The idea explained in the post, in a way that I don’t know what other reference already explains, is that there is a disconnect between the expected character of a mature goal engine, and the nature of the tools that are being developed under the name “AI safety”.
The idea explained in the post, in a way that I don’t know what other reference already explains, is that there is a disconnect between the expected character of a mature goal engine, and the nature of the tools that are being developed under the name “AI safety”.