DeepMind released an extensive document outlining its strategy for ensuring the safety of artificial general intelligence. However, some experts are not entirely convinced by the foundations on which it is based.