JP Addison🔸 comments on Can we evaluate the “tool versus agent” AGI prediction?