|
Should I step down as head of twitter This poll is closed. |
|||
|---|---|---|---|
| Yes |
|
420 | 4.43% |
| No |
|
69 | 0.73% |
| Goku |
|
9001 | 94.85% |
| Total: | 9490 votes | ||
|
I'm not gonna pretend to know the future and prognosticate on how Twitter will fare under Elon. I have no idea. I do know, however, that moderating a public forum is fraught with difficulties. I'll be interested to see what happens if it's true that Elon is in favor of maximalist free speech. The reality of what happens when you really do pursue that on a large public forum will be... educational?
|
#
¿
Apr 25, 2022 20:13
|
|
|
|
| # ¿ May 14, 2024 14:01 |
|
|
MyronMulch posted:How you do this? There might be a way to disable it in forum settings but if not you can also achieve this with an anti-tracking extension in your browser. For example, "Disconnect" will prevent Twitter embeds from loading under its default settings.
|
|
#
¿
Apr 25, 2022 20:33
|
|
|
https://twitter.com/3PSboyd/status/1589399244800413696 This drama and chaos really does have the vibe of forums drama from yesteryear. It's almost nostalgic for internet users of the right generation.
|
|
#
¿
Nov 7, 2022 06:27
|
|
|
This probably isn't an accurate take. First, the tweet reporting that moderation staff lost their access is more than a week old. Those who weren't laid off probably have access back now. Second, it is likely that the tweet is referring to Twitter corporate full time employees and not to the large number of low paid contract workers in somewhere like Kuala Lumpur who actually crank through tweets and make simple decisions like "is this porn y/n?" If these people are still working then a lot of moderation can still happen, likely even the majority of it. The corporate full timers in Trust & Safety do things like make policy, monitor system metrics like rate of false positives/ false negatives and overall rate of bad content, and set up the workflows for the contract workers. They're well paid tech workers. They also adapt policy and processes to changes in the abuse environment. Without these people, eventually the low paid contractor workflows will break down / become ineffective, but that would take some time. Of course, with Elon rapidly and capriciously changing policy it may be very difficult to maintain the workflows for the low paid contract workers to continue working effectively.
|
|
#
¿
Nov 11, 2022 03:05
|
|
|
Zil posted:
lmao D&D nerds absolutely dunking on the world's richest man. What a sight.
|
|
#
¿
Nov 11, 2022 03:21
|
|
|
DrankSinatra posted:It's important to note that the real Eli Lilly twitter still DOESN'T have an "official" checkmark. Don't be ridiculous, they're not testing in prod. They're making poo poo up as they go, swerving out of control trying to satisfy the capricious whims of their owner, who is a terrible leader and businessman. In prod. PITY BONER posted:Can any goons enlighten me on Dril? Someone somewhere said he was a goon? I don't quite understand why goons regularly post his tweets across the forum. I don't know who Dril was on the forums, but apparently he was (is?) indeed a goon. People post his tweets because they find him funny. He's part of what used to be called "weird Twitter", a Twitter community dedicated to, well, the kind of humor exemplified by Dril. That kind of humor has been (is) popular here and wherever people spend too much time on the internet. Don't overthink it: Goons find Dril funny, his humor meshes well with SA, so goons post his tweets a lot. Chimp_On_Stilts fucked around with this message at 10:53 on Nov 11, 2022 |
|
#
¿
Nov 11, 2022 10:44
|
|
|
Stoatbringer posted:Do they actually update the TOS and make all users re-agree to them, or does he make pronouncements and expect everyone to read his tweets and obey? I saw a claim that the Twitter policy enforcement teams are not informed of his latest brain droppings ahead of time, and instead are literally finding out about them the same way you are: via his public tweets. Then they scramble to try to interpret and enforce the mad king's latest idea. Updating the TOS would imply there's some kind of sane process being followed. There is not.
|
|
#
¿
Nov 11, 2022 10:57
|
|
|
PSA: If you have a Twitter account and want a list of the people you follow (so you can find them on other services or whatever else), you can use https://www.twtdata.com/ to produce a free .csv containing them. I just did it. The list is good, easily imported into my favorite spreadsheet program. It was free since I only had my personal account where I followed like 200 people. If you have a large list I guess the site may ask you to pay, but I am assuming that's for *huge* accounts like corporate brands with many thousands of follows or millions of followers. Now I can try to rebuild my follow list on, like, Mastodon or whatever. And I won't lose it if Twitter experiences extended downtime soon.
