- Compared to humans, the data that supercomputers have access to was all given to them. It was not learned by them through any form of sensory input. The data that they have was provided to them probably through searching and indexing terabytes of factual data, and storing it in a fault-tolerant database system. To me, that is like competing against a textbook. The human mind is multi-dimensional, from the point of view that all the data and information we learn and store in out minds/brains was as a result of taking in every sound, every visual image, and making sense out of it, and then remembering that association in the brain. To me, that is not the same as searching through a database of organized information that was basically sorted and given to you, and specifically designed to work in a game called Jeopardy. I mean, any human that will ever go against a supercomputer like Watson was processing every visual, auditory, and tactile input at the same time as trying to understand, process, and answer a question in Jeopardy. The processing power of our brains are always divided. The brain is processing bodily functions (like controlling hormones) and input (like deciding to scratch an itch in your sleeve, or wishing the air conditioning wasn’t so cold in the Jeopardy studio). The human mind may be wondering any moment whether the stove was left on, or if your wife is watching you on TV, or something, and still focuses on answering Jeopardy questions. Supercomputers have no idle thought, and no processing of physical functions. All they do is process that data they are designed to process. For me, the day that I admit to a supercomputer being intelligent is the day that the supercomputer can learn and process data the same way I can, which involves visual and audible input of your surroundings.
- Secondly, what I find intelligent are the people who write and develop the custom algorithms that process the speech input and the data that Watson has stored. Did Watson create or develop any of those algorithms for itself? Did it figure out how to learn a language? Or what words mean in different language constructs by itself, or intuitively? Someone, or rather teams of persons, had to figure out how to write the instructions to make Watson do it. So, in actual fact, the intelligent ones are the people who created all these algorithms, and figured out how to get it all to work together. Watson is the culmination of all of their intelligence. Personally, I think that it doesn’t do them justice to call the supercomputer intelligent and not recognize the intelligence of these individuals. I mean, from the time we are born, it takes us years before we fully understand the complexes and nuances of a language, and no one puts instructions into us as to how to do it. We, starting from nothing but being able to take in audible data, learn to understand language and the meaning of the sounds we hear. As of yet, I don’t know of any supercomputer that can do that. For me, I was more interested in who were the people who made Watson possible. They are pretty darn awesome!
Hi Everyone, I have been reading through a few articles about the Watson supercomputer, and how it beat the human competitors at Jeopardy. It really is a great feat of engineering and computer science. But, as I was reading through the various articles, at the back of my mind, I was thinking… “Why is everyone going on about how intelligent this computer is?”. For me, Watson represents a breakthrough in hardware design, computational algorithms, parallel computing, and speed. I was really impressed with it’s ability to process natural language, and interpret speech data correlating it to data that it has stored in its database. However, despite how impressed I was, I still think that computers have a long way to go before we can truly consider them to be intelligent. My critique of supercomputers like Watson, in terms of whether they should be considered as being ‘intelligent’, is based on the following observations: