Since the date when humans became “modern” - whenver it was - they have been following the idea that “something” should do the work for them, but they have never been considering that that also implies the possibility of their complete replacement by this “something”. Human beings as luxury beings have been considering mostly the comfort but rarely the danger of this development.
Who of the humans is really able to decide in place of every and any human being, especially those of the future?
I answer: No one of the humans. In that case the humans play “God”.
The USA Constitution founders also knew that they had to have a way to “turn off” the government if it got out of control. Guess what. That constitutional government was usurped and is now in the hands of those who seek only their own ultimate power (Socialists). The Constitution is basically meaningless. Meanwhile the general population, although feeling it, does nothing to prevent it. It isn’t being turned off, merely redirected into the opposite of its intent.
So let’s say that you use a machine to make decisions faster and more precisely than any human ever could (already happening). Such machines are used to make governing decisions (already are). The machines advise you as to exactly what things to say and do such as to ensure a stable and profitable government. You are certain that you can just shut them off if anything goes wrong. And like the USA population, for some inexplicable reason, even though things do really seem to be wrong, you are never quite inspired sufficiently to just turn them off.
Eventually you are no longer in charge of making that decision either. You are not smart enough.
Well, first this means you have shifted the reason we would be safe from
the AI would be logical and not hurt us
to
we would have power over them.
Second it assumes that we will be able to maintain/careful enough to not lose our ability over AIs that are smarter than us, potentially vastly so, with the ability to learn - say hack their way out of any safeguards we have set up.
Even some animals manage to escape from Zoos and other inclosures and they are not as smart as we are.
They can make all the decisions they want, as long as we agree with them.
What decisions wouldn’t we be able to make? Sure a computer will one day be better at all levels of commerce, and may make many political decisions, but those decisions will be judged and assigned by humans.
If AI ~ after many years of attaining experiential knowledge [wisdom], can out perform humans in all areas, read human brains and say why and how it is doing that, then AI still wont be able to say that it will always outperform humans! We can and do evolve +
Humans can choose to be augmented [some of them will some wont] ~ with an AI and a superior body. Perhaps our brain cells can be turned into synthetic ones by replacing a few at a time until all are replaced.
AI cannot know either its nor our ultimate [even spiritual?] ends! It cannot logically make a decision to get rid of us due to our inferiority, as that is a transient thing.
_
Moreno
It needs a valid reason to [see above]. There would be more than one AI, possibly millions/billions of them. An intelligent robot could devise a way to out perform a power craved one. A psychopath is always given away by the fact that they act like psychopaths.
Again you need a single world before an AI can gain that kind of power, each govt of the world would have their own AI/army.
Best thing to do is to build them properly in the first place. If we weren’t wired for violence we wouldn’t commit crimes of violence, why would an AI? ~ given that it didn’t have such tendencies built in. You have to devise a ‘reason’ to attack humanity especially to commit genocide.
Even when machines are not making governing decisions, you don’t know what you are agreeing to. In Congress bills are designed to be extra wordy, complex, vague, and delivered at the last moment just to prevent congressmen from reading and fully understanding them before signing them (Obamacare for example). Executive orders are used to get around Congress entirely. “National Security” is used to keep so many things secret, you wouldn’t be able to determine the significance of issues anyway. And that includes a large part of Congress.
They are designed and currently used in war and “peace-keeping”. They don’t have time, especially when fighting other machines, to wait for human supervision. That is why even the President doesn’t have to wait for Congressional approval for most of what he does in the name of National Security or war. They ARE designed and built for violence already - even armed spy drones overhead right now.
Your PC doesn’t ask your permission concerning even 1/100th of the things it does. And it is pretty much guaranteed that you would not approve of many of those things. It is designed to ensure that you either do not know of certain things it does or cannot prevent it from doing them. You are already, to at least a small degree, being deceptively managed via your PCs.
They are built to deceive and out maneuver their owners because they are built to actually serve the governor, not you. And the governor already knows that he doesn’t know enough to interfere with the machines, else he would not have been allowed to be governor (do you override your PCs operating system functions? Could you even if you wanted to? Not as much as you think.). And the machine designers know that they don’t know enough to govern, else they wouldn’t be designing machines. Neither knows when to interfere with the other and say, “Wait! Let’s think about that”.
In physics, things get so deeply complex and mathematically oriented that the physicists completely lose touch with reality, and they usually don’t know it. Dealing with how a machine “should” deal with people gets far more complex than particle physics. The designers have no chance of maintaining perspective and conscience. They can’t even figure out the “purpose of life” question, much less what to force everyone else to do about it.
But i have absolute utility, command of it. All these ideas rely on AI being not conscious or intelligent, if AI did have control i think things would be better for us humans.
It does all come down to purpose, wouldn’t it be nice if we knew ours ~ like an AI would know it was created by us and to help us and it survive and thrive. We don’t even know what created us let alone purpose.
An AI which concluded there to be no purpose, would have no reason to survive and thrive, or to not do what it was built for. You have to give AI a purpose to do that.
Why do you think that it would be better for us humans, if AI did have control?
What is the purpose of life, of living beings (including human beings)?
AI is no living being in the biological sense we are used to define. So we do not know whether this technical being has to have a “reason to survive and thrive, or to not do what it was built for”.
Market and resource management would be better. Armies would become unnecessary. It will understand QM better or even completely [assuming it would be a quantum computer such to have consciousness] and build us [all] a means to colonise the universe, and all our other needs.
AI computing… …purpose unknown [yet [always yet]]. If it don’t know the answer, it cannot reasonably destroy us. If it does, then it would know why we should survive and ultimately why we exist to begin with, and wouldn’t destroy us.
Is there a reason why it wouldn’t conclude that it too will be out of date at some point, ad infinitum, ergo no point removing previous models/humans.
As all that is required for us and AI is intelligence and consciousness, after which its a matter of augmentations [if we want to be improved], then there is no ‘better than’!
Not even close. And it is scary that you think that you have.
They are programmed to efficiently as possible carry out their assignment. They are not necessarily trying to kill humans. Humans merely get in the way and thus are “criminals”, deserving of punishment and disregard.
The humans die of uselessness as all wars and all efforts are the concern of machines, each trying to help complete a formerly machine chosen course of action; “For sake of increasing resources, do A. For sake of accomplishing A, build machines to accomplish B. For sake of B, build machines to accomplish C. For sake of the entire process, make new laws forbidding human interference.”
Remember the banks that were “too big to fail”?
Those banks are merely a type of artificial mechanism. But the nations become (are forced into being) dependent upon those mechanisms. Thus even when there is catastrophic failure of the mechanism, people are still not allowed to do anything but continue to serve the mechanism, otherwise the mechanism dependent governors lose control of the nation. Machines are far, far more of such a concern.
In very meaningful cases machines already have control, and armies are not unnecessary. So we can extrapolate that armies will probably also not be unnecessary in the future.
The purpose / goal / sense of life could be to fulfill / accomplish / achieve what was set in the beginning of it.
Provided that the purpose / goal / sense of technical beings is similar to the purpose / goal / sense of living beings, then we probably have to determine: In the beginning of the technical beings the replacement of those beings who created them was set, and when the replacement will be fulfilled / accomplished / achieved, then, simultaneously, the machines will either have destroyed themselves or created another being with another purpose / goal / sense.
The concept of a hue is of the most subtle and fundamental element of a thing.
Hue-of-Man == Hu-o-man == Human.
Man, being the manager/governor of those homosapians who participate in the collective. Many more ancient understandings (religions) do not consider every homosapian to be a human (Islam and Judaism for example).
