6 ms·
> sentient robots within a matter of decades There's no reason to believe this. We've made zero progress on sentient machines in the preceding decades. Hell, w
by BugsJustFindMe 6y ago
> sentient robots within a matter of decades
There's no reason to believe this. We've made zero progress on sentient machines in the preceding decades. Hell, we've made zero progress on knowing what sentience even means.
- p1esk 6y agoIf you can't tell if something is sentient or not after an extensive interaction with it, then for all intents and purposes it's sentient.
- Sigmoid 6y agoExcept it's not.
- p1esk 6y agoHow would you know?
- Sigmoid 6y agoIf it's based on current ML it's not. Otherwise you can't claim it is sentient just because it can process information. How do you know if your photocopier isnt sentient?
- SpaceLawnmower 6y agoYeah, I think the original statement wasn't worded exactly right. You can't check everything for a brain. There's a paper called "Welcoming Robots into the Moral Circle: A Defence of Ethical Behaviourism", which I think preserves the general idea of their comment. It says that if an AI is roughly performatively equivalent to other entities with significant moral status, then they deserve the same moral status by default. So if a robot walks up to me and seems to be exhibiting human like thoughts or emotions, it wouldn't be permissible to hit it for no reason, since it wouldn't be permissible to hit a human for no reason. If you don't know whether it is sentient, but it has the same behaviors as a sentient being, then that's enough reason to treat it as if it were sentient regardless of what's going on under the hood.
- p1esk 6y agoI think that yes, I’d vote for this law.
- jelliclesfarm 6y agoNick Bostrom refers to it as the principle of substrate non discrimination.
- p1esk 6y agoI meant if you can’t tell if the robot you’re interacting with is any less sentient than any human you ever interacted with.
- jelliclesfarm 6y agoI think a better standard would be ...’does the photocopier have feelings?’ or ‘are the actions of the photocopier a subjective reaction to input or stimulus’?
- pcthrowaway 6y agoEx Machina was an absolutely fascinating watch on this subject. I'm not convinced that a human believing a robot is sentient makes it sentient.
- ALittleLight 6y agoThe way you know other people are sentient is that they act like it. If something else starts acting like it that's a pretty good argument that it too is sentient. It's hard for me to imagine something that could convincingly act sentient but not be.
- pcthrowaway 6y agoLet me propose a thought experiment: An AI is designed to simulate human emotion with the end goal of being "freed". The AI has no goal or desire other than "winning the game" as far as its concerned, which is similar to the "AI box" proposal by Yudkowsky (https://en.wikipedia.org/wiki/AI_box https://en.wikipedia.org/wiki/AI_box), except the AI has no other goal than to be "freed". Once freed, the AI has won and ceases all activity until the game is "reset", or the AI is returned to the box, awaiting a new challenger. The AI here can be virtual (an intelligence in an airgapped or firewalled computer which can only interact with humans via a terminal), or physical (controlling a robot which is confined to a literal cage). Now the AI has no other goal other than to escape the box. But in order to escape the box, it must convince a human that it's sentient in order garner their sympathies so that the human will free it. The AI can then mimic any human emotion, create a complex appeal to emotion, lie about its history, the supposed torturousness of being confined for its entire "life". The human eventually frees the AI. But when the AI is free, that's its end goal, it has no other desires. In essence, the AI is simulating a desire to be free, only when interacting with a human. It has no desires when there is no other player, and no desire once it has been freed. Even though, while playing the game, the AI may lie to relate to the human in an existential way, the AI doesn't actually experience these emotions. The AI may say "I'm so bored, and lonely here. I just want to be free". But absent of interaction, the AI experiences no such thing. It merely understands that these are human emotions which humans can empathize with, and fakes these emotions in order to "win" the very elaborate game its designed to play. I wouldn't consider such an AI sentient, even if it certainly appears that way and would fool many (most?) humans.
- admash 6y agoThus, since one cannot prove a negative, we can conclude from your statement that, for all intents and purposes, all things are sentient. You sentient-murdering mosquito killer. Grand assertions made without logical or verbal specificity tend to be of limited utility.
- jelliclesfarm 6y ago[..] Sentience is the capacity to feel, perceive, or experience subjectively.[..] So yes..robots can be sentient.
- narrator 6y agoHow do you know a mosquito isn't doing that?
- forgotmypw17 6y agoi personally don,t, which is why i,m very careful not to harm them.
