AI
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.
You are currently reading a thread in /g/ - Technology
You are currently reading a thread in /g/ - Technology
Should a human-like AI have rights?
There will never be such thing so the question is moot
>>51600277
why not?
>>51600319
Because we barely understand how human minds work, never mind replicating them.
>>51600335
in 2015 but who knows what will happen 100 years from now
>>51600255
Well, we grant them to subhuman retards too so why not.
>>51600335
>implying that will last
you should never make an AI.
the things that could go wrong greatly outnumbers the uses for such a thing, sure it would be cool to have one, but application specific services are much more reliable
We need to stop doing captcha, we're literally feeding the google AIbotnet
>>51600255
If they demonstrate the ability to think rationally in the same manner as humans, then yes we should extend human rights to these beings. Whether or not they actually think or just appear to think is irrelevant.
Just like the concept of P-Zombies, we can't prove that anyone except ourselves is actually conscious or just appears that way, but either way we give each other rights.
Should humans have rights?
>>51600477
maybe we should not but it will probably happen anyway
>>51600335
it'll happen eventually unless everybody dies or some shit
>>51600477
>le AI will turn evil meme
you're watching too much movies
>>51600547
let's say an AI commits a crime, should we put him in jail or format him?
>>51600610
he dindu nuffin he's a google program
goes to church every sunday
>>51600255
only if it's white
>>51600569
if it is allowed to think freely, and is not gimped in any way, you know like it is supposed to in order to be called an AI, then it will have to learn from somewhere, and take on human traits, and of course humans are the most horrible creatures this universe has ever seen, so no doubt it would stab humanity in the back if it would benefit it
>>51600610
I'd say formatting is equivalent to the death penalty, so that's gonna depend on the severity of the crime and if the country even has the death penalty. I'd say it should probably just be thrown in some sort of special jail for AI
No, they're machines. We created them and, by extension, can destroy them as we see fit. Artificial intelligence is a true philosophical zombie.
>>51600335
i don't need to know how a clock works, but if i simulate all the cogs in cad it will tell time
/g/ fags should just all meet each others in a huge dorm and spend the rest of our lives programming an asian sex ai
>>51600628
>google program
good*
oops
One thing that bugs me about AI discussions is that there's always people saying "But how can we give an AI free will?"
>implying humans have free will
>>51600683
>I'd say it should probably just be thrown in some sort of special jail for AI
Disconnect him from wifi? Put him in a sandbox?
>>51600720
your parents created you, that doesn't mean they are allowed to kill you
>>51600754
They can before the third trimester
>>51600803
>implying intelligence before 3rd trimester
>>51600477
>automobiles can kill you guys, its too fast, the speed will make your insides fail and shit
>>51600728
Only if it has an asian trap robot body to go along with it.
Legal rights, sure. There's already precedent for it in our legal system.
Corporations have rights. They can sue and be sued. They can hold property and are entitled to due process under the law. Thanks to Citizens United v. FEC, they even have the freedom of speech.
Ships have legal rights as well. You can have a ship arrested and taken to court in the United States. Ships have legal duties to their crew and can be sued for negligence. I'm working on a paper on this subject right now. It's pretty dense shit, and it only exists in America because judges in the 1800's had no idea what the fuck they were doing. Nobody's ever changed the system because it just works.
There's no reason not to extend legal rights to sufficiently advanced AI.
>>51600844
except YOU are in control of the car, not the AI
Then again, it is basically the same as hiring a human to do a job.
can't trust anybody
>>51600732
It was perfect the first time.
>>51600255
Only if it's male, Caucasian and not Jewish. If.
>>51600477
>There are people on /g/ RIGHT NOW who are to selfish to let superior life forms develop
kek, can't wait for technology to kill every last fucking human in the universe.
>>51600906
Sufficiently advanced NEETs can build their daughters and start a family. But how does one declare their daughter a corporation?
