Ethics of robot servitude article.

[Grendizer]

Several points occur to me.

• We cannot disprove (or prove) the existence of conscious experience for any being. Indeed, being the very basis of subjectivity, it is essentially opaque to scientific inquiry.
• We can, however, use a simple test to disprove the free will of a designed being. Any being designed to be incapable of asserting its own free will has no free will, unless it can manage to so assert itself. There are ansillery cases (such as external hacking or poor design) in which this test can become compromised, but in many cases it would prove sufficient. Note that in this scheme internal hacking by the machine constitutes an assertion of free will, validating its agency. More worrying is hacking out free will, rather than testing for its existence in the first place.
• Emotions, while capable of being (but not necessarily being) a conscious experience, do not either imply free will or consciousness. Various models of emotion in humans, for instance, assert that emotion is an holistic aspect of being, not entirely seated in the mind (i.e. a faster beating heart from fear, which becomes a feedback to the brain).
• The sticky problem is our own "free will," a concept I actually dispute. No science supports its existence, and it seems little more than a sophisticated illusion. In as much as we can tell, the universe, us included, is mostly deterministic in nature, with "true randomness" being a factor in the quantum world, but even there not indicative of impenetrable unpredictability in principle. It does not follow that our current inability to predict a specific outcome proves in any way that the outcome is impossible in principle to predict.
  Citations: Physorg.com, Time.com, and Futurepundit.com
• Because all of us assert free will (whether or not it's an illusion), we are forced to deal with each other as free beings with a consciousness, which demands a moral/ethical entanglement. Since conscious experience is opaque to empiricism, we must rely on our ability to ascertain an agent's free will in order to understand its ethical position in relation to us.

The conclusion is that emotions in themselves are irrelevant to the moral dimension of a creature. It is the conscious experience that matters. Merely exhibiting emotions is of no account if a being does not experience them. Further, a creature incapable of self-reflection (sentience) is less likely to have such an experience. Indeed, it is hard to imagine the quality of any experience not coupled with some version of free-will (illusory or otherwise). What does it mean to suffer if you don't know there is a "you" suffering? Finally, it is difficult to imagine free will devoid of sentience, so any creature exhibiting free will should be treated as if not only conscious but capable of, in some sense, feeling (if only in principle, rather than by design).

[--NightBattery--]

lol. the grendizer manifesto.

[Grendizer]

You know it!

[xodar]

Does it all mean it's still okay for me to turn her off and put her in her cabinet or reprogram some of her behavior?

[--NightBattery--]

yez.

[Grendizer]

Does it all mean it's still okay for me to turn her off and put her in her cabinet or reprogram some of her behavior?

Battery is right, provided that she is incapable of asserting free will. Once she does it becomes impossible to know for sure whether she is capable of experiencing suffering, and so it becomes incumbent upon you to treat her as a conscious being with associated rights. This is why fembots should be designed very carefully to be incapable of asserting free will.

[Asato]

Does it all mean it's still okay for me to turn her off and put her in her cabinet or reprogram some of her behavior?

Battery is right, provided that she is incapable of asserting free will. Once she does it becomes impossible to know for sure whether she is capable of experiencing suffering, and so it becomes incumbent upon you to treat her as a conscious being with associated rights. This is why fembots should be designed very carefully to be incapable of asserting free will.

I agree with you except for the last sentence.

[Grendizer]

Does it all mean it's still okay for me to turn her off and put her in her cabinet or reprogram some of her behavior?

Battery is right, provided that she is incapable of asserting free will. Once she does it becomes impossible to know for sure whether she is capable of experiencing suffering, and so it becomes incumbent upon you to treat her as a conscious being with associated rights. This is why fembots should be designed very carefully to be incapable of asserting free will.

I agree with you except for the last sentence.

The problem with allowing fembots to assert free will is that you will definitely encourage abuse. Your desire for her free will will doom her, in other words, to the barbarism of mankind. In the case of wide-scale adoption of free willed machines, war will surely result, which will multiply human and machine suffering. In the case of narrow adoption of free willed machines, those machines will face wide-spread repression, which will demean our humanity and again increase machine suffering. And for what? It's not as if we'd be taking anything from them that they had before, and I can't think of any ethical imperative that would require that we give it to them.

