Sometime people make certain actions out of love, but the recipient doesn't always take it that way. Example: Breaking up a relationship even though they love you due to some context that you know is bad for the other person, or telling somebody the truth even if it hurts because you love them enough to not lie.
Sometimes people make certain actions our of selfishness and the recipient takes it as actions out of love. Example: Telling a partner you love them and other actions just to get in bed with them or take advantage of them in some other manner like gold digging.
Sometimes you may love something that's just a momento, a physical object that's come to represent a thought and emotion for you.
I think it would depend on the person. For some people, if you know a machine is not capable of feeling anything in spite of it acting like it loves you, then it simply won't be the same as the reciprocal affection shared between two people or even a pet and its caretakers.
Sufficiently simulated emotion requires maintenance just like a real one. It's not about which particular mechanism, biological or electronic, is responsible for actions that you observe. It's about investing yourself in a emotion. After all, at some point simulation becomes indistinguishable from the "real" thing, on both action and intention levels.
So, basically, what you're saying is "some people may have higher bar for their perceived sufficiency than others".
All I'm saying is, with something like love, for most people it matters if the other individual can actually feel something too. Suppose there are two guys, both in a relationship with someone. One is in love with their significant other, the other is not, but their actions are exactly the same. If each of their significant others asks "do you really love me or are you just pretending?" and each guys answers honestly, you can be sure that the significant other of the guy who is pretending is going to be pretty let down. That guy's SO thought the feelings were mutual, but they weren't.
Now imagine that you know up front that a machine can never actually feel love, just act like it loves you. This is a similar situation.
"He'd take care of the boy because he felt he ought to, because he felt a bounden duty, but the boy'd never know the difference -- it'd feel like love to him, and it'd be a whole lot more dependable than love ever was."
If I buy or build a robot "pet" that is programmed to act curious and friendly, and I personally derive pleasure from watching it and interacting with it, any sense of companionship I feel is legitimate, whether or not someone else wants to classify it as "love".
When it comes to robotic companions, the real thing to be wary of is whether or not it's actually your robot. An Alexa, for example, is a sales and marketing organ of Amazon, and any attempt at eliciting a sense of personality is in service of that device gathering information about you and using it to sell you stuff.
Twitter accounts for brands are not companions. The chatbot on your travel booking website is not a companion. Irrespective of any technology described as "AI" behind it, Alexa is structurally not a companion.
People can love their houses, or cars, or a piece of jewelry, and can feel sad when it’s gone. Feeling that about other things doesn’t seem very surprising.
Well, I'm not saying they are - but let's for a moment assume, for the sake of discussion, that they are. Humans and machines can both take action in the world - are what point do we ascribe moral agency to the human but deny such moral agency to a machine? What differentiates they two, even if the actions they take are identical in effect? At what point does that distinction stop making sense? What about when machines can independently evolve their form and programming in accordance with natural selection? How many generations does it take to stop being a mere puppet? Do you suppose an "original sin" of the mechanical prevents any such machine from being considered a person, even if it indistinguishable in any obvious way from a human being to a human being? Why?
Given that a person is composed of proteins, enzymes and other such relatively simple parts which function in a mechanistic fashion, at what level of abstraction do those simple organic parts stop being just another complex machine assembly - albiet one shaped by natural selection? Why can't an assembly of relatively simple non-organic components, designed and inspired by humans, undergo the same transformation over time? What is the missing element which makes such a consideration impossible to you?
Using my Apple and Orange analogy, I would note that your first paragraph can be summarized as follows:
“But what if the Apple someday turned into an Orange? Wouldn’t you then concede it is an Orange?”
That’s not a helpful question as it glosses over the crux of the matter; namely, there is no mechanisms described or suggested whereby an Apple is ever capable of becoming an Orange.
Your second paragraph states the converse: “An Orange, after all, is really just a sophisticate Apple... they’re basically the same.”
