If we encourage them to let AI do their thinking for them, there's much less need to worry about wrongthink...