Jack Paul 
Months after Elon Musk pledged to make the X algorithm more transparent, crypto users say the effort hasn’t lived up to expectations.
The pledge came on January 10. Musk promised the algorithm code would be published within seven days and refreshed every four weeks with detailed developer notes. The repository went live on January 17, but since then, there has been only one commit on the official xai-org page. No updated code or developer notes have appeared.
Crypto users had hoped for regular updates to understand how their content is ranked. But the silence has fueled frustration. Many now report weaker reach for crypto-related posts on the platform.
Reach Drops for Crypto Content
Several crypto users say their posts appear less frequently in feeds. Market watcher, Ethan, noted that feeds now show more politics, rage bait, and engagement bait. “Crypto content appears far less often than before,” he said. He added that X is losing the topic-based community structure that once made the platform useful for crypto discussions.
This isn’t the first time X’s algorithm transparency has faced criticism. A similar issue followed the release of the older Twitter/the-algorithm repository in 2023. That project also saw activity slow down after initial interest.
Missing Details in the Code
Ethereum co-founder Vitalik Buterin had questioned the transparency standard before the repository was released. His concern was whether X could provide enough detail for public review.
Now, the published code does show the final score formula. But it doesn’t show the weights attached to each predicted action. That missing detail limits outside analysis. Without those weights, reviewers can’t fully assess how posts are ranked or why some content gains reach.
The Phoenix module README also says its transformer is representative of the internal model, except for scaling optimizations. Critics say this means the public code differs from the deployed system.
Negative Signal Concerns
Crypto users have also raised concerns about negative signals. The model could learn from reports and blocks. Some critics say this could make coordinated bot activity a possible tool to suppress content. However, others point to Farcaster, which publishes forkable protocols instead of limited sample code, as a better example of transparency.
Related: XRP traders are watching a Gaussian retest as an $8 Fibonacci target emerges.