AI craze for replicating natural intelligence is rising. Here is another example of a robot who trying score top marks in Japan most difficult university entrance test.
Recently there was AI program(I think the algorithm was developed by Prof. David McAllester, TTI) installed in IBM machine who won the quiz game called Jeopardy. Prior to that there was IBM Blue gene who defeated Garry Kasparov in chess game. So from all those incident it quite clear that computational algorithm has some extend capability to surpass human ability to think and process the algorithmic problems. So I believe primarily the goal of AI is enhance the human ability in data processing and problem solving that are of algorithmic in nature. So this is a good technological outcome.
However the original problem that was formulated by AI giant Alan Turing is the true challenge that need to be solved, how to know who is intelligence or what we mean by intelligence? So the problem is whether any program can ever pass Turing test?
More over, the human being not just machine which has some capability to solve the problems. We have multiple characteristics, emotion, behaviour, feelings, desire, love, sympathy etc. Now the great question is whether all these things are also fundamentally a computational problems or something more than that. If these problems are also a kind of computational in nature than what is that? or if they are beyond computation then what ways they are different, how to characterize these problems as something non-computational. These are open problems and I am interested to solve these.
Jaynarayan, Bangalore, India