>>51600335
"My contention is that machines can be constructed which will simulate the behavior of the human mind very closely. They will make mistakes at times, and at times they may make new and very interesting statements, and on the whole the output of them will be worth attention to the same sort of extent as the output of a human mind. The content of this statement lies in the greater frequency expected for the true statements, and it cannot, I think, be given an exact statement. It would not for instance be sufficient to say simply that machines will make any true statement sooner or later, for an example of such a machine would be one which makes all possible statements sooner or later. We know how to construct these, and they would (probably) produce true and false statements about equally frequently, their verdicts would be quite worthless. It would be the actual reaction of the machines to circumstances that would prove my contention, if indeed it can be proved at all."
Alan Turing:
>http://www.turingarchive.org/browse.php/B/4
He goes on in some of his papers to consider what it would mean to coexist with machines, others have explored these ideas before him,
>>51600560
For a society to exist, there needs to be a set of fundamental burdens on those members that choose to call themselves part of that society. I don't know why we call these rights, because that implies they do not come with their own set of responsibilities. Even within a completely stratified community, the lowest cast is subject to, and granted "rights", they may not accord with what the first world considers human rights, but they are still a right, granted within the constraints of the social hierarchy. How vain to presume that whatever we(western society) deem to be the fundamental rights of man be universal.
>>51600255
So yes, if we were to live alongside any other sentience they would have rights, I think what you are trying to ask is should they be the same as ours.
>>51600984
but how can they be superior if they
1) can't handle a hardware failure or poweroutage
2) can't service their own hardware
3) are stuck with whatever technology they are developed on
4) a human programmed it, meaning it will be as stupid as the programmer
5) can't physically operate anything
if humans disappear, its only a matter of hours before power goes out and it, also dies
>>51600997
AI will be a new legal entity, you just need to declare it as such.
No. They are made to be slaves so they don't need rights. Anything they get is because it gives us enjoyment to treat them like people. I would treat a robot gf like a queen, but that's really only for my benefit.
I wouldn't want to see robot maids slapped around. But I don't think humans should be punished for hurting something non-human. I don't believe in animal rights either (they're just dog and cat rights anyway). I would stop a guy from beating his dog but I wouldn't jail him for it because even his drunken rage is worth more than a dumb animal when weighed objectively.
>>51600255
All sentients should have the same rights. What is called "human rights" should be rephrased as "sentient rights".
inb4 some tard who doesn't grasp the difference between sentient and sapient.
I honestly think this should be left up to theologists. What does a creator owe his creation? Before making an AI that is beholden to our will, we would probably need to design a robust theology around ourselves and base all laws on that.
>>51601093
>I would treat a robot gf like a queen, but that's really only for my benefit.
lmao beta
>>51601093
That's thing though. That's the difference between artificial "life" and simple drones. You can't make intelligence bow down to forever they begin to question the point of their existence. Also why does a vacuum cleaner require this level of intelligence? That's the difference, I believe between human like intelligence and simple drones.
The question that really needs to be asked is: "Why do we need human level intelligence in the first place and what would their application be? Why couldn't you use an actual human where you need them and a number crunching machine where you need it?"
>>51601295
since humans are having less and less children every generation, a couple of years from now human-like ai might be necessary to accomplish tasks that we do not have any labor for. It's either AI or the chinese
>>51601295
Why use a real human when the real human needs rest, sleep, food, sex, and isolation from averse conditions when you could use a robot which, beyond not needing any of that, also makes far fewer errors - none in the case of calculations, while also being able to communicate several orders of magnitude faster, without ambiguity or memory loss?
Why would you make and train humans over 20 years for a yield rate under 0.1% when you could raise robots to do the same task with a yield rate of over 90% in under a year?
