Watson on the other hand is a little different. It has not passed the Touring Test but it is a huge step in the right direction. I found the below excerpt from an article with Ray Kurtzweil talking about Watson:
The most successful example of natural-language processing so far is IBM’s computer Watson, which in 2011 went on the US quiz showJeopardy and won. “And Jeopardy is a pretty broad task. It involves similes and jokes and riddles. For example, it was given “a long tiresome speech delivered by a frothy pie topping” in the rhyme category and quickly responded: “A meringue harangue.” Which is pretty clever: the humans didn’t get it. And what’s not generally appreciated is that Watson’s knowledge was not hand-coded by engineers. Watson got it by reading. Wikipedia – all of it.
Speech is another area where A.I. could fail to ever be like us. the above quote is promising, it appears that the machine thought of something clever to say to an old poem or something? Honestly I have no idea what “a long tiresome speech delivered by a frothy pie topping” in the rhyme category and quickly responded: “A meringue harangue,” means, but the humans in the experiment were impressed. If you read below you will find why that is not so impressive, but lets talk about the idea of speaking to machines and vice versa. This is a big issue as for them to be intelligent they will need to understand what is being said along with the context it is being used. Some humans have trouble with this, so how can we program this ability into a machine. Also, what about a simple joke? Would a machine be able to understand why a joke is funny or would it just take the joke at face value?
There are so many different ways to look at A.I. Some experts sit in awe of Watson while others see the machine as the result of clever programming nothing more nothing less. Here is how Watson answers questions on Jeopardy. During the show Watson was not connected to the internet but it was allowed to read and store all the files or articles that are on Wikipedia. When the questions were read Watson searched through its files looking for the answer by associating words. It then found three answers and choose the one that was most likely the answer. This sounds cool and appears that the machine is “thinking” and using its “knowledge to come up with the correct answers. I read a few articles that state that Watson doesn’t “know” what it is doing nor does it “know” if it found the correct answer. It simply uses what it “thinks” is the best answer. I like the quote below am trying to use the words think and know very carefully. In an interview with Popular Mechanics Douglas Hofstadter a scientist at the University of Indiana explains it better:
Watson is basically a text search algorithm connected to a database just like Google search. It doesn’t understand what it’s reading. In fact, read is the wrong word. It’s not reading anything because it’s not comprehending anything. Watson is finding text without having a clue as to what the text means. In that sense, there’s no intelligence there. It’s clever, it’s impressive, but it’s absolutely vacuous.
Check out the full article here if you want- http://www.popularmechanics.com/technology/engineering/news/why-watson-and-siri-are-not-real-ai-16477207
As I stated above there are so many different views on this subject. In another article I read stated that we are looking at A.I. wrong, we should not be trying to necessarily figure out how to make a machine think. Instead we should be trying to figure out how to use machines to make our lives better. The article also stated that it is going to be near impossible to recreate the mind, how can we recreate something that we do not understand. For all of Kurtzweil’s enthusiasm and optimism I think this is an area where we should be more cautious than curious. He states that we do not need to understand the brain in every detail but instead just try to recreate it down to the molecular level. Exactly how this is done is explained in his book, but I do not agree with blindly trying to do this. I do not think messing with something that we do not entirely understand in this case is a good idea.
I do not share the optimism that Kurtzweil does. As cool as I think it would be to have machines that can think like us I also think it is dangerous. We have all seen countless sifi stories about killer robots that destroy mankind. I do not think this is a likely scenario. However, I do think that if we can create an Artificial Intelligence there will come a time, in the future, when resources have become so limited that we could potentially be competing for them against our creations. I do not think we will win that battle. The more I read about the subject, as much as I would love to be optimistic I do not know if the human brain can or will ever be duplicated in any artificial setting. Colin McGinn put it quite nicely :
“We’re rather like Neanderthals trying to understand astronomy or Shakespeare,” McGinn said. Human brains suffer from a “cognitive gap” in understanding their own consciousness.”