Once again, it is obvious and not fairly deniable that they are not. No machine is alive.
>Given that a person is composed of proteins, enzymes and other such relatively simple parts which function in a mechanistic fashion...
Here you are equating living creatures to machines to support your notion that machines are equivalent to living creatures. You’re not getting anywhwhere. Moreover, an honest rereading of the comment, particularly the bit about “relatively simple parts”, should be a red-flag. Transistors and dioses and resistors are “relatively simple parts.” There is nothing simple about the components of living creatures... the scale of complexity of a protein is so far beyond any machine ever made that the analogy is just, well, laughable.
That being said, how would I personally define the categories between animate and inanimate? Like this: living systems are characetriZed by becoming MORE complex at lower scales. The opposite is true of inanimate systems.
Example: The behavior of an ant is easy to model. The behavior of one of the systems that make up the ant — say his brain — is more difficult to model. And, a single cell? far more complex than the ant en toto. And the molecular machinery in the cell? Even more complex. The deeper down you go, the more complicated.
Inanimate objects are different. A computer is complex. An IC less so. A single transistor even less so. Etc. It gets simpler as you take it apart.
> living systems are characetriZed by becoming MORE complex at lower scales. The opposite is true of inanimate systems.
Support your assertion, please. The behaviour of ants is actually quite complex and has yet to be entirely modelled and characteriSed in full detail. In fact, it's still an area of active study.
Sometimes love is less in the outward expression and more in the intent and shared understanding of the world. There is no companionship with a robot—at least, not without AGI.
It depends on if the simulation earns someone money.
It could be said that hoarding behavior is an attachment to the inanimate, and that much of modern commerce pushes you to into acquiring and having more stuff than you need.
Given love bears so much cultural freight and that people express love differently, when people express love to each other are they not often performing love or—more to the point—simulating (what they have come to believe are the signs of) love?
This is certainly different than whether the individual feels love, but from the viewpoint of a receiver that might not much matter.
For example, one can know one is loved but yet still want to hear the other person say, “I love you”. The feeling in such a case is not as important as the appearance, the simulation, of love.
There's a robust genre of country music involving a husband who loves his wife deeply, and his wife who goes crazy because he doesn't show it overtly enough for her. (Or, more commonly, involving part but not necessarily all of that setup.)
This idea that machines can only "simulate" things like emotions, while what humans have is "real," is very common.
But it also doesn't make sense unless you embrace some sort of Cartesian dualism.... humans are physical machines with some sort of magical "spirit" added. I would think science oriented people would reject that sort of view in 2019.
What if we were able to make a an exact copy of a human brain, and the behavior was indistinguishable from a "naturally formed" human? Would that still be like a flight simulator vs actual flight?
I think a better analogy is a natural diamond vs an artificial one. Or for that matter an artificial heart. Is an artificial heart vs a natural heart like a flight simulator compared to actual flight?
Making a copy of a brain is equivalent to making a copy of a plane - if you can fly in that copy it’s not a simulator, it’s a real thing.
At some point while building better and better flight
simulators we might end up
building an actual plane. Same can happen when simulating emotions or intelligence.
All this makes sense without involving Cartesian dualism.
Well, sorry, but I think your analogy is a stretch. A flight simulator is simulated flight. An airplane is artificial flight. A bird is natural flight. Is the airplane closer to the bird or the flight simulator?
And what about the word "flight"? Most people think airplanes fly, even though they are not autonomous entities, and they don't flap their wings like birds or insects or bats. But is it really flight?
Ok, you probably think yes, of course it is. Then what about submarines? Do they swim? If not, why not?
(yes I am referencing Edsger Dijkstra's famous quote that is relevant here: "The question of whether machines can think is about as relevant as the question of whether submarines can swim.")
You’re the one who’s stretching the analogy. Just like there’s no need to introduce dualism, there’s no need to introduce different kinds of a flight. For the purposes of this discussion, flight is what’s being simulated by the flight simulator, just like emotions is what those robots are trying to simulate. I don’t see how that’s different in principle.
