The basic motivation here is that having progress in our capacity to engineer the physical basis for information processing grind to a halt would be a bad thing.

Why? Continuation of Moore's law might make AGI come sooner and thus more likely to be unfriendly.