Nikita Bier (Product lead):
“In October, a new myth started circulating in CT that you need to reply hundreds of times per day to grow an account.
However, each time you post, it uses some of your reach for that day. (We can't show all your posts to all your followers because the average user only views 20-30 posts per day).
So CT ends up wasting all their reach on replying "gm" hundreds of times and when they finally post real content like a project announcement, it only gets shown to 3 people.
CT is dying from suicide, not from the algorithm.”
Half of the people in DegenerateNews post comments that he is lying (followers can not find them), some people agreed with him saying that “Gm” is lower quality, while others successfully gain traffic from consistently reply to hot tweets (currently growing, etc)
Some Chinese Speaking users listed in their tweets: (translated using grok):
1. Semantic value judgment (Grok intervention): The algorithm already has content understanding capabilities. Pure greetings with no informational increment (GM/GN) will be deemed low-value content and downranked.
2. Account trust tiering: The system may layer and label accounts. High-frequency sending of homogeneous content (like GM) will be tagged negatively (similar to Shadowban), and once labeled, even subsequent high-quality content will struggle to enter the recommendation pool.
3. Attention's zero-sum game: Given that users only have a 20-30 post reading budget per day, the algorithm must prioritize content with high retention potential, i.e., content users are most likely to be interested in.
What are your thoughts about this? I am open for discussion.
“In October, a new myth started circulating in CT that you need to reply hundreds of times per day to grow an account.
However, each time you post, it uses some of your reach for that day. (We can't show all your posts to all your followers because the average user only views 20-30 posts per day).
So CT ends up wasting all their reach on replying "gm" hundreds of times and when they finally post real content like a project announcement, it only gets shown to 3 people.
CT is dying from suicide, not from the algorithm.”
Half of the people in DegenerateNews post comments that he is lying (followers can not find them), some people agreed with him saying that “Gm” is lower quality, while others successfully gain traffic from consistently reply to hot tweets (currently growing, etc)
Some Chinese Speaking users listed in their tweets: (translated using grok):
1. Semantic value judgment (Grok intervention): The algorithm already has content understanding capabilities. Pure greetings with no informational increment (GM/GN) will be deemed low-value content and downranked.
2. Account trust tiering: The system may layer and label accounts. High-frequency sending of homogeneous content (like GM) will be tagged negatively (similar to Shadowban), and once labeled, even subsequent high-quality content will struggle to enter the recommendation pool.
3. Attention's zero-sum game: Given that users only have a 20-30 post reading budget per day, the algorithm must prioritize content with high retention potential, i.e., content users are most likely to be interested in.
What are your thoughts about this? I am open for discussion.