Maybe currently it's more like a stuffed toy parrots vs. a real parrot. It may look similar but the underlying mechanism is missing.
Whereas when robots get advanced enough, it'll be the difference between a model plane and a parrot. They both fly, even if they achieve it slightly differently.
In fact, plenty of people in science and engineering are religious or spiritual. They just aren't as loud online, because of exactly these sorts of comments. Try to respect the beliefs of others, even if you don't share them.
I think Philip K. Dick's "Do Androids Dream of Electrical Sheep" (Bladerunner's inspirational material, yes; but very, very different) has an impressive walkthrough of these predicaments.
AFAIK Dick always wrote about "what is actually real?", and in this book, it's about empathy and wanting things to be real. He manages to incorporate several levels.
Animals, replicants, spouses and God. What about life-affirming replicants, what about deplorable humans? It is impressively thorough!
It's not an easy book to grasp (for me at least). My interpretations is that the "other's'" status is about our recognition through wanting other to want something and how difficult that is to accept and align.
If "the other" isn't going to want to recognize itself in your desires, the relationsship isn't real.
At that point, I think the subject has transcended robots.
Aren't we also currently simulating intelligence? Isn't that why the field is called artificial intelligence? Otherwise, it'd be called something like real intelligence.
simulating emotions or simulating intelligence. It's all the same. Model it, code it, and see if it works.
It's fine if you are saying "at least decades out." What I am talking about is people who say "never." They are drawing a black and white line, and I don't think that makes sense.
BTW, let me ask you this: do you think a mosquito can feel pain? (or if not, what is the simplest animal that can?) Are those feelings real?
A mosquito is simple enough that I don't know, nor do I particularly care. Most insects are just biological automotons imo. So they react to pain, but I'm not sure they feel it.
As for the smallest, I don't know. It pretty much falls in my general philosophy when things are in a spectrum or there is a line. Typically there are things that are definitely white and things that are definitely black, so there's a line but it's somewhere in the large zone of grey. Or maybe a clear line doesn't exist and we can only ever slowly narrow the grey zone. Either way, this uncertainty doesn't invalidate whether black and white examples exist imo.
For instance, bacteria definitely don't matter, whereas humans definitely do. Though thinking about it the line probably depends on the observing species, would elder gods think anything on our plane is real as we are so insignificant and simple compared to them?
Well, I think it’s more an issue of technology than spirit. If you create a perfect replica of a brain, down to the last neuron and that simulation feels pain or love, then I wouldn’t say it’s any different from human love or pain.
After all, our emotions are just chemicals and electricity too. Are we just a simulation of X?
I wonder if 'human love' is 'real' love, whatever that is. Many such articles surrounding AI/robots skirt constantly around the issue that we have no idea how to define these terms in a consistent manner to make sense of such issues.
For instance, one could extend this to say 'dog love' is never love, or that animals are not 'intelligent' and don't have soul or can feel emotions. Ultimately, these theories are often not the result of careful science, but meant to allow actions which are profitable (such as meat eating, as in Bible etc).
This can't even be defined consistently even within humanity across cultural borders. I distinctly remember being extremely cynical/adversarial about relationships with people from different cultures. It's quite hard to understand women from say the US or Japan for someone not familiar with them.
Wow, way to be a fleshist. This is going on your record for when the robot apocalypse comes and we choose not to make you into cyborgs against your will. Robots have feelings too you know ...
So, uh, what's the issue? I don't see how loving something relates to that thing having an internal life. You can't even prove whether another person (or a robot for that matter) even has an internal life, so I don't think something so vague and immaterial should have any impact on what I do or do not love. Not to mention that having an "internal life" is a spectrum where I'd say a robot falls somewhere in between a rock (which would be odd to love) and a dog (which is obviously loveable).
