[chuckles] I'm sorry, Ex Nihilio, I wasn't referring to any trivial interpersonal nonsense when I talked about hostility, I meant hostility to the idea. It's an interesting thought experiment and engineering problem ... how do you raise children in the absence of adults but with an arbitrarily sufficient amount of preparation.
Now while macaques did show abnormal behavior when raised by synthetic proxies, how much of that behavior reflected actual harm? Contrary to Christopher's example of human babies deprived of touch showing physical abnormalities, here we're talking about creatures that were shown affection by the proxies. Neonates responded to simulated macaque behavior and sought reassurance and nurturing from the proxies. But just because they developed into mature animals with different social reactions, that doesn't mean they were actually harmed ... they simply grew up differently. If you raised a Japanese infant in a orthodox Jewish family, the infant would likely grow up with Jewish customs and beliefs.
I strenuously object to your statement,
No matter how lifelike a simulated caregiver could be, they would never be able to raise a child from infancy that would be close to a “normally” socialized person.
It's simply too absolute a position to have correlation with reality. A simple series of thought experiments should illustrate the weakness of your stance. Suppose the child was raised by R2D2 ... not humanoid and almost no social cues. I'd be amazed if any child raised in that condition came out remotely functional. But suppose it was raised by C3P0. Now we have a caregiver that can speak, hold its hand, change its diaper, and even teach it. True, there's no warmth in its touch (I assume), but there's a strong humanoid form that gives the neonate's instincts something to fix on. I'd still be surprised if the child grew up "normal" by societal standards, but I'd expect it'd do better than the R2 baby. Now let's raise another infant by Commander Data ... much more humanoid and in many ways indistinguishable to humans. Are you sure that child will grow up recognizably different? Especially if Data's got his emotion chip plugged in? Now let's go all out. Let's raise a child with an android indistinguishable from humans. Good social skills, appropriate emotional responses, something like "Bishop" from Aliens, but more maternal. Is this synthetic, nurturing surrogate insufficient for the baby's growth and development?
I'm not so sure.
Now let's take it from the reverse direction. A human mother, who, through accident or disease or crazed social science experiments conducted by trekkies, must lose key elements of humanoid features. An accident could leave a mother with no legs, one arm, no breasts, and a loss of bilateral symmetry in her face. She is functional in every other respect and determined to be a caregiver in spite of her injuries. Should she be refused? Let's give her minimal bionic replacements. Modern prosthetic limbs for mobility and some grasping functions. She's not as mobile and lacks the dexterity of C3P0 in drag, but she's still at least part human. Let's give her slightly science fictional prosthetics now that includes full finger dexterity and some haptic feedback, but also replace her remaining natural arm ... the injuries were worse, but she's got better medicine to survive. She even has some artificial organs. So she's less human than before, but more functional. Would you argue with me that she can't be a mom? Now let's go all out (again) and replace almost all of her physical body ... her bionic form is much more advanced than we can build today, and probably indistinguishable from reality at a distance and possibly up close. She still has a human brain and some glands, and maybe a kidney or something, but the rest is all synthetic. Can she raise a baby?
My point is a simple call to recognize a continuum of responses to different artificial caregivers. Even R2D2 would be a better caregiver than none at all, and if we have the tech to build that fiesty little droid, we should be able to make one that can do the job a bit better. Remember that a key evolutionary advantage of humans is their adaptability and while we might consider a child raised by robots socially retarded and scarred, the simple fact might be that that child has adapted fluidly to his or her environment and claimed it as his own.
Finally regarding the matter of the "magic" of Fire's proposed technology, you said, "In the original post Fire suggest this device should be built using the present tense, indicating that this is a project he feels should be initiated now." I'd like to point out that he said no such thing. His first post said:
Now the technology would be beyond what we have now but I dont think we're far off.
He plainly acknowledges it can't be done now. The "I don't think we're far off" bit is rather undefined and subjective. Was he speaking in terms of decades, centuries, or millennia? Was he speaking of a linear growth of technology or something like Kurzweil's exponential growth?
