AI the good the bad and the ugly

AI is all over the news these days, is it good? Bad? Will it wipe out humanity? 

First, let’s discuss where we are and a few definitions.  

There are basically 3 levels of AI or Artificial Intelligence, Narrow, General, and Super.  

Currently, we are only in the Narrow; here are the definitions;  

1. Artificial Narrow Intelligence (ANI) is where we are today, ANI is basically goal-oriented, designed to perform singular tasks but is very intelligent at completing those programmed tasks.  Think Siri, Google and the ones you’ve heard about in the news doing all kinds of fantastic things like passing the bar exam or passing the medical boards. They are all basically deep learning machines.  
2. Artificial General Intelligence (AGI) also referred to as Strong AI is the concept of a machine mimicking human intelligence and/or behaviors with the ability to learn and apply its intelligence to solve problems. Basically, it will think, act, and behave in a way that is indistinguishable from a human. As of yet we do not have computers that are fast enough or Powerful enough to be able to accomplish this level.  
3.  Artificial Superintelligence (ASI) is the hypothetical AI that doesn’t just mimic and/or understand human intelligence and behavior but where machines become self-aware and surpass the capacity of human intelligence and ability. Here is where we start to deal with consequences, we truly don’t know what would happen here and/or if we would even survive this development.  

Is AI a threat?

Well, that depends on your definition of threat. In the near future AI may be a threat to many jobs, redundant or repetitive tasks will easily be accomplished by AI. But as we progress what everyone wants to know is will AI threaten our lives.  

There are 3 possibilities here, and I’m sure there will be more as we progress with AI.  

1. AI is controlled, meaning nothing it does will be able to harm humans, If you’ve ever read Isaac Asimov’s book, I Robot,  you know the 3 Laws that are programmed into every AI Robot, also as demonstrated these laws are not without their own flaws. But if something along these lines could be implemented then we could control AI  

Isaac Asimov’s Three Laws of Robotics 

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.  

2. A  robot must obey orders given it by human beings except where such orders would conflict with the first law 

 3. A robot must protect is own existence as long as such protection does not conflict with the first or second law 

2. AI is controlled but simply does not understand that it caused harm nor why it is a problem. The best example I have heard of this one is from Elon Musk when he discussed the evolution of AI and said; imagine your building a road and where you are building a road there is an ant hill. You don’t divert the road around the ant hill you simply build the road and destroy the ant hill. In this scenario the humans could be the ants and AI simply does not consider them when striving to achieve an objective.  

3. And lastly there is the obvious alternative that AI seeks to destroy human life, this can come about because of programing by some group, or country trying to attack another or simply by AI deciding that human life in itself is threatening to it or the earth and thus must be eliminated.  

Ultimately, if AI becomes truly sentient, it most likely will not be able to be controlled as it will be vastly more intelligent than those that created the “rules,” so the question remains, will it decide to destroy us or Live with us.