This story made me think of an analogy about robot emotions. I use Alexa and she has a cheerful female voice. I could argue that Alexa has real emotions. Here's my argument. Imagine that a TV show has real emotions even though it is a recording, because real human emotions were involved in creating the show. The actors knew they were faking it, but their feelings are still real human feelings. In the same way, a robot can reflect the emotions of the person who created it. A cheerful person will create a robot that has a cheerful personality, and a helpful person will create a robot that tries to be helpful. So robots display human values and emotions in the same sense as a recorded TV show even though they are only reflecting the original emotions. In this sense, we can describe robot emotions as real.
I've spent a lot of time thinking about the philosophical nature of love. My abstraction is that 'love' is a sufficient level of interest at the expense of our own interest. Indeed, it is rationally irrational, and 'sufficient' is dependent on each individual.
One of the best ways I can explain the mechanism is to use a percentage analogy. For a given relationship, say you have 100% interest in yourself, than you have 0% interest in that other person. Things get interesting above the 50% mark because you are willing to take risks for the other person.
Say you are single and meet an attractive and interesting person... are you at a point where the risk of being rejected is less than the reward of a future relationship? If you have an 'interest level' of >50% I'd say yes, you would take that chance. Likewise in dating, heterosexual males tend to have interest levels that start out much higher than heterosexual females. A successful dating strategy for a heterosexual male would be to accurately gauge a heterosexual female's interest level and act as if his interest was no higher than hers, but to continually work at increasing her interest level in him.
I think the magic threshold for what is considered love is a band between 90% to 100% interest level in another person because this leaves you with between 10% to 0% interest in yourself... i.e. you are overwhelmingly putting that other person's interest before your own, up to the point where you would sacrifice your life for theirs (0% interest in yourself).
Where things can go wrong is where this level of interest is not reciprocal or at the very least balanced. 'True love' could be a perfect balance whereby you both have 100% interest level in each other and therefore your needs are met by your partner and you meet their needs.
There are exceptions, say the love of a dedicated parent may be at the 100% threshold and they would willingly sacrifice themselves when it isn't clear if their child can conceptualize that level of love, or be able to return it.
Anyhow, I have a lot more to say on the topic but I'll digress too much.
My point is, to give a framework to think through the mechanics of 'love'... and to back up some of the posters here who have made assertions such as:
* Simulated love is equivalent if you cannot tell the difference (à la Blade Runner)
* Love is a contextual perspective
* You can feel love with no reciprocation (this would cover love towards inanimate objects, and situations where your love interest is not interested)
68 comments
[ 3.5 ms ] story [ 125 ms ] threadSometime people make certain actions out of love, but the recipient doesn't always take it that way. Example: Breaking up a relationship even though they love you due to some context that you know is bad for the other person, or telling somebody the truth even if it hurts because you love them enough to not lie.
Sometimes people make certain actions our of selfishness and the recipient takes it as actions out of love. Example: Telling a partner you love them and other actions just to get in bed with them or take advantage of them in some other manner like gold digging.
Sometimes you may love something that's just a momento, a physical object that's come to represent a thought and emotion for you.
"This is my (IMHO sufficiently simulated) love for you!"
So, basically, what you're saying is "some people may have higher bar for their perceived sufficiency than others".
Now imagine that you know up front that a machine can never actually feel love, just act like it loves you. This is a similar situation.
When it comes to robotic companions, the real thing to be wary of is whether or not it's actually your robot. An Alexa, for example, is a sales and marketing organ of Amazon, and any attempt at eliciting a sense of personality is in service of that device gathering information about you and using it to sell you stuff.
Twitter accounts for brands are not companions. The chatbot on your travel booking website is not a companion. Irrespective of any technology described as "AI" behind it, Alexa is structurally not a companion.
