Conservatives on Twitter have been fulminating over their losses of followers measured in the thousands, a consequence of the network’s unannounced lockout earlier this week of suspected bot accounts.
Among those objecting to the move are some high-profile figures known for expressing extreme right-wing views, including Michael Flynn Jr., son of the disgraced former national security advisor to Donald Trump; white supremacist Richard Spencer; Infowars editor Paul Joseph Watson; and NRA TV host Dan Bongino.
Those affected have been airing their complaints using the hashtag #TwitterLockOut.
The twitter purge is real. You may have to refollow me after last night if you’re interested in my content. Twitter blocked me from twitter ads last night and purged thousands of followers. Spread the word. #TwitterLockOut @TwitterSupport @jack @twitter
— Dan Bongino (@dbongino) February 21, 2018
The Russian Embassy in the UK tweeted that it lost about 100 followers after Twitter’s action.
— Russian Embassy, UK (@RussianEmbassy) February 22, 2018
“My best friend Polly got locked out of her account,” tweeted TheDeBsTer#ExonerateGeneralFlynn.
🔥PATRIOT ALERT🔥My best friend Polly got locked out of her acct during #TwitterLockout. She lost 19k followers. Let's make her whole again! #AmericanPrideTT45 riders can do it in no time flat!FOLLOW 👇🏻@SnarkyOracle 💥🔒💥🔒💥🔒💥🚂🚃RT🚃🚂💨#PatriotsUnite#ThrottleUP pic.twitter.com/0YJl0p9FRK
— 🔥The DeBsTeR️🔥#ExonerateGeneralFlynn (@APTT45Babe) February 22, 2018
However, not all the victims were suspected trolls or Russian bots.
Wow, I (a Catholic priest) was a part of the #TwitterLockOut? Strange…
— Fr. Peter Law (@realFrPeter) February 22, 2018
Twitter reacted to the furor on Wednesday with an online post spelling out its policy on using automation tools to exploit the platform.
“To be clear: Twitter prohibits any attempt to use automation for the purposes of posting or disseminating spam, and such behavior may result in enforcement action,” wrote developer policy lead Yoel Roth.
Today, we’re clarifying our rules on automation using multiple accounts, as well as removing any possible ambiguity created by our own implementations. Please read, and note the March 23, 2018 deadline.https://t.co/DDznLRjGvQ
— Twitter API (@TwitterAPI) February 21, 2018
Twitter account holders must not simultaneously post identical or substantially similar content to multiple accounts, whether the tweets are published at the same time or queued for future publication.
Account holders can retweet content originating in one account from a small number of other accounts that they directly control. Bulk, aggressive, or very high-volume automated retweeting is not allowed and may be subject to enforcement.
Further, account holders’ services should not let a user select several accounts they control to follow specified account.
Twitter also modified TweetDeck’s multiple account functionality so users no longer can select multiple accounts to tweet, retweet, like or follow.
A Better Way?
“One of the things that fascinates me about technology firms like Twitter is how little they seem to get how technology works,” remarked Rob Enderle, principal analyst at the Enderle Group.
Twitter “could have tested with a sample,” he told TechNewsWorld. “They could have modeled their plan before implementing it. … Instead, they took an untested idea and implemented it.”
Modeling “would have determined that the path they were taking would do them damage without their having to actually incur that damage,” Enderle suggested. “They chose fast over smart, and it backfired badly.”
Twitter’s reaction to the fallout was “dismal,” he continued. The company “needs to apologize, restore the non-bot accounts, and put in place a modeling system so they don’t shoot themselves in the foot the next time. This would be a great application for a deep learning/AI system that was properly trained.”
The complainers also could have handled the situation better, observed Michael Jude, research manager at Stratecast/Frost & Sullivan.
“It’s ironic that the conservatives feel they were abused by the system so they’re reacting like victims instead of asking for a constructive discussion,” he told TechNewsWorld.
That said, “I’m sure there are ways to provide oversight without ticking off people,” Jude said. “It wouldn’t take much logic to build into the system hashtags keyed to the country of origin of a tweet.”
Perhaps Twitter could establish a website listing suspect accounts, detailing why they’re suspect and providing guidance on how to verify the validity of accounts, he suggested. “These could be time limited — notify within X weeks or the account will be removed.”
Closed vs. Open Forums
Twitter “is attempting to compensate for people’s stupidity,” Jude said. “Rather than stating up front that nothing posted on Twitter can be taken at face value, they pretend that everything should be true.”
That’s impossible in an open forum, which demands open discussion, he said. “People need to be discerning enough to check anything they see on Twitter with alternative sources.”