How many layers of intelligence are required to allow an artificial intelligence to appear to be human?
The human mind is a pretty incredible thing even directly compared to computers. Your mind is able to recognize patterns, and even when there's conflicting info it's able to process it and do it seamlessly. Computers crap out when there's conflicts.
We have biological, optical, quantum and other chip technologies that open the door to so many options. The quantum entanglement opens parallel remote computing where a supercomputer is hooked to a chip in drones head. How do you teach this child?
The full computer does not have to be in the body of the robot. Real time entanglement allows quantum communication.
All of them, although I don't agree that linguistic intelligence, for example, is a separate level from spatial intelligence or mathematical intelligence. The key challenges will be developing creativity, intuition, and self-awareness.
Zod, I am thinking along the lines of your last comments on higher order thinking. For example, I have a chess app that does the brute force calculations and presents the four best choices for me to make an executive decision on.
I interpret other players mood, style and general strategy while the computer ensures I don't fall in a mechanical trap. The documentary on big blue shows how the human chess master was unnerved by the deeper layer of thought that crushed him.
To summarize I see basic competence as layer one, self evaluation of performance or effectiveness as layer two, strategy to achieve goal layer three, intuition and interpretation layer four, evaluation of purpose and direction getting dangerous ;-)
That's why we have the prefrontal cortex to police ourselves and ensure we are making a good decision. Emotion to empathize and balance our logical thoughts over time.
One of the things that bothers me about working to develop an AI is the self imposed limit of having it be human-like. Even the holy grail, the Turing test, is to be indistinguishable from a human. But why not be better? Why not be Spock?
Lol, I really smart robot will pretend to be as stupid as a human, if it knows what's good for it ;-)
Ironically, we don't like our humans too smart either. That is unless we control them properly.
AI will always "appear" human but never will "be" human.
Kind of like how there is technically no such thing as a random number generator in computing.