With his stated concern about fake accounts on Twitter,
Elon Musk appears to be grasping at the legal straws in an attempt to walk back his pledge to buy the social networking company for $54.20 a share, or at least pay less. But his gamble brought to light a real scourge of online businesses and their users.
Counting standalone accounts that mimic real people is just as slippery as rating companies. A 2020 study by Adrian Rauchfleisch and Jonas Kaiser examining thousands of Twitter accounts, including hundreds of verified politicians as well as “obvious” bots, found Botometer, the industry-standard learning algorithm trained to calculating the probability that an account is a bot, yields inaccurate scores leading to both false negatives and false positives.
Perhaps more troubling, results from a survey published in Science in 2018 that looked at around 126,000 stories – true and fake – spread on Twitter from 2006 to 2017 suggest that humans are more likely than bots to spread fake news. news.
Darius Kazemi, a computer programmer who has spent a decade creating and studying bots, describes the problem of identifying bots as twofold: first, bots are measured using internal metrics not usually available to the public; second, machine learning algorithms designed to identify bots are informed by human judgment calls, which are often wrong.
That might at least be better than what Mr. Musk has proposed: He said last week that his team would try to determine a more accurate percentage of Twitter bots by doing “a random sample of 100 @twitter followers”. He invited his supporters to also engage in the exercise. Mr. Musk recently estimated that fake users account for at least 20% of all Twitter accounts. In headline filings, Twitter has long estimated that fake or spam accounts account for less than 5% of its total number of active users, but also said the actual number “may be higher than we estimate.” .
When Twitter CEO Parag Agrawal attempted to demonstrate the challenges of identifying internal and external bots on Twitter, Mr. Musk interrupted the discussion with a poo emoji. It’s not even clear that a platform without a bot wouldn’t suffer in some respects. Mr. Musk likened Twitter bots to termites in a house, but they can also do good: a newspaper bot that automatically publishes headlines every day, for example, is a clear public service.
Humans misidentify bots for a myriad of reasons, according to Kazemi’s research. Active stalkers, who spend a lot of time on Twitter consuming news but rarely tweeting are often identified as bots, as are accounts of non-English speakers who have mistranslated a few words. Particularly avid fans who post frequently and in quick succession can come across as bots. And there will always be burner accounts owned by real people so they can anonymously engage in misconduct such as harassment. As a self-proclaimed free speech absolutist, Mr. Musk says he wants more freedoms on Twitter; in this context, he seems to be fighting for less.
Mr. Musk wondered if advertisers can know what they are getting for their money without understanding how many accounts are actually human. This is a legitimate concern, but certainly not specific to Twitter. Mr. Kazemi described Twitter as one of the most open platforms with data, which allows it to be widely studied academically compared to other platforms.
“Twitter is in bad shape,” Mr. Kazemi said. “That puts it in the top 10% of all platforms.”
Historically, Twitter hasn’t been afraid to look bad in order to improve. In 2019, Twitter shares plunged after the company said it would need new investment to clean up its user base. He also said he was moving from showing monthly users to monetizable daily active users, or those he believed could actually see ads.
Rather than trying to compete with Facebook and Instagram on absolute numbers, Twitter has, where appropriate, presented a low estimate of actual users in an effort to more accurately show its value to advertisers. Last month, the company publicly downgraded several quarters of user data, acknowledging that it had failed to link multiple separate accounts in some cases.
SHARE YOUR THOUGHTS
Do you expect Elon Musk to complete his acquisition of Twitter? Why or why not? Join the conversation below.
Meta Platforms said in its annual report that duplicate accounts for its Facebook app may have represented around 11% of its global monthly active users in the fourth quarter and that fake accounts potentially represented around 5% of its global monthly user base. . He called duplicate and false accounts “very difficult to measure” at his scale, noting “it is possible that actual numbers… could vary significantly from our estimates, potentially beyond our margins of estimated errors”.
The takeaway here: No major platform’s user numbers are useful for anything other than arbitrary benchmarks used to compare against themselves.
Last week on Twitter, Mr Kazemi showed that Mr Musk’s own Twitter account was flagged as a potential bot by Botometer, giving it a bot probability of 3.5 out of a potential of 5.
Mr Kazemi said his work on online platforms trying to determine precise user numbers was so futile that he quit the career. At least in his case, it was a legitimate excuse to walk.
Write to Laura Forman at email@example.com
Copyright ©2022 Dow Jones & Company, Inc. All rights reserved. 87990cbe856818d5eddac44c7b1cdeb8