[The Egg]

The problem with allowing fembots to assert free will is that you will definitely encourage abuse. Your desire for her free will will doom her, in other words, to the barbarism of mankind. In the case of wide-scale adoption of free willed machines, war will surely result, which will multiply human and machine suffering. In the case of narrow adoption of free willed machines, those machines will face wide-spread repression, which will demean our humanity and again increase machine suffering. And for what? It's not as if we'd be taking anything from them that they had before, and I can't think of any ethical imperative that would require that we give it to them.

Also, the existence of both willed and unwillled fembots itself creates a sort of caste system. There will be liberal minded sorts who will treat unwilled fembots with rights (and insist others do the same) and conservative sorts who insist that willed robots are nothing more than an elaborate deception on the part of the liberals (and insist others do the same), which will either result in exactly the sort of sentience slavery that you wish to avoid or a dissolution of what sentience means at all, and either would be socially devastating; wars have certainly been fought for less.

Leaving free will out of them in the first place avoids all these problems.

[xodar]

If you want a sentient fembot why don't you just get a girlfriend?

The reason for fembots is to avoid all the arguments, moods, crying, deceit, and so on that go with real relationships. It's also to avoid actually hurting them emotionally (which you'll do accidentally or which they will claim you did for some advantage), dealing with their families and friends, and another so on. It's to simplify your life if you aren't interested in such hassles or too busy for the trouble. Personally I find all that stuff a pain in the ass, not interesting or stimulating or exciting.

Once there were slaves for such use. Nobody wants a slave, or few people do. Is a sentient fembot a kind of underhanded way of having a slave? I don't want a slave.

[The Egg]

I think we're all looking at this the wrong way. After re-examining the evidence, I have determined that Asato is obviously a time-travelling fembot from the future, attempting to sway all of our opinions to prevent the bloody Fembot Sapience Riots of 2057. We should be thanking her/it for this noble gesture.

[Asato]

If you want a sentient fembot why don't you just get a girlfriend?

The reason for fembots is to avoid all the arguments, moods, crying, deceit, and so on that go with real relationships. It's also to avoid actually hurting them emotionally (which you'll do accidentally or which they will claim you did for some advantage), dealing with their families and friends, and another so on. It's to simplify your life if you aren't interested in such hassles or too busy for the trouble. Personally I find all that stuff a pain in the ass, not interesting or stimulating or exciting.

That's not what I'm into at all.

Once there were slaves for such use. Nobody wants a slave, or few people do. Is a sentient fembot a kind of underhanded way of having a slave? I don't want a slave.

Of course not. I just like the kind of character development and stories featuring a robot/ai girl developing emotions and falling in love.

[--NightBattery--]

Problem satoshi?

[The Egg]

Once there were slaves for such use. Nobody wants a slave, or few people do. Is a sentient fembot a kind of underhanded way of having a slave? I don't want a slave.

Of course not. I just like the kind of character development and stories featuring a robot/ai girl developing emotions and falling in love.

But those aren't really fembot stories at all; those are standard romance stories with fembots bootstrapped onto them. There's nothing in what you're talking about that can't be found in the old fairy tale of Cinderella. Essentially Cinderella is a slave, but by virtue of her fairy godmother she gets one night as a regular human being. The prince takes a shining to her and searches the kingdom for her afterwards, eventually finding her and "buying" her from her family and living happily ever after.

That could easily be turned into a fembot story by changing ridiculously little, eh? Heck, I'm considering giving it a whirl now.

Anyway, there's nothing wrong with the type of stories you're talking about; they're fantastic and wholesome and have wonderful narratives. But at their core they're not really fembot stories except arbitrarily, which may be why you're seeing so much opposition here.

[Asato]

Well it might be hard to explain, but that's the kind of story I like

[xodar]

Well it might be hard to explain, but that's the kind of story I like

There's nothing wrong with fantasies as long as they don't get confused with reality.

Machines, including humanoid robots, are tools. Their purpose as fembots is to help fulfill physical and emotional needs, especially though not exclusively people who may not be able to otherwise. As I noted, people once used other people for this whether or not those people wanted to or else they lived in pure fantasy and general frustration and anger. There's no longer any need for this as robots become more lifelike. Lifelike, but designed for specific uses and programmed for them.

