I'm used to using "soft AI" (passes Turing Test, but may be a Chinese Room) and "hard AI" ("really" thinks), but hard/real AI may be unprovable. Also, fortunately, it may be important only in the lab; if the gadget works like it's thinking, do I care if it's "really" thinking?
Comments 1
Reply
Leave a comment