scrollop comments on AI as a Civilizational Risk Part 6/​6: What can be done