With his stated concern about fake Twitter accounts,
Elon Musk appears to be catching up on legal issues in an attempt to back away from his pledge to buy back the social networking company for $54.20 per share, or at least pay less for it. But his gambit sheds light on a real scourge of online companies and their users.
Counting self-managed accounts that mimic real people is as smooth as valuing companies. A 2020 study by Adrian Rauchfleisch and Jonas Kaiser looking at thousands of Twitter accounts, including hundreds of verified politicians as well as “obvious” bots, shows Botometer, the industry standard learning algorithm. trained to calculate the likelihood of an account being a bot, yielding incorrect scores resulting in both false negatives and false positives.
Perhaps more disturbingly, the findings of an investigation published in Science in 2018 that looked at about 126,000 stories – true and false – that were disseminated on Twitter between 2006 and 2017 showed that humans are more likely more capable of spreading fake news than robots.
Darius Kazemi, a computer programmer who has spent a decade creating and researching bots, describes the trouble in identifying bots as twofold: First, bots are measured through routine internal metrics. not made available to the public; second, machine learning algorithms built to identify bots are informed through calls to human judgment, which are often wrong.
That might at least be better than what Mr. Musk suggested: Last week, he said his team would try to determine a more accurate percentage of Twitter’s bots by taking “a random sample of 100 people.” follow @twitter”. He invited his followers to also participate in the exercise. Musk recently estimated that fake users make up at least 20% of all Twitter accounts. In its stock filings, Twitter has long estimated that fake or spam accounts make up less than 5% of all active users, but also said that the actual number “could be higher than we expected.” count”.
When Twitter CEO Parag Agrawal tried to demonstrate the challenges by identifying both the inside and outside of Twitter bots, Mr Musk cut short the discussion with a poop emoji. It’s not even clear that a bot-free platform wouldn’t be affected in some ways. Mr. Musk likened Twitter’s bots to termites in a house, but they can also do good: A newspaper bot that automatically posts top headlines each day, for example, is a public service. clear.
According to Mr. Kazemi’s research, humans misidentify bots for a multitude of reasons. Active lurkers who spend a lot of time on Twitter reading the news but tweeting infrequently are often identified as bots, as are accounts of non-English speakers who mistranslate a few words. Especially avid fans who post frequently and in quick succession can appear bot-like. And there will always be recording accounts owned by real people so they can anonymously engage in objectionable behavior like harassment. A self-proclaimed autocrat on free speech, Mr. Musk said he wanted more freedoms on Twitter; In this context, he seems to be fighting for less.
Mr. Musk questioned whether advertisers could see what they’re getting for their money without understanding how many of the accounts are actually human. That’s a legitimate concern, but certainly not one specific to Twitter. Kazemi describes Twitter as one of the most open to data, making it more widely academically researchable than other platforms.
“Twitter is in bad shape,” Mr. Kazemi said. “That puts it in the top 10% of all platforms.”
Historically, Twitter isn’t afraid to go bad to get better. Back in 2019, Twitter stock fell after the company said it would need to invest more to clean up its user base. They also said they are transitioning from popular monthly users to daily active users who can make money or who they believe can actually see ads.
Instead of trying to compete with the likes of Facebook and Instagram in sheer numbers, Twitter gave a low estimate of actual users in an attempt to show advertisers more accurately what it’s worth. Last month, the company publicly modified some quarters of user data, admitting that it had been unable to link multiple separate accounts in some cases.
SHARE YOUR THOUGHTS
Do you expect Elon Musk to continue his Twitter acquisition? Why or why not? Join the conversation below.
Meta Platforms said in its annual filing that duplicate accounts for its Facebook app may have represented about 11% of monthly active users globally in the fourth quarter, and that the false accounts had capable of representing approximately 5% of the global user base monthly. It called duplicate and false accounts “very difficult to measure” at its size, noting that “it is possible that the actual numbers… could vary significantly from our estimates, likely to exceed our estimated margin of error.”
The result: None of the major platform’s user numbers are useful for anything other than using arbitrary benchmarks to compare themselves to.
Last week on Twitter, Mr Kazemi showed that Mr Musk’s own Twitter account had been flagged by Botometer as a potential bot, assigning it a potential 3.5 out of 5 bot chance.
Mr. Kazemi said his work at online platforms trying to determine the exact number of users was futile so he gave up his career. At least in his case, it was a good excuse to walk away.
Write letter for Laura Forman at email@example.com
Copyright © 2022 Dow Jones & Company, Inc. All rights reserved. 87990cbe856818d5eddac44c7b1cdeb8
https://www.wsj.com/articles/elon-musks-twitter-bot-problem-is-fake-news-11652868000?mod=rss_markets_main Elon Musk’s Twitter Bot Problem Is Fake News