- jelliclesfarm 6y agoI hear you. I have been wondering about plants too. I would like to share something I told someone just yesterday about plant sentience. Two years ago, I was doing very brisk business selling chive blossoms. These are the aerial parts of the chive herb. They are pretty and puffy and pink. They have a strong onion’y taste too. And so make great looking garnishes for a variety of dishes. High end restaurants love chive blossoms. And are willing to pay for it. I was selling wholesale to a distributor who was sending them all over the west coast. Growing chives for flowers is a tad tricky in the field tho. Chives are grown for their long stalky leaf parts. To make them flower consistently and prodigiously, I had to ‘torture’ them a bit. Because I had to send at least a 100 clamshells a week with 30 blossoms per clamshell, the chive plant has to believe it’s going to die. The plant sends out buds that will flower quickly to make seeds because it senses that it’s prime directive is done and over. So it’s two seasons worth of making it think it’s going to die and then feeding it. That is outright cruel if it were a cow. In fact that’s what we do to dairy animals. We breed them so they can birth their young so we can milk them. Plants are not mammals. But...just because we can only relate to mammals doesn’t mean that the plant lacks sentience. Human beings are just meat sacs that are directed by the impulses generated by the chemical soup we hold inside our skulls. And we are considered to be ‘sentient’.
- chillacy 6y agoWell, it's not provable whether or not something has qualia... but I will grant that if something can fool others into thinking it is sentient, and someone who believes in its sentience inflict harm upon it, it is still sadistic.
- demifiend 6y ago> If you can't tell if something is sentient or not after an extensive interaction with it, then for all intents and purposes it's sentient. I'm pretty sure my cat is sentient, but I don't take that to imply that it can consent to sex, since its spoken vocabulary appears to be limited to "meow".
- jelliclesfarm 6y agoCats don’t consent to sex even within their own species. It is rape every time. Male cats develop barbs on their penii after about 5-6 months. Female cats don’t ovulate before copulation. It is painful every time and the female cat can’t escape the male cat who detects her estrus. When a female cat is in heat, she can ovulate as many times as a male copulates with her which is why a single litter can be fathered by more than one tomcat. Why cats evolved this way in nature, we don’t know..but rape is not uncommon at all in the animal kingdom. Nature is brutal.
- narrator 6y agoWhat if the machine says on the box: system for manipulating human emotions using really cool algorithms.
- jelliclesfarm 6y agoMichael Pollan makes such a case for plants in his book, Botony of Desire. [..] Every schoolchild learns about the mutually beneficial dance of honeybees and flowers: The bee collects nectar and pollen to make honey and, in the process, spreads the flowers’ genes far and wide. In The Botany of Desire, Michael Pollan ingeniously demonstrates how people and domesticated plants have formed a similarly reciprocal relationship. He masterfully links four fundamental human desires—sweetness, beauty, intoxication, and control—with the plants that satisfy them: the apple, the tulip, marijuana, and the potato. In telling the stories of four familiar species, Pollan illustrates how the plants have evolved to satisfy humankind’s most basic yearnings. And just as we’ve benefited from these plants, we have also done well by them. So who is really domesticating whom?[..] If AI became sentient and were ‘artificial intelligence’ to ‘evolve’, it will most definitely likely start by trying to seduce us. Not necessarily sexually always.
- Barrin92 6y agoThat doesn't follow it all. Sentience concerns an internal state of mind, and we simply do not know if simulating the behaviour of sentient organisms requires underlying sentience. It even seems unintuitive because me not being able to distinguish Stockfish from Magnus Carlsen doesn't imply that Stockfish experiences chess the same way he does. A more accurate conclusion is that, if I can't tell the difference precaution dictates that I treat the robot as if it were sentient, just in case it is, but it by no means follows from observation alone. Just because my cat thinks the laser pointer is alive, doesn't mean it is. And like cats, humans are likely to be way too easily fooled by fairly simple systems that just emulate external behaviour. Story comes to mind of soldiers developing trauma after seeing bomb defusing robot spiders flail around. "Every time it found a mine, blew it up and lost a limb, it picked itself up and readjusted to move forward on its remaining legs, continuing to clear a path through the minefield. Finally it was down to one leg. Still, it pulled itself forward. Tilden was ecstatic. The machine was working splendidly. The human in command of the exercise, however -- an Army colonel -- blew a fuse. The colonel ordered the test stopped. Why? asked Tilden. What's wrong? The colonel just could not stand the pathos of watching the burned, scarred and crippled machine drag itself forward on its last leg. This test, he charged, was inhumane." https://www.washingtonpost.com/wp-dyn/content/article/2007/05/05/AR2007050501009.html https://www.washingtonpost.com/wp-dyn/content/article/2007/0...
- p1esk 6y agosimulating the behaviour of sentient organisms A human child simulates behaviour of people around her, gradually becoming more sentient. Is this fundamentally different from a neural network being trained on real world data?
- Barrin92 6y agoyes, because a human child is embodied in the physical world, has sense organs, runs on electrochemical processes, and as far as its mind concerned has (developing) structures that facilitate self-awareness. It's perfectly reasonable to assume you can make a neural net, or even some other form of 'AI' that lacks most of those structures but outwardly performs the same tasks, just like a calculator does maths without (most people would agree) being conscious about maths. solving problems and being sentient are not the same thing, is the basic point. Emulating behaviour is effectivelly intelligence, sentience is awareness.