Many interpretations and suggestions regarding X's public recommendation algorithm have emerged, and overall, the direction is quite good. For most people, adhering to some of these suggestions will be very helpful. However, if you focus on the details and the underlying principles of the algorithm, some interpretations become less accurate. I haven't fully figured it out yet, and I'm still trying to guess the design rationale. Understanding the "why" is more important.