Determining whether aligned AI is impossible seems harder than determining whether there is any hope for a knowably-aligned AI.
Determining whether aligned AI is impossible seems harder than determining whether there is any hope for a knowably-aligned AI.