According to the article on Computerworld, artificial intelligence (AI) is a sub-field of computer science. The overall goal of AI is to develop computers that are able to complete tasks that are normally done by people. It does not matter how the AI accomplishes the task so long as it mimics what an intelligent human could accomplish. There are two camps of AI, “strong” and “weak.” The strong group is trying to genuinely simulate human reasoning but there are currently no real world examples of this. The weak group is currently what we have for AI machines that mimic human behavior and respond in normal fashions like Siri. There is a grey area for the group that merges the two ideas and the result is IBM Watson. For me AI is currently a long ways away from human intelligence because computers are able to think in ways of answering questions even if they are abstract, but they cannot synthesize the information and use it to develop and improve themselves which is a distinct characteristic of human intelligence.
In regard to AlphaGo, Deep Blue and Watson, I feel it is a mixed bag on whether they are real AI or just gimmicks. Deep Blue and Watson are definitely tricks/gimmicks because they are not learning or adapting, they either simply search a database for an answer based on key word inputs or evaluate a board with a program that gives value to chess positions. However, AlphaGo is definitely more AI and interesting as a result because Go takes much more intuition and there are no set “rule books” for a program to follow and get the best result. According to the Quantamagazine article, “Instead, by analyzing thousands of prior games and engaging in a lot of self-play, AlphaGo created a policy network through billions of tiny adjustments, each intended to make just a tiny incremental improvement. That, in turn, helped AlphaGo build a valuation system that captures something very similar to a good Go player’s intuition about the value of different board positions.” The way that AlphaGo got better at Go is in a sense the exact same way a human player would, by trying different strategies and slowly developing and learning what works.
I do not think the Turing Test is a valid measure of true AI. The basis of AI is whether they can think on their own. The Turing test does not prove this because it is simply an imitation game and not true thought. Also, Turing himself stated that using the test to decide if machines can think is meaningless and not worth discussion.
Finally, I do think a computer system could become a mind once it has the ability to abstract, synthesize and develop its own thoughts and ideas. Currently, computers have not reached that point so they do not have minds in the sense that humans have minds. I do think of the human brain as a biological computer, it in a way runs calculations telling our body what it needs to survive, it gives commands to our body and is essentially what controls everything within us. For ethical implications, I do think eventually we could have to consider basic rights for computers, if the computer is able to have sentient and individual thought it should be treated just like a member of our society (or else it might turn on us).
programming ethics 