Comment by pardon_me

First we would run into the spam-filter problem no different to email. Then we have to choose: do we concede to viewing the world through a lens of WhatEverAI, or train it locally on our own thoughts/views on the world, and hope that AI model is never compromised.