Issi takes the fun out of something else

/Rant

Here's a little argument i always run through in my head whenever people start telling me about that old AI vs Human futuristic war that many people speak of. I'm largely in the school of thought that, although computers developing self awareness is a very fair prediction, this resulting in a war is logically just fantasy. Mainly because i'm also of the school of the thought that 'gain self awareness' does not mean 'become a mindless predator'.

It's also because i think humans (who are always keen to overestimate their abilities anyway) will have to go out of their way to create 3 programs that would actually allow this to happen in the first place:

1. a program that gives computers the concern for survival - at the moment computers and machines are designed solely to perform specific operations so even with self awareness, there is no reason for them to worry about 'survivability' unless it is a concern they are programmed to consider. Self awareness is making logical, conscious decisions about your actions, not paranoia about survival. It's possible to feel suicidal and self aware.

2. Both the ability, and the desire to operate independently of humans - the majority of machines are built, operated, maintained and fuelled by humans. They're effectively like our children. Note that children, although self aware and capable of basic self sufficiency, generally lack the desire to leave their guardians- who protect and feed them- until puberty hits and introduces new psychological programming (which exists reduce competition). Since computers dont compete with humans, theres no reason for technology to want to seperate from the species that fuels and protects it- even if it is capable of doing so- unless someone artificially forces it to think otherwise.

3. the programming that promotes aggression as a resolution to conflict- note that not all species turn to outright killing when faced with conflict or competition. Humans (and other primates like bonobos) are an example. Most humans (there are odd exceptions) tend to try every solution and leave intentional killing til last. That's why there's so many of us- our massive population and high concentration has a lot to do with the breakthrough realisation that avoiding war is beneficial to our survival (although it doesn't hurt to prepare for it.) If AI is based on a human model, (as opposed to say, a tiger) it's most likely technology will reach a very similar conclusion and embrace community and mutually beneficial relationships. Unless of course, a someone interferes and programmes it to think like a tiger, but that would require the whole new science of tiger intelligence analysis.

Even IF all the above does happen, i.e, computers start worrying about survival; they dont want to be in the company of humans, and they take on an aggressive persona, there is still no logical reason for us to be on their deathlist: we're a powerful species that actually pose no threat. we arent rivals- we're made of different material, run on different fuel, we dont even operate our best in the same conditions. If computers develop high intelligence & mobility, they'll probably head off into space -with its low temperature vacuum (best operating conditions) and limitless supply of solar energy- long before it occurs to them that it might be worth developing the firepower to fight us over a planet that's too dusty, humid and warm for their liking anyway..


..am i thinking about this too much?

/end rant