The Trouble With Robobabes

copyright 2005 by Pat Powers

I've written several movie and hentai reviews that have led me to rant about the idiocy of the assumptions made about the nature of robobabes in such works (i.e., female robots designed to slake their owners' sexual desires have all the feelings of a normal human woman).

That's because I've made a study of the actual technology that might be involved in developing a robobabe, and I'm going to tell you about it here, so you, too, can knowingly razz the pathetic efforts of screenwriters whose appreciation of technology begins and ends with cell phones.

OK, to start with, what do you need a robobabe to do if she's to work as a sex toy? Well, to be honest, there's only one place a robobabe really needs to shine, and that's in the bedroom. She doesn't need to go about in public or anything like that. In fact, she doesn't even need to be able to walk. Be nice if she could, but crawling about on all fours is about all that's needed, and if necessary just the ability to kneel in various postures would do the trick.

(Some guys might fantasize about robobabes they can take home to mother, but frankly, such guys are seriously deranged. If you want a sex partner that's just like a real live girl in every respect, date a real live girl, OK? Sure, they'll tell you "no" but they'll also tell you "yes" sometimes. But back to our robobabes, who will only tell you "no" if you program them to, which is kind of the whole point of robobabes, though you wouldn't think it from watching hentai featuring robobabes).

Robobabes just have to be able to move well enough to simulate sex, something some human women don't even do well. Ideally, robobabes should handle this better than human women, after all, it's their specific purpose.

Robobabes don't have to be able to mimic human speech in every respect, but they should be able to moan and sigh and cry out in passion at the very least. Some ability to make small talk would be a plus, and of course, the ability to say "Yes master" and such will be needed for slavegirl robobabes.

The really important stuff relating to a robobabe has to do with how she looks, feels, smells and moves. Obviously, the more she looks, feels, smells and moves human, the better. This is also the hardest part.

Now, how might we get from where we are to a fully functional robobabe?

Let's start with locomotion. A lot of work has already been done and there are Japanese robots that can walk about in a sort of human manner. The slow and clumsy gait of such 'bots is a far cry from a sexy, slinky stride, but it's mostly a matter of fine tuning. We may already have the technological capacity to create a robobabe that can walk and/or crawl in a convincingly sexy human manner, and if we don't, it's hard to imagine that it's more than ten years in the future. It's a fairly difficult problem but it's already well on the way to being solved. In fact, there are military apps for any robot that can walk reliably over uneven terrain, and it's likely that eventually we'll have robots that greatly exceed human capacities in this respect.

Speech is another easily solved problem, primarily because you don't need artificial intelligence to mimic human speech. All you need is an expert system. Capacity here far exceeds what's needed -- in fact, a very early expert system called Eliza, designed to conduct psychological therapy sessions and written almost as a joke, was so successful in fooling patients into thinking it had some insight into their characters and problems that it alarmed its creator.

Given the nature of a robobabe's services, it would probably be difficult to design a speech program that didn't emotionally involve owners in the robobabe. This is one case where fooling owners is so easy it's almost scary.

Mimicking other human attributes is much more difficult, but often has been mostly solved due to other applications.

For instance, scientists have developed a substitute for human skin that pretty much matches the real thing as a temporary substitute for burn victims to wear while they heal. I had thought this might be one of the tougher problems but fortuitously, it's already been solved. Whether this substance has the strength, durability and most of all, FEEL of human skin is a question I don't know the answer to, but in any event, owners of Real Dolls (more about them later) claim that their latex exteriors have a remarkably human feel.

There's also the matter of odor. In any sexual activity, odor is obviously going to play an important part, but perfumers have been working on it for centuries, and there are some pretty good solutions to that one in the bottle already. We're not going to have to do anything more than build in some scent holders in a few strategic places.

Now we come to the real difficulty, the problem that has not as yet been solved. Eye-hand coordination, specifically, realistic, human-like eye-hand coordination.

This problem has already been attacked very determinedly for reasons that have nothing to do with robobabes and everything to do with mass production.

Factory owners would very much like to be able to purchase and install in their plants machines which could mimic the ability of human hands and arms to move, combined with the abiity of human eyes to see and understand what they see. Then they wouldn't have to give money to those nasty, smelly humans who keep insisting on more money, safe working conditions, reasonable hours, and who sometimes get you in trouble for being kids or for being enslaved.

So there's plenty of desire out there for working robotic hands, but so far, no results. This is gonna be a tough one. There's no reason to suppose it can't be done, but every reason to suppose it will take awhile -- at least ten years, more likely twenty years, maybe fifty years.

With all these trends in existence, it seems logical to suppose that robobabes won't be designed as such from scratch, but will "evolve" from more primitive devices.

We already have a precursor to robobabes, a product called Real Dolls. They're lifesized latex replicas of women, and they're pretty darned lifelike. The early models had perpetually-open mouths and vaginas that gave them a blow-up doll look but the later models have solved that problem and are reputed to have a lifelike look AND feel. Inside and outside, if you know what I mean.