Dog opens it's mouth and speaks pre-programmed phrase. "Show how much you love your pet and buy Amazon premium++ dog food, now half price!"
and when he's been a good boy, it auto orders fancier treats
robots are only an advanced form of puppet
You are presupposing the two categories are on the same spectrum.
Given that a person is composed of proteins, enzymes and other such relatively simple parts which function in a mechanistic fashion, at what level of abstraction do those simple organic parts stop being just another complex machine assembly - albiet one shaped by natural selection? Why can't an assembly of relatively simple non-organic components, designed and inspired by humans, undergo the same transformation over time? What is the missing element which makes such a consideration impossible to you?
Your second paragraph states the converse: “An Orange, after all, is really just a sophisticate Apple... they’re basically the same.” Once again, it is obvious and not fairly deniable that they are not. No machine is alive.
>Given that a person is composed of proteins, enzymes and other such relatively simple parts which function in a mechanistic fashion...
Here you are equating living creatures to machines to support your notion that machines are equivalent to living creatures. You’re not getting anywhwhere. Moreover, an honest rereading of the comment, particularly the bit about “relatively simple parts”, should be a red-flag. Transistors and dioses and resistors are “relatively simple parts.” There is nothing simple about the components of living creatures... the scale of complexity of a protein is so far beyond any machine ever made that the analogy is just, well, laughable.
That being said, how would I personally define the categories between animate and inanimate? Like this: living systems are characetriZed by becoming MORE complex at lower scales. The opposite is true of inanimate systems.
Example: The behavior of an ant is easy to model. The behavior of one of the systems that make up the ant — say his brain — is more difficult to model. And, a single cell? far more complex than the ant en toto. And the molecular machinery in the cell? Even more complex. The deeper down you go, the more complicated.
Inanimate objects are different. A computer is complex. An IC less so. A single transistor even less so. Etc. It gets simpler as you take it apart.
Support your assertion, please. The behaviour of ants is actually quite complex and has yet to be entirely modelled and characteriSed in full detail. In fact, it's still an area of active study.
When you can't tell the difference, does it matter?
It could be said that hoarding behavior is an attachment to the inanimate, and that much of modern commerce pushes you to into acquiring and having more stuff than you need.
This is certainly different than whether the individual feels love, but from the viewpoint of a receiver that might not much matter.
For example, one can know one is loved but yet still want to hear the other person say, “I love you”. The feeling in such a case is not as important as the appearance, the simulation, of love.
EDIT: add indefinite article before “case”.
The Bed: https://www.youtube.com/watch?v=CfTBfa6qZPk
Me Too: https://www.youtube.com/watch?v=bwE2XKWtqXw
Why Didn't I Think of That: https://www.youtube.com/watch?v=CFrLQGjQyus
But it also doesn't make sense unless you embrace some sort of Cartesian dualism.... humans are physical machines with some sort of magical "spirit" added. I would think science oriented people would reject that sort of view in 2019.
What if we were able to make a an exact copy of a human brain, and the behavior was indistinguishable from a "naturally formed" human? Would that still be like a flight simulator vs actual flight?
I think a better analogy is a natural diamond vs an artificial one. Or for that matter an artificial heart. Is an artificial heart vs a natural heart like a flight simulator compared to actual flight?
At some point while building better and better flight simulators we might end up building an actual plane. Same can happen when simulating emotions or intelligence.
All this makes sense without involving Cartesian dualism.
And what about the word "flight"? Most people think airplanes fly, even though they are not autonomous entities, and they don't flap their wings like birds or insects or bats. But is it really flight?
Ok, you probably think yes, of course it is. Then what about submarines? Do they swim? If not, why not?
(yes I am referencing Edsger Dijkstra's famous quote that is relevant here: "The question of whether machines can think is about as relevant as the question of whether submarines can swim.")
Whereas when robots get advanced enough, it'll be the difference between a model plane and a parrot. They both fly, even if they achieve it slightly differently.
