Elon Musk’s previous promise has now come true, as Twitter has released the source code for its recommendation algorithm, with sources indicating that something interesting is already visible in the first scripts.
After poring over the code, several people discovered an intriguing section that revealed the media is particularly tracking stats for users divided into four separate groups, notably Power User, Republican, Democrat, and Elon.
The Discussion After the Twitter’s Release
Following the release of the new algorithm, Twitter put up an audio discussion called Twitter Spaces, where Musk and a couple of developers addressed questions from listeners. Around this period, a couple of people mentioned this particular section of the algorithm during the conversation.
Musk was questioned about the programming that tracked accounts in these four groups, including one specifically for his tweets. He remarked that it was the first time he had seen that. “A lot of silly and embarrassing things are being demonstrated by making the code open source,” the billionaire added.
Although Musk now maintains that he lacks sufficient awareness of the algorithm’s metric tracking component, it is worth remembering that he has previously admitted that Twitter was reviewing his account’s data.
This assertion surfaced last month when he attempted to explain the issue that he called for an emergency Twitter meeting to elevate his tweets in the algorithm. This occurred after President Joseph Biden’s Super Bowl tweet outperformed Musk’s own tweet about the game.
One of the guest speakers on Space expressed worry about categorizing accounts as “democratic” or “republican,” but one of the developers emphasized that this code was only used to collect statistics.
A remark in the algorithm source code, on the other hand, argues that the metric collection is intended to verify that modifications to Twitter do not negatively affect any of these four groups. Furthermore, Twitter has sought to allay this issue by asserting that this metric tracking is used if the algorithm update benefits an entire group of users, but if it negatively impacts the group composed just of Elon Musk, the algorithm change will not be implemented.