you're right about the ethics of rushing into slamming this kind of tech into everywhere possible in all places in all aspects of human life and treating the outcome as an inevitable good, but, like, a meta-heuristic managed vision-native set of <insert ML technique here>? that's my jam.
