :D).
Silicone is a rubbery polymer which contains silicon. :D (Ex, window calking, breast implants, etc.)
> Google was banning people's email and their entire digital life because of buying multiple pixel phones.
I thought they unbanned most of the actual people. a large majority were just bot accounts used to buy up the pixel.
They unbanned them because the story took off and it was bad PR for Google. Had the story never taken off, I'm pretty sure those accounts would still be banned.
It's stuff like this that's going to bring more regulation to the Internet. The game industry purposefully created an independent group to regulate their businesses to avoid government regulation that was imminent. I agree that Internet companies need to do the same. No one wants regulation, but it's going to happen one way or the other. Either we do it ourselves on our own terms (like the ESRB) or the government does it on their terms (like HIPAA).
I wonder if Twitter is susceptible to mass report attacks. I strongly suspect many of these systems just automatically suspend someone for too many reports. This makes it convenient if you don't like someone else's politics to just get a group to mass report them and you can silence whomever you want.
This is definitely a problem which some users have been exploiting on twitter.
Happens on youtube all the time. There are raids to take other people's videos down, usually communities between which there is a feud.
Also, people like ASMR artists have seen a lot of takedowns of completely innocuous videos where they crinkle paper or something, just because people flagged it because they thought it's stupid and ridiculous.
It is. I won't go into details, but yes, it can be gamed.
This illustrates the limitations of machine learning and automated processes when applied to the real world. Until GAI is achieved, you really need to couple ML with humans to correct the mistakes made by ML.
Think Iron Man, not Ultron.
> just setting up my twttr...again
> (account suspension was an internal mistake)
twttr is what Twitter used to be known as, and Jack's tweet is in reference to https://twitter.com/jack/status/20
There is a lot of color to the context of this tweet which is very well described in Hatching Twitter [0].
Definitely would recommend, even for entertainment value, but definitely for the SV lore and understanding these references.
[0] https://www.amazon.com/Hatching-Twitter-Story-Friendship-Bet...
throwback to his first tweet: https://twitter.com/jack/status/20
Back then, Twitter was called twttr.
Not just his first tweet, Twitter's first tweet.
Can't be, because it is tweet number 21 as you can see from the ID (assuming the first tweet hat ID 0). Tweet number 1 is just not available.
Bull shit. We all know the first tweet was either 'test' or 'hello, world'.
...followed by a whole lot of "gudsfgjhgfkjhgdasf"
Heh, probably but if you have database access then it can be anything you want.
Compiler Error?
Did they restore his followers from a backup or do you keep them through suspension?
It's a suspension, not a deletion, which implies it's possible to resume the account. It seems logical that the state of the account would be entirely preserved for later restoration.
so why does the article report that only 3.8 out of 3.9 million followers were restored? While overall I would say that is pretty much a rounding error, if we look at it very carefully you have to ask: did literally 100,000 people proactively choose to unsubscribe, from an account that obviously remains important, and moreover during a time when it was inactive? Would his tweet have been so 'spammy' that 1 out of 39 of his every follower decided on that basis alone to unsubscribe?
If the full number of followers were not restored, it raises the question of why not?
I don't know how to get historical "number of followers" information - as of this moment :
the numbers I see are " 20.6K tweets, following: 2,2264, followers: 2.63M, likes: 15.1K, lists: 3"
for good measure and lest someone try to rewrite history here's a screenshot I just took: http://i.imgur.com/JJQKPI9.png
The figure CNN quotes is: "Soon after Dorsey was reinstated, his number of followers was showing up as only about 145 -- a steep drop from the roughly 3.9 million he had previously. The figure later popped back up to around 3.8 million."
So what exactly happened to cause 1.2M twitter followers to decide to unfollow him?
Sure, he's a heavy tweeter - I wouldn't follow him due to the level of spam - I mean who send 20,000 tweets in 7 years - that's 7+ per day at least.
but the people who did choose to follow him - if the CNN quote above is accurate - aren't people I would expect to unfollow him exactly around this time.
So if CNN's figure is accurate -- what happened? And why the drop of 1 million subscribers, or in other words one in three subscribers that he had had?
in distributed systems, counts are often incrementally maintained over time. There are many ways to trigger an unfollow (e.g. account deleted or country banned), and all the edge cases must be explicitly accounted for, or error creeps in, which is probably what happened. 3.8M is probably the more accurate count.
I have no idea why anyone would downvote the current comment. I've edited to be clear I'm being totally open-ended.
>3.8M is probably the more accurate count.
You mean "than 3.9" right? So in an open-ended way, genuinely curious here, would you say the 2.63M he's at now (as I screenshotted for you) an "even more accurate" figure?
just trying to understand the change in figures - I don't have a horse in this race and can certainly accept your reasoning. I don't know much about how these counts work.
I don't know at all, but a number that high may be fuzzy like with reddit's voting system. At some point to save database queries, they may just approximate the number.
that's an interesting perspective - though fuzzing by 46% seems more than I would expect (but may be possible.) I got 46% by doing 3.8/2.6 (you can check because 2.6 * 1.46 == 3.8).
-----------------------------------
EDIT: I didn't change above the line with this edit but - why would anyone downvote this comment too? when all I say is it's an interesting perspective. (There are no comment replies at this time, just a downvote.) Also I reread my comment and I noticed that it was not actually listed as 2.6 - but 2.63 -- so two sig figs.
I actually took screenshots (for myself) with my clock, this is an interesting change:
it's -0.11m change over about 2 hours 21 min. (take 7 seconds). I don't know what part of that is "fuzzing" (like reddit does) what part of that is lack of coherence (from distributed DB), and what part of that is the actual change.
I would think, if anything, it would move up during this time due to the extra attention - maybe HN, CNN and other readers would sub to him after reading this story. at any rate it's certainly a puzzle to me why anyone would downvote this comment. (I don't think it's malicious and there's just one downvote.)
Someone in this thread is down voting just about everything. I wouldn't take it personally.
the big idea here is that distributed systems are complicated and accuracy is often sacrificed to make an intractable problem easier.
Twitter has long had a number of "unfollow bugs", where users sometimes just find themselves no longer following people they definitely intended to follow. shrug.
Perhaps he was suspended as part of a bulk suspension and 1.2m of his followers were also suspended at the same time?
> It's a suspension, not a deletion
I'd doubt that the underlying data is deleted even if you choose to delete your account.
Read this some where on some other forum, is this true?
> So yesterday someone found out that you can get any account banned if reports reach ~1000.
Is this somehow related?