Artificial Intelligence, Can it become a human?

                People are creating to make our life more convenient; people invented cars to move to somewhere quickly, phones to communicate to other people instantly, and typewriters to record something easily. Now, people are trying to invent artificial intelligence (A.I.) that will basically think for them; is this even possible? Can A.I. think like a human?

                There are many fictions that talks about A.I., and A.I. in those fictions can think like a human; is this going to come true? For instance, in the novel Neuromancer, two A.I., Wintermute and Neuromancer, combines themselves and become a cyberspace with its own conscious. It has its own mind and behaves like a human, or like a superior creature than a human. Also, in the movie Matrix, robots can think and even rebel against human because human didn’t accept robot as a living creature.

                However, fictional A.I. is a far more advanced than real A.I. at the moment;  for now, there is no real A.I.; A.I.s in real life are just a program that acts based on algorithm that collects data and updates itself; they collect data from people’s input, but they can’t use that data to infer from it. Also, there is no chatterbot - programs that simulates people’s communication through voice or text - that passed Turing Test, a test to figure out whether it’s human or computer. Furthermore, Turing Test isn’t the best way to figure that whether I’m talking to human or A.I. since Turing Test is from 1950s; considering this, it will take much more time to improve A.I. to think like a human.

                Then how are we going to improve an A.I.? Well, first of all, people need motives; if there aren’t people who are trying to create a perfect A.I., this technology will not be advanced. Fortunately, we have Loebner Prize that is a yearly held competition for chatterbots. It applies the Turing Test, it someone creates a chatterbot that passes the Turing Test, it will give 100,000$, but there is no one who got that prize; it also gives around 2,000$ ~ 3,000$ to chatterbot that is the most similar to human.

                Second, people need to analyze and completely understand about human’s brain. Beside the existence of soul, people behave as their brains order through electric signals. People can’t build something that they don’t know; they need to know about human’s brain first so that they can program it. If human can comprehend how brain works, then programmers may able to apply that knowledge into their program and make an A.I.

                Third, people need to create a body for the A.I. Perfect A.I. can’t be created without having a body because they will not feel. Feeling is very important to act like a human; people learn from what they feel. Let’s say there is a baby near a fireplace; the baby might go near the fire, try to touch the fire and get hurt. Then the baby learns that fire is dangerous, and will not touch anything that feels hot; the baby learned from one’s feeling. If scientist can understand the reactions between human’s brain and nerve system, they can probably create a body that can feel, and they can put A.I. into those bodies so that A.I. can feel.

                There are so many steps to make the perfect A.I.; people don’t know completely about human’s brain and nerve system. Even though people succeed to create the perfect A.I., there will be more problems we will face; how will people treat those perfect A.I.s? If people create A.I. that is smarter than human, how can we control it? Would we accept those A.I.s like one of “us”? We have to think about these questions before we creates the perfect A.I.