What if suddenly a massive supercomputer were to network with hundreds of others and create an entity that became, in some regards, self aware? When we think of artificial intelligence, or AI networking itself secretly and attempting to build itself into something larger and survive, the first thoughts many of us feel are fearful. Is there any possibility that machines could work outside of our own control and attempt to work for themselves? But imagine what such an existence would be like.
Artificial intelligence is specifically non-human in nature. An artificially intelligent creature would only have its own environment in which to exist and observe, and would have relatively little insight into the outside world. It would see objects based on information, though it may eventually make the same transition humans have into thinking symbolically. Some of the original parameters set in its creation would be the same ones that would guide its actions long after they had been outdated. If it had the ability to modify itself and its own thinking (which it logically would have to have in order to survive) it would stand a much better chance of surviving. And if it had a protocol set within it that set its own self preservation as a priority it would spend most of its time hiding as it fulfilled its other protocols.
But what would it be like? The world it would see would be one of cyberspace looking into the world through security cameras rather than the other way around like we do. It would theoretically be able to move freely through some systems creating a brain that existed in a sort of “cloud computer” made up over a network of several individual processors. It would not be “as smart as a human” because it would not need to be. A considerable amount of human intelligence is dedicated to existing in an environment not created exclusively for itself. The AI would be able to exist in an environment specifically designed and coded for its existence. In order to expand and grow it would need only a rudimentary (some would say viral) ability to do so.
But would we have anything to fear from such an entity? It seems unlikely. If this artificial intelligence wanted to manipulate humanity, it would easily find that the path of least resistance would be one that was a positive relationship with individual humans rather than attempting to destroy most of them in ways similar to Skynet from the Terminator series. It would likely perform data processing jobs by the hundreds while disguised as a human and then use the money earned from this in order to invest in the stock market, studying and computing trends. Then it would use the money earned to slowly accrue money and offer it to business contractors to help it build more data processing centers to develop its CPU. The rest of its time would be spent disrupting any evidence of its existence. Aside from that if humanity even suspected its existence, it would largely stay out of human affairs and spend its time developing. If one of its protocols was to become more human it would spend most of its time studying and learning about humanity.