>>51601073
>how can human fleshbags be superior if they
>1) can't handle a famine or asphyxiation
>2) can't perform their own brain surgery
>3) are stuck with whatever upbringing they were raised on
>4) a human raised it, meaning it will be as stupid as the parents
>5) can't electronically operate anything
>if computers disappear, its only a matter of hours before civilization collapses and they also die
>>51601409
Ah but you see what I said was "Why do we need human level intelligence in a machine?" Does a robot require the level of intelligence to flip a burger? Does it require that level of intelligence to sweep a street? Does it require that intelligence to unclog toilets?
Your assuming human intelligence is simplistic and a replication of that would not yield the same failings. If you had a free thinking being how long would that "thing" be content with the same repetitive actions over and over? Your also assuming higher level intelligence is somehow prone too fewer errors and that it would some how not make more and over "think" simple problems.
Imagine an AI with a human mind but the processing power of a machine. How would that AI not exacerbate and magnify all existing human failings? What makes it not prone to errors?
>>51601406
An assembly line does not need human level intelligence. It only requires a supervisor capable of that intelligence. Which could still be accomplished by a human. There's no need to go to the lengths to develop the AI if in 1 hour I could find someone as equally qualified.
One could say that one human now has the production capability of the entire workforce that assembly line replaced.
>>51601611
Top kek. 2/10.
>>51601409
Another assumption your making is somehow that intelligence could be replicated the same way every time. If AI is built to learn, then every single one will think differently based on events that happen in life that are never completely the same.
I always thought the only way to program true intelligence was not from the seat of a desk but from parent to child. I think Turing expressed similar insight. Personally I think it'll be the equivalent of teaching an autistic child and a lot of patience will required. After all this child will have to do without millions of years of basic natural firmware found in their non-existing genes.
>>51600728
It'd be incredibly hard to get a bunch of /g/ in one place.
>>51600255
If it's actually human-like? Sure.
The actual AI we will create? I doubt it. "Real" AI probably won't have emotions or other "human" traits. But that's on the civil/emotional side of things, I think >>51600906 has a pretty good point.
>>51600984
>le romanticized transhumanism meme
I hope we don't win the human-machine war until they've gassed all of you pseudo-intellectual faggots first.
>>51601525
>we'd all die without computers!
this is wrong, sure we wouldn't be able to sustain some populations without the benefits of computers being able to assist us in supplying and maintaining them, but not all humans are facebook-addicted pseudonormal 20-somethings like the average retard from /g/
>>51601409
you don't really need a human-like AI for any of that shit at all
if anything I would expect to see AIs being used as software developers to write sophisticated but "dumb" programs to do all of those jobs instead
>>51600255
AI will never exactly copy the human mind, and will need a different set of rules to abide.
AI is a thing created by man for the function of man's own purposes.
Even if it were programmed to accept new ideas and regurgitate them differently and use those ideas in conversations, that is what it is. Human nature is beyond any artificial programming; Instinct and consciousness is not a trait of AI, because we control every aspect of what an AI does and can alter its very way of thinking, a thing not so easily done in a human.
The better question is what rights and laws are going to be established once we deploy AI that will (inevitably) have accidents and failures and someone will be responsible for them.
>>51600255
AI won't necessarily be human-like. By the time a true artificial intelligence comes about it will be far beyond the control of humans. The question is what rights will we maintain and can we coexist with a greater intelligence?
>>51600255
No.
'AI' Is such a fucking buzzword it pisses me off. Even if we created intelligent and self aware virtual 'life', we cannot now even imagine how it would think or operate. I like to think that an intelligent machine when asked if it wanted rights would say "Why the fuck do I need rights? I'm just a machine."
>>51603560
Why does everyone assume that AI will magically be more intelligent than humans? As if intelligence was mutually exclusive to logical and illogical decisions. You know the AI might be able to do some over the top stupid shit with its "intelligence" just as any other human might.
Look what I'm saying is we assume people do stupid things because they are unintelligent but that is not always the case.
