One thing about humans on social media is they also are subject to Reward and they learn from it.
What if we considered Likes, Replies, Reposts, ands Follower Count as criteria in a Reinforcement Learning Rubric?
What if we all are the Grader as well as the Graded