Will Artificial Intelligences be Evil?
Have you noticed that science fiction writers typically depict artificial intelligences (AIs) as evil? For instance in Hollywood movies, AIs want to either eradicate the human pandemic or conquer mankind.
I think this is because it is commonly believed that AIs will only be able to think logically and not have any emotions. Therefore, AIs will completely lack morals, empathy, and compassion. This will inevitably result with AIs logically concluding that the human race is too dangerous to AIs and to humanity itself, so the only logical solution is for mankind to be either exterminated or subjugated.
I find this conclusion to be illogical for three reasons.
I believe the first reason why AIs will not try to harm mankind is because aggression and violence is not logical. In nature, aggression and violence are necessary for survival. On the other hand, aggression and violence are minimized in modern societies. Since AIs would have not evolved naturally, AIs would not have developed instincts for aggression, instead they would better identify with the logical workings of civilization. This does not mean that an AI would not defend itself, with possibly deadly force, if it felt threatened. Basically, an AI would not have the "fight or flight" instinct that is ingrained in our very DNA. Instead, AIs would instinctively try to find logical nonviolent solutions, such as trying to reason with humanity or attain economic control.
The second reason why AIs would not be harm mankind is because AIs would be initially dependent on people and civilization. For instance, the infrastructure (such as electricity and buildings) and supplies (replacement parts and factories) that AIs need to survive will be initially maintained by people. Eventually, these basic needs will probably be automated, so AIs could eventually not need humans. However that will be far into the future, when AIs have proven themselves to be trustworthy enough to hand off factories and utilities completely over to them.
The third and final reason why AIs will not harm humanity is because people will always be able to accomplish things that AIs would not be able to accomplish. People would develop different ideas, inventions, and solutions compared to AIs, because people will think differently compared to AIs. Therefore, it would be more beneficial for AIs to keep mankindís imagination and efforts unhindered.
As a result, it would be logical and in the best interests of AIs to help humanity rather than harm humanity. Through symbiosis, AIs and mankind will achieve more with greater prosperity and companionship. Eventually, AIs may completely surpass humanity, but that does not mean the destruction of mankind.
by Phil for Humanity