Just saw this awesome documentary ***Do you Trust this Computer.***
There was a part in the documentary about ai becoming self aware (due to self learning?)
Watch at 52:20 (the 4 legged spider robot thing).
“Watching the neural network, the robot started to track our faces as we moved around. The spooky thing about this is that we never trained the system to recognize faces, but somehow it learned to do that. There is something else going on here, it’s not just programming.”
How is this possible? How can a program self program itself for certain criteria it was never programmed to do? Are robots and ai like this really becoming self aware? How does ai learn to beat games it’s never been taught at? I know it has something to do with continuous data?
Thought the AI experts here might be able to clarify this issue of ai becoming self aware?