Feel like X’s algorithm needs work? Elon Musk agrees.
Musk isn’t a fan of X’s algorithm, and he’s documenting the company’s efforts to improve the technology that decides what you see in the platform’s “For You” feed.
Musk said he is overhauling the recommendation system that decides which user-generated posts and advertisements are shown when you open the app. He also pledged to make that system transparent in the process, with the underlying code getting published on GitHub every four weeks. Musk said any changes to the code will be explained with every post.
“We know the algorithm is dumb and needs massive improvements, but at least you can see us struggle to make it better in real-time and with transparency,” Musk wrote on X. “No other social media companies do this.”
X’s engineering team has also said the new recommendation system is built on the same architecture as Grok, the chatbot developed by Musk’s artificial intelligence company xAI.
While X has not said the two systems share training data or model weights, the comparison signals a shift toward more AI-driven content ranking, moving beyond the rules-based and engagement-heavy systems the app has historically relied on.
Key questions remain about how the system operates in practice. X did not immediately respond to Business Insider’s request for more specifics on the algorithm.
The social media app, formerly known as Twitter, hasn’t always relied on algorithms to serve its users.
From its launch in 2006 through 2015, it presented posts in reverse chronological order, showing users the most recent tweets from accounts they followed.
But, in 2016, after several updates, a “Show me the best tweets first” option became the default for all users.
Over time, the system grew more complex, incorporating engagement signals such as likes, replies, shares, and blocks to predict what content users were most likely to interact with or remove.
Musk has long voiced frustrations with that algorithm.
Before acquiring Twitter for $44 billion in 2022, he repeatedly described the platform’s algorithm as a “black box,” arguing that it introduced political bias and failed to reflect users’ true preferences — particularly for politically conservative users.
His complaints continued after the purchase. In a 2024 post, he said the rules-based algorithms had become too divisive.
“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.”
He recommended that users send videos of puppies to friends to combat the issue.
