Daniel Kokotajlo comments on Why I’m Optimistic About Near-Term AI Risk