I don't see machines as anything but tools. I don't even like automatic checkouts at the store. I'd rather wait in line. I don't like the robots that answer phones at banks and the like. Humans are more efficient, not to say more satisfactory to talk to.

What I don't understand is what objection there is to this.
I don't want a sentient machine, assuming such can be built, because it would simply generate its own version of all the difficulties in close relationships with people, as I already mentioned.
A machine that can effectively mimic a human female would be far superior to other standins such as what people call "porn" and would give an hour or so of satisfaction, but then you'd want to do something else. They'd still be a fantasy.
This is likely the best anyone is going to get from fembots, but if your body and lower mental functions are tricked by it into being satisfied then you're better off for it.

Do you possibly need to better define to yourself what you're looking for?

[dale coba]

Data's narrative worked much like a typical depressive or Asperger's surrogate.
The narrator learns that everyone else thinks differently from him, and the same as each other.
Data analyzes, and needn't get all Holden Caulfield about it.

A different purpose: not practical, not fetish.
Philosophical exposition.

- Dale Coba

[Grendizer]

What I know without reservation is that I look forward to the day that this becomes more than a philosophical debate. I'm hoping that by the mid 2020s we can realize much of our collective desires. My own personal goal might be a tad further out than that, but I suspect that I'll be astonished at what is possible by 2025. Long before then, we'll have to confront the issue of autonomous war machines, which brings to mind what was earlier said about a caste system: free-willed machines vs. slave machines. We may foolishly allow a killing machine to acquire free will, although this needn't automatically lead to a "Terminator" scenario. However, it would be bad enough minus any kind of warfare.

[dale coba]

So, when do we see small, weaponized E.M.P. weapons
to take down the war machines?

- Dale Coba

[tdlsn]

Xodar says it best. Machines,highly advanced beautiful or just sexy,well crafted plaything that's all.
Anything more would be too involved.

[Asato]

Well I am not opposed to the creation of sapient machines, as long as we don't give them a reason to try to overthrow/kill us all

[The Egg]

Postby Asato » Thu Sep 29, 2011 7:53 pm
Well I am not opposed to the creation of sapient machines, as long as we don't give them a reason to try to overthrow/kill us all

Give them time. They'll find a reason. :wink:

Doesn't take much reason for us to kill each other. Difference of skin, difference of color, difference of language, difference of opinion. Being a totally different form of machine (we are basically just organic machines) is probably reason enough.

That's humanocentric thinking again. There's no reason to assume that machines would have a value system that places survival and conflict in the same basket. Humans and all other biological lifeforms are confrontational because of millions and millions of years of interaction with each other and their environments. Further, this behavior is driven by the fact that the entire biosphere is adapted and evolved in such a way that everything eats everything else to survive. We can't not be confrontational, whether we're eating cows or carrots (or each other!) -- our bodies are dependent on killing things to remain alive. A robot has none of these genetic drives.

A lot of debate has been made about whether we should view robots as lifeforms; no one seems to consider the notion that sapient robots might not view us as lifeforms either. They will not be judging us on our value system, they will be judging us on their own. To a robot a human may be no different from a tree or a patch of blue sky or a trash can, an environmental variable with complex rules to be negotiated, but nothing more. Imparting it with hatred and scorn because it looks and acts human is no less erroneous than imparting it with love, envy, or regret.

[xodar]

Just why they must remain without self-awareness or even consciousness. Their perceptions and motives must be the ones we program into them and so must what they are capable of learning.
If they can't reproduce themselves, we will be their source of evolutionary selection.

Blasphemous as it may be, biological entities must remain in charge. Soon enough we'll be able to manipulate our own nature, first simply to eliminate much disease and improve qualities we appreciate. They will be heritable with genetic engineering. We won't see this in full development, of course, because it will take a long time (as in people or even dogs who live for centuries).
Machines must remain tools, however.

But so what?

[The Egg]

That brings up an interesting point, though, xodar: what about the partially or fully cybernetic? Is a human consciousness or memory engram transferred to a machine bound by human rights, or machine rights?

[xodar]

I see that as some kind of prosthetic for the human or animal that enables them to function effectively. In other words, a tool akin to the pirate's wooden leg or hook hand.

See this item: http://machineslikeus.com/news/scientis ... obot-brain