I really don't expect a lot of progress on this front until an armature is developed that allows the doll to move in a lifelike way. Being able to speak or moan is technically possible right now. But in the absence of movement that's not much.

As soon as a reasonably priced armature that can move a Real Doll or something like it in a lifelike way is developed, things will move very quickly because then all the other advances will make huge amounts of sense.

So as I see it, there are two key issues: developing lifelike movement capabilities and developing it at a reasonable price. I really can't say how difficult the price problem will be. It could be a matter of developing software algorithms that can drive existing technology, which might well get reasonable in price quickly, as the Japanese are going to take the lead in both developing and paying for the early development of robotic walkers. By the time the software is ready for use in robobabes, most of the development costs will have been covered, and it'll be relatively cheap to install software that will allow a robobabe to locomote about.

Of course, suitable writhing and wriggling software will also need to be developed, but that challenge will also be met by the Japanese, who have a little cultural problem with deformity. Seems that a guy in Japan who's lost a limb or otherwise varies from the Japanese norm in some dramatic way has little or no chance of getting laid in Japan, even by prostitutes. They just can't handle it. So the Japanese in their bizarrely kind/cruel way, have designed lifelike, lifesized dolls for such men. (I really don't know if women with physical deformities have trouble getting laid or not -- on the one hand, hentai have demonstrated that Japanese men will do just about anything for an orgasm. But the whole physical deformity thing in Japan is kinda strange so I can't say for sure.)

Anyway, with the Japanese absorbing the initial development costs with their greater cultural interest in robotics and love dolls, it'll get cheap by the time it gets to the U.S.

So, what I see happening commercially is a Real Doll equivalent that "moves and moans just like a real woman". That's gonna be popular, and it'll also arouse some pretty strong reaction from the general public of the "isn't it awful that men go for these when there are lonely real women around?" sort -- for a short while, until the media get distracted by something else. (Say, 2 days max to judge by modern standards.)

Still, I don't think there'll be a real robobabe around that's capable of fooling anyone until they come up with a realistically moving face and mouth, i.e., one that'll do more than just open wide enough to admit a cock.

To be really, dangerously lifelike, a robobabe should have facial expressions that change from moment to moment as is demanded by the situation and a mouth that can move in a realistic simulation of speech.

Frankly, the mouth challenge is the lesser of the two. Anyone who's ever watched a film for several minutes before coming the the realization that it is dubbed will realize that a mouth doesn't have to move exactly in synch with speech to pass muster under ordinary circumstances. Designing a robobabe that moves its mouth slightly whenever speech is occurring, opening wider when volume is louder and less when it's lower, should do the trick just fine, even if it's detectable if you work at it to any extent. Let's face it, robobabe customers are going to want to believe -- in fact, the suspension of disbelief will be the strongest factor in a robobabe's favor with regard to success in the marketplace.

The matter of expressions will be much harder. Canards about male disinterest in female faces aside, we are all trained from babyhood to be very attentive to human faces and the expressions they wear. That's why beginning artists find that faces are the hardest thing to draw. It's not that faces are that much more complex than other objects an artist might draw, it's that your average person knows what a face should look like to a much greater extent than they know what anything else in the world should look like. Frex, an artist can draw a chickadee that's anatomically incorrect in any number of ways so that no self-respecting bird fancier would call it a chickadee, but so long as it doesn't have tentacles or razor sharp fangs, most folks will think it a fair representation of a chickadee.

But draw a human face with a nose slightly askew, a mouth positioned too low on the face, eyes a millimeter too close together or any other of thousands of subtle errors, and people will immediately spot the error and describe the work as bad art.

The same problem will plague the manufacturers of robobabes if they are not careful. Oh, it'll be no problem to create a robobabe face that looks natural in repose, but it will be very hard to create a robobabe face that looks natural when it moves. A robobabe whose facial "muscles" aren't exactly, precisely engineered to human specs is going to look like she is grimacing, has facial tics or perhaps even as if she's having a seizure. This is off-putting under ordinary circumstances, but is extraordinarily off-putting during sex.

Frex, I once worked with a woman who was quite the beauty -- long blond hair, creamy complexion, beautiful features, slim figure -- but she didn't have much control over her face at times. Like, when she ate. Her face would convulse in vaguely alarming ways. Once you'd seen her face contorting as she downed a banana or a spoonful or forkful of anything, it wasn't hard to imagine her face contorting in exactly the same way during a blowjob. It wasn't a pleasant prospect, though not enough to throw one off the notion of getting a blowjob from her, of course. (Sadly, she was a very proper woman and the prospect of blowjobs was distant, to say the least.)

Strangely enough, it's a seemingly little thing like this that might hold back acceptance of robobabes long after the basic problem of locomotion is fixed, because there's no real money working on it for other reasons.

But surprisingly, there ARE people working on a very similar series of problems who might develop solutions that are transportable to the robobabe dilemma.

