[txrx0000]

Allow me to make one important distinction I forgot to do in the previous reply: there is a difference between conflicts that arise from delusions vs conflicts that arise from preferences. The former is a subset of the latter. We can prevent escalation of the former with client-side filtering and ranking, but the latter is a force of nature. Two people that share an identical model of reality may decide to take different actions anyway. Minimizing echo chambers aligns people's reality models, but it doesn't align their preferences. So while people will want to do less ranking and filtering over time, they will still want to set a non-zero base level of ranking and filtering for themselves, and congregate into homogenous groups, not because they want to be delusional, but because they're fundamentally different.