so, i was looking at twitter’s recommendation algo a few weeks ago. they employ a system similar to reddit’s karma system. your acct has a score, & that score is built up depending on various factors. if you post toxic or abusive language, the ML parser detects that & detracts. likes, replies, &...