|
|
#
¿
Nov 18, 2022 08:21
|
|
|
I said come in! posted:What is John Carmack even doing nowadays? I miss when he worked at ID Software. He is in a kind of advisory role at Meta, but not there full time. He has stated that his goal is to be a sort of "Gentleman Scientist", Victorian era style. What he means by that is he's rich, doesn't need any income to live comfortably day by day, and so wants to spend his time working on interesting problems at his leisure. He has recently spoken about wanting to work on nuclear technology and artificial intelligence. Carmack gave a long interview on a podcast recently where he discusses these things: https://www.youtube.com/watch?v=I845O57ZSy4 The interview is interesting, I enjoyed it. However I did cringe a few times when the host went out of his way to kiss Carmack's rear end, insisting that he's a Super Mega Turbo Ultra Genius. It is my impression from the interview that Carmack is not super in love with the way VR is being managed over at Meta. Also from the interview, Carmack acknowledges that he is often falsely credited with inventing the fast inverse square root algorithm. He used it, but did not invent it. Chimp_On_Stilts fucked around with this message at 23:59 on Nov 18, 2022 |
|
#
¿
Nov 18, 2022 23:56
|
|
|
Dabir posted:Carmack's also like a big time sociopath isn't he? Like he's not going out and actively doing anything wrong or anything right now but only cause he doesn't want to. He's very libertarian and, in the interview I posted earlier, makes it clear that he sees social and ethical issues as secondary to pure technology. Specifically, in the interview he is asked about ethical implications of general artificial intelligence (an AI that can learn general skills and concepts, like a human - contrast with highly specialized AIs which are only capable of functioning in a narrow field like choosing which pictures contain an image of a cat). He says he thinks it's a waste of time to consider ethical and moral implications of such technology until it has advanced to at least the intelligence of a learning disabled toddler (his words, not mine). As a computer toucher myself, I disagree with this. Waiting until potentially abusable / harmful technology is basically up and running before even considering how to mitigate harms has a poor track record.
|
|
#
¿
Nov 19, 2022 00:32
|
|
|
koshmar posted:He's trying to channel Trump with this vague revelation thing Gives me Qanon vibes, too. "The storm is coming" - a moment when all will be revealed, your Enemies will finally suffer, and you will be vindicated at last. There's people who expect to see something like source code which states "if tweet.sentiment = conservative, then shadowban", or an official human written list titled "Conservatives To Suppress" from Twitter Trust & Safety with a signature at the bottom which states "personally approved by Hillary Clinton". The reality, of course, is that content moderation is messy and difficult and its grey area corner cases all the way down. The actual files would be more like "Here's a chain of two hundred emails and seven documents, with comments, where we deliberated exactly how to define Hate Speech for six months and, in the end, some anti-trans stuff fell under this policy so a few accounts got whacked for repeatedly deadnaming people with intent to antagonize". "Even then, two of the suspensions required director level approval because the accounts were prominent." Chimp_On_Stilts fucked around with this message at 22:41 on Nov 28, 2022 |
|
#
¿
Nov 28, 2022 22:38
|
|
|
roffles posted:lol elon paid 44bn to build his own total perspective vortex and its full of nazis This is actually a really good analogy because he looked inside and learned he was not the most important man in the universe, so it destroyed his mind.
|
|
#
¿
Nov 28, 2022 22:43
|
|
|
Lamech posted:Didn't Apple not give a poo poo about losing Fortnight to the "30% is unfair" argument From Apple's perspective, in the Epic Games / Fortnite dispute they're defending their 30% cut on every sale in the app store, which in 2021 meant 30% of $85.1 billion according to this source I spent exactly zero seconds vetting. Even if that source is off, I'm sure it's shitloads of money. If dumping Fortnite on iOS means saving the remaining $XX billion, Apple will happily do it. It's the obviously correct financial choice from Apple's perspective. "Do we forgo a small amount of revenue to save a very large amount of revenue?" - yes, it's a no brainer. Meanwhile Twitter is struggling to make money at all. Financially to the App Store it's probably insignificant. So Apple's decision making here wouldn't be primarily financially motivated, it would be motivated by the reputational risks to Apple. They're choosing between the pain of hosting an app which contains content they don't want associated with the Apple brand vs. the pain of removing a popular app. Both would create backlash. Which choice is worse? That's a VP and maybe even CEO level decision at Apple. Chimp_On_Stilts fucked around with this message at 23:06 on Nov 28, 2022 |
|
#
¿
Nov 28, 2022 23:04
|
|
|
PhazonLink posted:isnt this also the dumbass that said the chatbot they were feeding stories about chatbots becoming true AIs said that the chatbot had become a true AI. No. These are two different incidents. The guy with the biotruths letter is James Damore. The chatbot-is-sentient guy is Blake Lemoine. You can Google their names for info about their respective news cycles.
|
|
#
¿
Dec 4, 2022 10:30
|
|
|
ymgve posted:so, I know DMs on twitter arent encrypted or anything, but does the admin control panel literally have a button on the left to access any user's private messages with a single click? I have never worked at Twitter and I have never seen that control panel before, but I can make some informed guesses. First, that button might take you straight to their DMs in plaintext. But it also might not - it may show you metadata about the DMs, it may show only DMs which have been reported as spam or abuse, etc. Regardless, access will be logged and audited, and any employee who can load the control panel will have agreed (likely in writing) that access is to be used strictly for predefined use cases like anti abuse. Furthermore, just because the button appears on the page doesn't mean the employee can load whatever it leads to. Remember, these are internal tools - they don't need to be as "pretty" as the user facing page. It may show every button by default, but in reality each employee will only have access to the subset of tools necessary for their specific role. This absolutely does not mean employee abuse of the tool is impossible! That's not what I am saying at all. What I am saying is there's probably more insider risk controls in place than is readily apparent from the screenshot. As for the "guano" button, that may simply be the name of a project or tool, or a dumb joke that ended up becoming the tool's common name. Those kind of oddball names are common when a company makes its own tools.
