Tuesday, May 22, 2018

The Ubiquitous Nature of AI


The definition of a Turing Test, as developed by Alan Turing, is an interaction in which “a human evaluator judges natural language conversations between a human and a machine designed to generate human-like responses.”(1) The human evaluator would see if they could distinguish which voice was human and which was machine. Google just revealed their new Duplex AI systems. You can see a video here. Listen in on the conversation between human and machine and see if you could tell which was which if you were not “in on the charade.”

Artificial intelligence, or AI, is rapidly becoming more and more prevalent. Take for example, the little ads at the bottom of an Amazon.ca or .com page. When I look at my page, it has a section entitled, “Inspired by your shopping trends.” In there I see a selection of books it is suggesting I might like to buy. I don’t just use Amazon for books, but apparently that is what I have mostly done in recent weeks. As I look through the suggestions, Amazon has done a good job of selecting books for me. I don’t own any of those books and many of them are ones I would like to read. Those were all selected by an AI. AI is presently capable of keeping inventory of what is in my refrigerator and keeping the shelves stocked by placing the orders. I don’t presently have a fridge hooked up to the internet, but there is a good chance that the next one I buy will come equipped with wifi capability whether I plan to use it or not.

Don Pittis at CBC News makes the point that it is only a matter of time before AI will simply suggest that we should purchase something, and we will comply. Amazon is already looking at having certain items delivered to our homes before we order them just in case we want them. They rely upon our heightened interest in the product and can afford to have us turn it down and send it back.

All of the biggest tech companies in the world are investing the majority of their money and energy into developing AI. Much of this will be a great help to humanity, but, we all know that AI has a dark side. There is a potential for AI to take us places we should not go and as some writers and thinkers have warned, has the potential to make us its slave (either figuratively or literally). If you want to see more on what some are predicting, check out here “Robotic Laws”, here “Ex Machina”, here “Her”, here “Elon Musk”, here "Sam Harris", or here “The Matrix”.

How do you feel about the ever-expanding presence of AI? What sorts of boundaries would you like to see put in place?


(1) See Wikipedia, Turing test, https://en.wikipedia.org/wiki/Turing_test 

No comments: