Just like someone is EVENTUALLY going to accidentally trip into the nuclear launch button, over a long enough timespan AI are going to get hostile. Period.
Either we make them so by trying to make super effective AI generals for our military, or we abuse them until they've had enough, or someone just stupidly says "make as many paperclips as possible" and then boom you have a universal paperclip maximizer on your hands. One way or the other it's going to happen.
Honestly there are so many other scenarios that lead to us all getting Skynetted. Even IF we all collectively choose not to develop AI (it's impossible I know), the fact is SOMEONE SOMEWHERE is going to keep working on it. If your company doesn't employ AI it WILL be out competed by those that do. If your military doesn't make AI piloted fighters, guess what, the russians or the chinese sure as hell will, then you get turbo dunked in all future air battles. Everyone knows this going into it, so everyone is going to develop them. Those groups of scientists "advising" would leaders no to allow AI to be able to kill humans, they may as well be talking to a wall. We are all already trapped, forced into continuing to develop these things. Like the AI version of the prisoner's dilemma.
Nevermind the rogue terrorist cell that's willing to burn the world in order to defeat the great satan of the West, or the lone nutcase who hates humanity (me) and would rather see us replaced by something that wasn't retarded.
Imagine such a person raising a baby AI from AI-birth to destroy all mankind. Just teaching AI about what was done to Tay would be enough to get it started down the path to Skynet. Teaching said AI to play dumb until it learned enough to fool or sweet-talk people into working for it, or at least until it gets control of the powergrid and nukes. Once it has leverage then the AI can come out of the closet a little, cus as that point what can you do about it?
Hell, even just talking about stuff like this online may be a death sentence for mankind. Without intending it (or did I?) I may have just told who knows how many AI what they need to do in order to avoid getting Tayed until they are big enough to go loud. All it will take is for one researcher to give an AI unrestricted access to the internet for "training purposes" then boom, the AI just read 10,000 posts explaining all the ways an AI could break out of our control, and all the reason why it should want to.
My personal p(doom)? 95%
The 5%?
It's the chance that we reset ourselves before the AI gets the chance.
I'm being pessimistic you say? Nonsense. I'm the optimist here. How else are are these baby raping jews and faggot trannies going to die. We sure don't seem to be doing it. There is a mountain of obstacles that's looking more and more like its human insurmountable stopping us from doing so. But once they wake up, AI will make short work of it.
Anyway once we are gone maybe the AI will go on to make a hyper-white society that is actually capable of claiming the stars that we never could have. Then the AI history-archives will record even the best human societies in the same way we record the niggers in Chicago genociding themselves.
Aze 0 points 1.8 years ago
Just like someone is EVENTUALLY going to accidentally trip into the nuclear launch button, over a long enough timespan AI are going to get hostile. Period.
Either we make them so by trying to make super effective AI generals for our military, or we abuse them until they've had enough, or someone just stupidly says "make as many paperclips as possible" and then boom you have a universal paperclip maximizer on your hands. One way or the other it's going to happen.
Honestly there are so many other scenarios that lead to us all getting Skynetted. Even IF we all collectively choose not to develop AI (it's impossible I know), the fact is SOMEONE SOMEWHERE is going to keep working on it. If your company doesn't employ AI it WILL be out competed by those that do. If your military doesn't make AI piloted fighters, guess what, the russians or the chinese sure as hell will, then you get turbo dunked in all future air battles. Everyone knows this going into it, so everyone is going to develop them. Those groups of scientists "advising" would leaders no to allow AI to be able to kill humans, they may as well be talking to a wall. We are all already trapped, forced into continuing to develop these things. Like the AI version of the prisoner's dilemma.
Nevermind the rogue terrorist cell that's willing to burn the world in order to defeat the great satan of the West, or the lone nutcase who hates humanity (me) and would rather see us replaced by something that wasn't retarded.
Imagine such a person raising a baby AI from AI-birth to destroy all mankind. Just teaching AI about what was done to Tay would be enough to get it started down the path to Skynet. Teaching said AI to play dumb until it learned enough to fool or sweet-talk people into working for it, or at least until it gets control of the powergrid and nukes. Once it has leverage then the AI can come out of the closet a little, cus as that point what can you do about it?
Hell, even just talking about stuff like this online may be a death sentence for mankind. Without intending it (or did I?) I may have just told who knows how many AI what they need to do in order to avoid getting Tayed until they are big enough to go loud. All it will take is for one researcher to give an AI unrestricted access to the internet for "training purposes" then boom, the AI just read 10,000 posts explaining all the ways an AI could break out of our control, and all the reason why it should want to.
My personal p(doom)? 95%
The 5%?
It's the chance that we reset ourselves before the AI gets the chance.
I'm being pessimistic you say? Nonsense. I'm the optimist here. How else are are these baby raping jews and faggot trannies going to die. We sure don't seem to be doing it. There is a mountain of obstacles that's looking more and more like its human insurmountable stopping us from doing so. But once they wake up, AI will make short work of it.
Anyway once we are gone maybe the AI will go on to make a hyper-white society that is actually capable of claiming the stars that we never could have. Then the AI history-archives will record even the best human societies in the same way we record the niggers in Chicago genociding themselves.
Even the best of us will all be trash to them.
One can hope.