|
|
#
¿
Dec 9, 2022 09:48
|
|
|
Zulily Zoetrope posted:There probably were a lot of backend protections in place at least. Yeah that's a potential problem! Access is hard to audit if the auditors were all laid off.
|
|
#
¿
Dec 9, 2022 09:58
|
|
|
learnincurve posted:Lot of early cars were built by coach builders, or in the case of Rolls Royce, using tooling meant for coach building so that’s why the early cars were all slightly wider than a horse’s arse. These kind of standards which exist now only due to an arbitrary choice in the past or due to technical limitations which no longer exists are interesting to a specific kind of nerd. Many contemporary tech companies making software you use daily still write code in lines of just 80 characters or less because that's how many punches wide a physical IBM punch card was.
|
|
#
¿
Dec 16, 2022 00:37
|
|
|
https://twitter.com/veryimportant/status/1605976420479406080
|
|
#
¿
Dec 22, 2022 21:29
|
|
|
Tei posted:I don't think bots are made that way. Both exist! The kind of spam and abuse that is often referred to as "bot" activity comes from diverse sources. Some of it is fully automated, as you illustrated. But some of it really does come from groups of machines run by human operators. Each method has its own advanges. I'll share a few examples. Automated: - Cheap - Scales well - Can be operated remotely - Does not require the logistical hassle of maintaining a bunch of machines - Very few human operators required (maybe just a single person!) Real machines: - Activity looks more "real", can be harder to detect - Requires less programming and sysadmin skill - There are places where phones and labor are actually quite cheap - Leaves room for human input, creativity Both methods get used because each has its own advantages. For example, if you just want to crank out large numbers of identical posts containing a link to some product where you get a cut of every sale, the automated solution is probably your most cost effective option. But if you want to try to stir up political controversy, having humans who can respond quickly to the latest headlines with bespoke messages, and can argue with other posters, etc., is much better. You'll want a proper troll farm full of humans for that. There's also considerations with respect to having your operation detected and shut down. Every spam method gives off signals which a good Trust & Safety team at places like Twitter (... before it was gutted) will be looking for. Often - but not always - fully automated spam operations are easier to detect. Real groups of humans aren't all running the exact same kind of computer, and logging in from the same IP, and posting identical text within milliseconds of each other - to give a trivial example. Anyway, it's all very interesting if you're the right kind of computer security nerd. Don't underestimate the spammers, they're highly motivated (it can be lucrative) and they're smart / creative. They'll try anything, even mounting a hundred phones on a rack for some person in a third world country to use mechanical turk style.
|
|
#
¿
Dec 23, 2022 11:12
|
|
|
Senor Tron posted:It's so easy to underestimate how much money he has blown on this. You're absolutely correct. The opportunity cost to mankind is huge - so much better could have been done with this money. The list of potential projects is endless. However, it's also important to note that Musk is so unfathomably rich that this has absolutely no impact on his lifestyle, personally. Society bears a large opportunity cost. He bears no meaningful personal cost. This gigantic discrepancy should be cause for massive public outrage and policy change. I know it won't be, but that's what I think about.
|
|
#
¿
Mar 19, 2023 02:59
|
|
|
Devils Affricate posted:This part is just so strange to me. It's not like the premium features are scoped to the device you use to sign up for them. Even if you vastly prefer to use Twitter on mobile, why not take 30 seconds to just sign up for Blue on your laptop/desktop to save some money on a recurring monthly fee? Hell, I'm pretty sure you could just do it through your phone by going to Twitter in your browser instead of using the app. - They may not know - They may not care - They may prefer the lower friction of paying via a system that is already set up to take payments (1 tap vs. digging out your credit card and adding it to Twitter) - They may feel that paying via iOS / Android is more private, since they don't have to share financial info directly with yet another website - They may not want to bother switching devices (for many people, their phone is their primary computer today) But most of all: Never underestimate the power of defaults. Users will just do the default drat near always because it takes the least effort and zero new understanding. Defaults are wildly powerful in UX design.
|
|
#
¿
Mar 29, 2023 06:26
|
|
|
|
| # ¿ May 14, 2024 14:01 |
|
|
There's a huge number of companies / services which have handled loads far larger than this. The first company I would point to on that list is... Twitter. Twitter has been at the forefront of major breaking news, world events, sports events, etc. The service has been a global go-to to get immediate information about massive breaking stories for like a decade. They'd routinely get slammed with traffic when something big happened and for the last several years could handle it. Twitter has always been smaller than, say, YouTube or Facebook, but they could still play ball with the supergiants when big news hit. Obviously, that's no longer true. And honestly, to have lost so much ground is probably worse than to have never had it in the first place. Just think of the degree of mismanagement necessary to have fallen so far backward in less than a year.
|
|
#
¿
May 25, 2023 00:31
|
|