They are the Poser art community. They have been working on the problem of making 3D wireframes move in a reasonable simulacrum of human movement. Except that Poser artists have been doing things like developing realistic walks for Poser wireframes wearing high heels, and improved ways of realistically wrapping ropes, chains, tentacles and such around Poser wireframe bodies. In short, they're my kind of artists.

While it's very obvious that you couldn't directly transport the code from a Poser wireframe to a real world powered armature designed to move a robobabe around, but clearly the kind of thinking that it takes to make virtual armatures move in a sexy way will also be useful in solving the problem of making real armatures move in a sexy way.

In any event, the business of getting armatures and so forth to move in the right way is likely to prove difficult, but let's face it, we're not talking about FTL drives, time travel or anything like that. We're really talking about applying current knowledge to existing technology in new ways. It's doable, the only question is, how long will it take before it's good enough? After all, it's unlikely we'll need analogs for all 53 muscles in the human face to fake some pretty realistic expressions. As few as a dozen might do the job well enough for most purposes, especially those having to do with sex.

Somebody will probably buy Real Dolls which have the gait and the facial expressiveness of Frankenstein, so long as they look like Britney Spears (or name your personal icon of Hot Babery here). But when Robobabes look and MOVE like Britney Spears (and Britney DO know how to move) then a LOT of guys will want them. To be precise, ALL guys will want one. It will be the new iteration of video porn, something everyone wants, with every authority figure and church group deploring the fact that everyone wants one. And it'll be about as "stopable" as video porn was (and is) at that point.

I mean, let's face it. Once the robobabe is fully developed, it will mean that any guy with X amount of dollars can own a simulacrum that looks like Jennifer Lopez (need I mention that the faces and figures will be interchangeable) moves like a young Tawny Kitaen, smells like an incredibly horny woman in a field of flowers, has the voice of Anita Morris and the sexual inhibitions of Jenna Jameson -- he's gonna want one.

In fact, robobabes could well become the television of the future, the big ticket home appliance that everyone making more than minimum wage MUST HAVE.

In summation, robobabes are feasible, within the range of existing tech or tech that's very likely to come online in the next few years, and are very likely to prove an irresistible purchase.

But that's not really the point of this essay. The point of the essay is that, having thought about robobabes as a technology, as a product as it were, I've come up with all sorts of points on which to hang a plot, what's more an original plot, not Pinocchio version 5010.2.

Frex, take the problem with facial expressions. You could hang a nice story there, of a guy whose robobabe has an incredibly beautiful but immobile face who is overjoyed when he gets the chance to purchase a hardware/software upgrade that will allow her to "Smile, laugh, frown, cry, look puzzled and emulate up to 14 other human expressions!"

He thinks this will make her even more real and thus more valuable to him, but finds that instead it makes her seem less attractive, too much like real women. He discovers too late that he has invested a certainly otherworldly sexiness in the inhuman calm of his robobabe's face, and as it loses the calm, it loses the sexiness.

Or the problem with hands. You could do a Telltale Hands kinda story about a guy who it totally OK with his robobabe except when he sees and/or feels the mechanical way her hands move. Then his erection goes bye-bye. What will he do? What WILL he do?

Or you could do a story about the first court case in which a robobabe is named as a causus belli in divorce proceedings and the prosecution/defense calls on her to testify.

If you've taken the time to do a little thinking about what you're actually writing about and how it might actually happen in the real world (as I have in the Jetta 3000 novel) you have a superior SF product going. Of course, it's ultimately all about the story -- no amount of trappings can make up for a story that doesn't have interesting characters and a compelling plot. But you'd be surprised how much more often interesting characters and compelling plotlines are associated with original storylines, as opposed to the sameold-sameold.

What's more, this article alone is an excellent start -- on the evidence, reading it is a lot more research than most people do when they're writing a story about a robot woman, man or child.

The true determining factor in whether or not we have a real lifelike robobabe as opposed to a sophisticated doll (its Turing test, so to speak) is this: could you have a guy go to a whorehouse and make love to one and never know it's not a real woman? Now, I'm not talking about a drunk guy, or a real idiot, but a sober regular guy who's out to get his Friday rocks off at the local House of Blue Lights, a guy who's been around the block recently enough to have a clear memory of what doing it to a real woman is like.

When it becomes commonplace that regular guys can be fooled, we will have real robobabes. When will it happen? I'm thinking 30 years for a bedroom-only robobabe that can crawl and suck cock prettily, and up to 100 years for the Real Thing to appear. I don't really think it WILL take 100 years -- more like 50 -- but I'm giving myself some fudge factor because it has become obvious that sometimes seemingly trivial Real World things can be very hard to emulate in a computer.

Finally, what's my personal feeling on robobabes? I think that they are absolutely topnotch masturbatory aids, nothing more. Because when "no" has no meaning, "yes" has no meaning as well. Some guys who could date women might find them a superior alternative, but for most they'll be at most a backup -- when a date with Ms. Right doesn't work out, Mrs. Right Now will be there, in her box, ready to be played with.