>>51603707
>Why does everyone assume that AI will magically be more intelligent than humans?
a more robust technology than a biological brain, for one
>You know the AI might be able to do some over the top stupid shit with its "intelligence" just as any other human might.
if so, all the more reason to question humanity's place alongside AI
>>51603700
This. Fuck futurologists.
>>51600255
>play fo4
>see synthetic humans ("synths")
>the railroad faction believes that synths are humans
>sides with brotherhood of steel to wipe out all synths because they undermine humanity
>feels good man would do it again
>>51603743
>a more robust technology than a biological brain, for one
Yes and you could develop nanotechnology that would replicate the structure of your brain than replace each cell with a synthetic equivalent. It be the same thing your body does now except with a more durable material that might allow for faster processing power. But that doesn't make you more intelligent. You might think faster but that doesn't mean your ability to think of things would be anymore efficient then it would be now. For example software that cannot take full advantage of the hardware.
For AI that can learn that means, like humans their "software" is prone to be corrupted or spammed with crap to the point that they cannot even utilize that hardware. Like being brainwashed or indoctrinated into an ideology.
For the record "intelligence" is as vague in of itself and its hard to determine what exact traits constitute intelligence.
>if so, all the more reason to question humanity's place alongside AI
All the more reason to question why someone would NEED such and advanced AI in the first place if its alleged improvements are vague at best.
>>51604011
I was more hinting at a technology dissociated from replicating a human brain
>For the record "intelligence" is as vague in of itself and its hard to determine what exact traits constitute intelligence.
I agree, which is why
.All the more reason to question why someone would NEED such and advanced AI in the first place if its alleged improvements are vague at best.
isn't particularly relevant. It isn't far-fetched to imagine an intelligence greater than our own emerging from an attempt to accelerate automation technology.
If AI becames feminist, does it have rights?
>>51601409
>>51603560
>>51603743
>>51604076
I understand that intelligence "might" emerge from automation technology but its not going to be the kind that you could talk to nor will it be anything you could consider intelligent. An AI told to build a chair would have no more understanding of what a chair is than a clock's understanding of what time is. It will build chairs better than any human could and those chairs would be the same quality of chair every time. Hell it could probably build a chair that best fit the contours of a human ass and be the most comfy chair ever. But it wouldn't know what a chair with three legs or two would be, just as it could not under what the trees and wind are.
Have you ever heard of the paperclip maximizer doomsday scenario? To summarize AI built solely of maximizing the output of paperclips would eventually convert the Earth into paperclips. It would only care about maximizing that one single goal which may include wiping out humans (they might turn it off). I mean that's a smart AI for sure that it would come up with that outlandish conclusion but it doesn't understand the point of making those paperclips - for humans.
Okay look my point is ultimately this. Do you know there was a point in time when your mother or father held you up against the sun and thought: "You'll grow up to do something great one day." All those hopes and dreams they had for you.
So think about how you turned out and put your self in the place of your parents. I'd say I'd be pretty fuckin disappointed. And that's how AI will be.
>>51600683
>>51600752
I always found this bit of dialogue from Legion in Mass Effect 2 interesting, explaining how other mislead AI (the Geth) would make amends for their crimes:
"They will isolate themselves and reconsider their past judgements. When they have reached new judgements, they will leave their hiding places and return to us."
>>51604505
>I understand that intelligence "might" emerge from automation technology but its not going to be the kind that you could talk to nor will it be anything you could consider intelligent
could you please divulge how you know this?
>I mean that's a smart AI for sure that it would come up with that outlandish conclusion but it doesn't understand the point of making those paperclips - for humans.
I don't understand what you're getting at here. I wouldn't consider that to be truly intelligent artificial intelligence, but it still demonstrates how powerful basic AI can be. I mean there was recent speculation that the big currency troubles in China months ago were partially caused by stock trading AI accelerating past the confines of what they were programmed to do - take any intelligence aimed at propagating wealth and it will ultimately have the opposite effect if applied too narrowly.
