Re:How We’ll Know an AI Is Conscious
by rbanffy on Monday March 29, 2021 @08:20AM
(#26620482)
My vote is "when it says so". I can't really prove it's not.
[ Parent
| on HN ]
Re:How We’ll Know an AI Is Conscious
by TaupeRanger on Monday March 29, 2021 @08:22AM
(#26620492)
GPT-3 already says so. You can't "prove it" either way. Such is the nature of consciousness. Good luck using that metric.
Who needs GPT-3?
printf("I'm conscious!\n");
There, it says so.`printf` for a literal, seriously?
OK... The compiler will do your job for you, drop the \n and use a puts instead, but, still, it'll be very disappointed with you.
That's how I test the compiler's level of consciousness. Does it just blindly apply some predefined optimisations like an automaton, or does it enter into a meaningful discussion with me to determine my real intent here, and the best way to fulfil it?
Re:How We’ll Know an AI Is Conscious
by beckman466 on Monday March 29, 2021 @08:27AM
(#26620542)
Holy shit, did I just witness the birth of Skynet on HN?
Re:How We’ll Know an AI Is Conscious
by auggierose on Monday March 29, 2021 @09:05AM
(#26620933)
No.
The real Skynet will answer like this.