1 Comment

This is one of the best explanations of this concept, one that a lot of people don't seem to get.

The way that current AI's aren't really programs is a key part of the AI risk argument.

Expand full comment