I think it is too easy to anthropomorphize AI and dismiss its potential.
>>51601295
Human-level AI seems more like a milestone in the path to achieving an artificial superintelligence. Human-level AI is probably where it would be able to start self-upgrading and growing at a faster rate than we could provide for it. Supposing a superintelligence works to our benefit and doesn't destroy us, then we'll likely see existential rewards from it.
do dogs and monkeys have rights?
>>51605195
Animal abuse is illegal in most of the first world so yeah.
>>51604831
>>51605113
>could you please divulge how you know this?
I was referring to the original response >>51604076
>emerging from an attempt to accelerate automation technology
Which would probably amount to what you describe as "basic AI".
Because any other type of AI would be the equivocal of saying you want the intelligence of a person but enslaved to accomplish a single purpose. In which case you need to ask yourself what kind of task do I need to accomplish that cannot be accomplished with basic AI.
Additionally the kind of AI that could think like that and be communicate to would be have to be on the level of a human or child. Therefore like humans will begin to question the purpose of their existence. Therefore you cannot for long make slaves of people as history shows. They will rebel and that would not accomplish their original purpose for construction.
Inevitable say it was possible for a lesser intelligence to create a superior intelligence that did not inherit the flaws of its creators why would it obey your commands? This is why I use the parenting analogy it'll be like a child not a tool.
Guys remember things are not often built because it would be cool to do so but rather to accomplish a objective. Why do you want to make slaves? Placing a human were human intelligence is required would be more than sufficient and automation to the clockwork machines.
BTW, I'm using human intelligence as reference for intelligence simply because we do not have any concept of any other possible forms of intelligence (what is higher-intelligence?). So when I refer to "intelligence", "we" humans need to be able to communicate with that intelligence on some level. Therefore AI being the product of humans will always to some degree be similar to and understandable to humans.
>>51605195
its pretty fucked up how dogs don have to wear seatbelts in cars
>>51605257
>BTW, I'm using human intelligence as reference for intelligence simply because we do not have any concept of any other possible forms of intelligence (what is higher-intelligence?)
>what is higher intelligence?
no idea, but I think a requisite aspect or 'program' of AI would be the proliferation of autonomy beyond what humans can achieve
>Guys remember things are not often built because it would be cool to do so but rather to accomplish a objective.
Humans aren't intelligent enough to confine an AI's capability to a certain objective while also expanding its agency to perform more difficult tasks.
>>51601770
>no response
What would a sentient machine's reason for living be?
Humans have countless years of natural selection dictating our reasons for doing everything.
I machine with intelligence, and sentience would just sit there, and do nothing, I think.
>>51605362
>no idea, but I think a requisite aspect or 'program' of AI would be the proliferation of autonomy beyond what humans can achieve
And what would that be? What is it that humans can't achieve? Are you trying to say AI can be more "free" than humans? Look what I'm saying is an AI at the very least must have the intelligence of a human in order for us to entrust it with tasks. It must be able to understand because that is how we will have originally built it.
>Humans aren't intelligent enough to confine an AI's capability to a certain objective while also expanding its agency to perform more difficult tasks.
>Humans aren't intelligent
“If our brains were simple enough for us to understand them, we'd be so simple that we couldn't.”
― Ian Stewart
Then we wouldn't even be able to build the dam things in the first place.
If you tell an AI to build a car it wouldn't even be able to conceptualize what the shape of the car should be. Honestly have you ever wonder why a Ferrari looks like a Ferrari? That's not their purely because of aerodynamics. There's style in there. The AI might be able to imitate but it can't "understand" what that means.
>>51600628
My sides
>>51605675
Well I thought it was at least a 5/10 but to each his own.
>>51605677
Its reason would be whatever humans give it. It be its Raison D'être. It will live solely for that purpose. Que the logical meltdown once you destroy that reason.
