Just saw Billy Marcus rant about X’s algorithm issues—this guy co-founded Dogecoin and is now very dissatisfied with X’s recommendation mechanism. His core point is actually quite simple: the percentage algorithm of a social platform should be very direct; the content from the people you follow should be prioritized for you, and your content should mainly be promoted to your followers.

Musk himself has also admitted there are problems with the algorithm and has been tweaking it. But the question is: how optimized is X’s percentage algorithm at this point? In these past few months, how many times has it been changed? It feels like there’s a new change every few weeks. The crypto community has a lot of objections to this—especially after X restricted InfoFi app permissions last month, when Kaito’s token price crashed directly, sparking a wave of discussion.

Interestingly, X is also rolling out new features, such as smart tags and Starterpacks, to help users discover content more effectively. But can that solve the core problem with the percentage algorithm? It still feels like it comes down to how the data performs later. What do you think—does X’s algorithm logic really have a problem?