It's also typically presented in a pretty oversimplified/watered down way -- like, "what if a computer could talk to you and you couldn't tell that it wasn't a human -- wouldn't that computer be reasonably described as 'intelligent'?"
In Turing's actual paper, he proposes a very specific and adversarial game -- with which I think current AIs would struggle greatly. Not that Turing's arguments that winning would mean AI are all that convincing either (as I recall he knocks down a bunch of strawmen for like 2/3 of the paper) -- but his game itself is deliberately very hard to win.
p
It's also typically presented in a pretty oversimplified/watered down way -- like, "what if a computer could talk to you and you couldn't tell that it wasn't a human -- wouldn't that computer be reasonably described as 'intelligent'?"
In Turing's actual paper, he proposes a very specific and adversarial game -- with which I think current AIs would struggle greatly. Not that Turing's arguments that winning would mean AI are all that convincing either (as I recall he knocks down a bunch of strawmen for like 2/3 of the paper) -- but his game itself is deliberately very hard to win.
More options
Context Copy link