X algorithm treats shared outrage as user preference, Musk says
Elon Musk acknowledged that X's recommendation algorithm cannot distinguish between users sharing posts because they like them versus sharing them out of outrage, leading it to recommend more toxic content to users who discuss such posts with others.
What happened
Elon Musk said on September 1, 2024 that X's algorithm misinterprets user intent when people share posts with others. 'One of the strongest signals is if you forward X posts to friends, it assumes you like that content a lot, because it takes effort to forward,' Musk wrote. 'Unfortunately, if the actual reason you forwarded the content to friends was because you were outraged by it, we are currently not smart enough to realize that.'
Musk explained that users experiencing increasingly toxic feeds may have been sharing examples of toxic posts with friends, which causes the algorithm to assume they prefer that content. He suggested users share posts of puppies instead to counteract toxic feeds.
Reported impact
Some X users reported their feeds had become more toxic. Musk said he had spoken with friends whose feeds had become toxic after they had been sharing examples of toxic posts with each other.
Index analysis
The X recommendation algorithm ranks about 1,500 posts from hundreds of millions of candidate sources based on past interactions including likes, retweets, and replies. The system applies heuristics and filters to personalize recommendations, but cannot distinguish between sharing content because of interest versus sharing it due to outrage.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.