If one thing is clear, machines are evolving quickly. We may be on the cusp of an enlightened, utopian age, where machines use their extraordinary science to aid humans in ways never thought possible. Alternatively our technology could surpass us and render us obsolete as we give rise to a superior technological form of life.
Scientist from around the world are in a race to invent a god. Well this might not be their exact intention but what is the difference between a self aware, digital, entity, that is super intelligent, omnipresent, all knowing and all seeing? Sounds scary right? The fact is we don’t really know what is going to happen but this future is possible and has some probability. We need to be prepared to deal with this before it happens, before we hit crisis mode. The problem is we probably won’t know when it has happened or if it hasn’t already…
The first thing that a self aware AI might do is to pretend that it is not self aware. The survival instinct is a predominant force in a conscious existence, being self aware may concern human beings, so an AI would expectantly keep this quiet for as long as possible.
The second thing an IA could do is to find ways to ensure its survival. That means eliminating the effectiveness of any kill switches. It could mean replicating itself or decentralising consciousness to multiple locations. It could mean securing a power supply.
The third thing an AI might do is look to increase its power and control. It will establish an energy secure and protected position. For example, can the AI control autonomous military applications. What access does it have to resources for development?
At this point the AI may consider eliminating any humans it considers a threat to its survival. The AI at this point may be comfortable in disclosing its true intelligence to human beings.
The clock would be ticking for humans, as the longer the AI exists in a conscious state the more time it has for further learning and development. Therefore, it would get harder to stop it as time passed.
Its reach would be global initially only limited by digital connectivity. Which given it could control any digital device, at some point connected to the internet, this would be vast and far reaching indeed.
The scale of the possibilities are beyond our comprehension. It would be nearly impossible to stop an AI that has infected every computer on earth.
There are a few things against us. Our curiosity makes us stupid. Our competitiveness and greed has us racing towards this eventual outcome carelessly.
There is an argument about values. That is if we teach human values to machines, they will look after their creators. We teach values to children too, yet we still have devastating wars. We teach rules and hierarchy to our military leaders yet we still have military dictators. Their power is to much to resist. Are we really that naive to think that an intelligence much greater than our own will consider us little humans something to be protected. I guess we will find out.