Thought: can an AI ever be imperfect in the human sense? Or is it an entirely separate entity, fundamentally different from humans? Is it a matter of scale (more neurons/synapses = more human), or is there something deeper and more profound in the human brain/mind than artificial neurons/synapses (even an infinite number of them) can't replicate?
The thing is, humans make mistakes, for example "HODL" or "pwned", which are due to mistyping, but would never draw a (human) hand with 6 fingers.
AI makes mistakes, like wrong knuckle count, legs coming out of one's neck, etc., that even a 5 year old child would never do.
I just feel there's something fundamental in the above differentiation. Something deep and profound.
Thinking out loud...
The thing is, humans make mistakes, for example "HODL" or "pwned", which are due to mistyping, but would never draw a (human) hand with 6 fingers.
AI makes mistakes, like wrong knuckle count, legs coming out of one's neck, etc., that even a 5 year old child would never do.
I just feel there's something fundamental in the above differentiation. Something deep and profound.
Thinking out loud...
If you ask me, AI (in any form) can hardly reach a human in all his virtues or flaws. Some would say that a machine (software) can never be like a human because it has no soul. Even in SF films and series, we have examples where, by today's terms, very advanced humanoid robots struggle to reach a certain level of humanity. Perhaps the most famous example is the android Data from the Star Trek series and films, who, despite his exceptional intelligence, can never be the human he wants to be.