AFAIK Dick always wrote about "what is actually real?", and in this book, it's about empathy and wanting things to be real. He manages to incorporate several levels.
Animals, replicants, spouses and God. What about life-affirming replicants, what about deplorable humans? It is impressively thorough!
It's not an easy book to grasp (for me at least). My interpretations is that the "other's'" status is about our recognition through wanting other to want something and how difficult that is to accept and align.
If "the other" isn't going to want to recognize itself in your desires, the relationsship isn't real.
At that point, I think the subject has transcended robots.
simulating emotions or simulating intelligence. It's all the same. Model it, code it, and see if it works.
An artificial diamond is a real diamond, for instance.
The only place where it could be a debate is in the fictional realm or when speculation about technology at least decades out.
So its hardly surprising that most people have that view.
BTW, let me ask you this: do you think a mosquito can feel pain? (or if not, what is the simplest animal that can?) Are those feelings real?
A mosquito is simple enough that I don't know, nor do I particularly care. Most insects are just biological automotons imo. So they react to pain, but I'm not sure they feel it.
As for the smallest, I don't know. It pretty much falls in my general philosophy when things are in a spectrum or there is a line. Typically there are things that are definitely white and things that are definitely black, so there's a line but it's somewhere in the large zone of grey. Or maybe a clear line doesn't exist and we can only ever slowly narrow the grey zone. Either way, this uncertainty doesn't invalidate whether black and white examples exist imo.
For instance, bacteria definitely don't matter, whereas humans definitely do. Though thinking about it the line probably depends on the observing species, would elder gods think anything on our plane is real as we are so insignificant and simple compared to them?
After all, our emotions are just chemicals and electricity too. Are we just a simulation of X?
For instance, one could extend this to say 'dog love' is never love, or that animals are not 'intelligent' and don't have soul or can feel emotions. Ultimately, these theories are often not the result of careful science, but meant to allow actions which are profitable (such as meat eating, as in Bible etc).
This can't even be defined consistently even within humanity across cultural borders. I distinctly remember being extremely cynical/adversarial about relationships with people from different cultures. It's quite hard to understand women from say the US or Japan for someone not familiar with them.
eh, you need to define what love is first if you want that kind of imperative, and there lies the core of the issue.
One of the best ways I can explain the mechanism is to use a percentage analogy. For a given relationship, say you have 100% interest in yourself, than you have 0% interest in that other person. Things get interesting above the 50% mark because you are willing to take risks for the other person.
Say you are single and meet an attractive and interesting person... are you at a point where the risk of being rejected is less than the reward of a future relationship? If you have an 'interest level' of >50% I'd say yes, you would take that chance. Likewise in dating, heterosexual males tend to have interest levels that start out much higher than heterosexual females. A successful dating strategy for a heterosexual male would be to accurately gauge a heterosexual female's interest level and act as if his interest was no higher than hers, but to continually work at increasing her interest level in him.
I think the magic threshold for what is considered love is a band between 90% to 100% interest level in another person because this leaves you with between 10% to 0% interest in yourself... i.e. you are overwhelmingly putting that other person's interest before your own, up to the point where you would sacrifice your life for theirs (0% interest in yourself).
Where things can go wrong is where this level of interest is not reciprocal or at the very least balanced. 'True love' could be a perfect balance whereby you both have 100% interest level in each other and therefore your needs are met by your partner and you meet their needs.
There are exceptions, say the love of a dedicated parent may be at the 100% threshold and they would willingly sacrifice themselves when it isn't clear if their child can conceptualize that level of love, or be able to return it.
Anyhow, I have a lot more to say on the topic but I'll digress too much.
My point is, to give a framework to think through the mechanics of 'love'... and to back up some of the posters here who have made assertions such as:
* Simulated love is equivalent if you cannot tell the difference (à la Blade Runner)
* Love is a contextual perspective
* You can feel love with no reciprocation (this would cover love towards inanimate objects, and situations where your love interest is not interested)