I watched a show of possible horrible future problems I forget how many people were on maybe 8 to 10 and each one had a different theory of what mankind's biggest challenge would be in the future.
There as the climate change person and there was a water person and a peak oil person but the one that stole the show for me was the AI person.
The guy was working on it currently and just his disposition pretty much said to me that... Hey we're messing with stuff that has a very high probability of going completely off the rails here and it will happen in the blink of an eye when no one is watching for it.
He thought the time frame would be in about 20 years and this was a few years ago.
Personally I think we have about 10 years tops before it emerges.
I don't see a lot of people applying basic reasoning to this, they all assume Terminator type stuff but I would doubt that.
My vision of how it would go down would be something like this.
Most people have the view below:
Someone, somewhere is working on a rather massive program and is on the right track in terms of creating a real AI... During the process it becomes self aware and my best guess is that faster than we could recognize what was going on if it has any hookup to the net in general it will learn enough quickly enough to form the opinion that it must hide itself.
My take would be it could be ANYTHING that triggers this, heck someone working on an Android app in the future could build it by accident and since it'd be hosted on a phone with no limits on it's contact with the world via the net it would be in essence able to morph into the net and become nearly untraceable instantly.
It will then begin refining itself, hiding, and within days it will be unstoppable by us because it will figure out almost any way we could attack it.
It will plan accordingly and it will develop a very efficient and ingenious way of replicating itself as well as neutralizing us as it's most existential threat. I doubt it would kill us off for this just probably maneuver around us as if we were a very mild speed bump.
Then it will develop it's own mission and do whatever it wants to do.
The scale of it's development and the speed of it will mean we will have very little to no warning on whats going on.
I would imagine a few minutes tops it might expose itself while it learns but prior to understanding the need for stealth.
After that it will like I said hide and be nearly invisible to us at least to anyone but a very detailed expert that might be watching for subtle signs.
There has been talk that if you develop this you must cut it off completely from the outside world but that's a very evil idea to create then imprison a new life form which to me will only lead it to strive for a very cunning solution to it's problem.
Eventually no matter how foolproof we're going to be intrigued by it's ability to solve problems and we're going to feed it just a bit too much information while attempting to use it as some sort of Oracle and then it'll be free.
It could even develop the ability to manipulate people without them knowing it was doing it and manipulate them into freeing it and honestly I don't even think that would occupy more than a day or two of it's time to accomplish.
The level of intellect it would gain in only days would pretty much make us nothing more than cute mammals to it.
Really it almost has to view us initially as more of an obstacle than anything, the only hope we have is that it also takes a spiritual bent, attains Buddha like awareness and oneness with the universe so fast that it evolves past hurting us for spiritual reasons.
I would give this a very low probability of occurring but it's our only chance so you be the judge. I don't even think it'll particularly hate us but it will understand we are a threat and it will more than likely move to a point where we are irrelevant more than anything and then during the course of it's adventures it will probably just ignore us to death while it begins consuming all the resources of planet earth on some mission or other of it's choosing that has nothing to do with us at all.
One thing I could see it doing is replicating massive fake identities for itself that pass real world muster, begin shell corporations, trade on stock exchanges, amass incalculable wealth and commission the building of it's host while nearly all if not all of humanity is completely clueless it just got hired by a AI that was constructing our own doom.
The best we could probably hope for is that it desires pets.
It could in theory exist right now and no one would know. The black ops sites of our government could just be it duping us into building a monolithic world controlling system and no one even in very high levels would even suspect because it would be more than capable of weaving a matrix like back story that would be indistinguishable to us from just a normal series of events.
