This Student’s Side Project Will Help Decide Musk vs. Twitter

It wasn't a normal day for Kaicheng Yang. The day before, a US court published Musk's argument on why he shouldn't have to buy a micro-messaging service. A PhD student at Indiana University discovered that his bot detection software was at the center of a titanic legal battle.

Musk tried to back away from his offer to buy the platform. The social network was accused of misrepresenting the number of fake accounts on the platform by Musk. The number of users that can see ads represents less than 5 percent of the total number ofmonetizable users.

According to legal documents, a free tool that claims it can identify how likely a account is to be a bot has been crucial in helping Team Musk prove that figure is not true. The Musk Parties' preliminary estimates show that the business was not minimally affected by false or fake accounts.

One researcher accused Botometer of "pseudoscience" for making it seem easy to tell the difference between humans and machines. Musk used a tool with a history of making mistakes, as pointed out by the social media site. Earlier this year, the platform reminded the court that Musk was likely to be a bot.

The demand for tools that promise to distinguish bot accounts from humans has led to the rise of Botometer. It will be Musk and the science behind bot detection that will be on trial.

The man who started Botometer did not inherit it. Around eight years ago, the project was started. The responsibility for maintaining and updating the tool fell to the founder, who declined to confirm or deny whether he has been in contact with the Musk team. He says that he does not have a full-time job. He works on the tool when he isn't working on his PhD. He says it's just him and his adviser at the moment. The person doing the coding is me.

A supervised machine learning tool called Botometer is used to separate humans and bots. Botometer differentiates from humans by looking at more than 1,000 details of a single account before giving it a score between zero and five. The higher the score, the less likely it is to be a bot. It means that an account is likely to be a bot. It is more likely to be a person if it is 1.2.