cousin_it comments on Friendly AI and the limits of computational epistemology

[cousin_it]

Your point sounds similar to Wei’s point that solving FAI requires metaphilosophy.