HBO released a new version of “Westworld” based on the old movie about a robot-based western theme park. The show hasn’t excited me yet — it repeats many of the old tropes on robots/AI becoming aware — but I’m interested in the same thing the original talked about — simulated experiences for entertainment.
The new show misses what’s changed since the original. I think it’s more likely they will build a world like this with a combination of VR, AI and specialty remotely controlled actuators rather than with independent self-contained robots.
One can understand the appeal of presenting the simulation in a mostly real environment. But the advantages of the VR experience are many. In particular, with the top-quality, retinal resolution light-field VR we hope to see in the future, the big advantage is you don’t need to make the physical things look real. You will have synthetic bodies, but they only have to feel right, and only just where you touch them. They don’t have to look right. In particular, they can have cables coming out of them connecting them to external computing and power. You don’t see the cables, nor the other manipulators that are keeping the cables out of your way (even briefly unplugging them) as you and they move.
This is important to get data to the devices — they are not robots as their control logic is elsewhere, though we will call them robots — but even more important for power. Perhaps the most science fictional thing about most TV robots is that they can run for days on internal power. That’s actually very hard.
The VR has to be much better than we have today, but it’s not as much of a leap as the robots in the show. It needs to be at full retinal resolution (though only in the spot your eyes are looking) and it needs to be able to simulate the “light field” which means making the light from different distances converge correctly so you focus your eyes at those distances. It has to be lightweight enough that you forget you have it on. It has to have an amazing frame-rate and accuracy, and we are years from that. It would be nice if it were also untethered, but the option is also open for a tether which is suspended from the ceiling and constantly moved by manipulators so you never feel its weight or encounter it with your arms. (That might include short disconnections.) However, a tracking laser combined with wireless power could also do the trick to give us full bandwidth and full power without weight.
It’s probably not possible to let you touch the area around your eyes and not feel a headset, but add a little SF magic and it might be reduced to feeling like a pair of glasses.
The advantages of this are huge:
- You don’t have to make anything look realistic, you just need to be able to render that in VR.
- You don’t even have to build things that nobody will touch, or go to, including most backgrounds and scenery.
- You don’t even need to keep rooms around, if you can quickly have machines put in the props when needed before a player enters the room.
- In many cases, instead of some physical objects, a very fast manipulator might be able to quickly place in your way textures and surfaces you are about to touch. For example, imagine if, instead of a wall, a machine with a few squares of wall surface quickly holds one out anywhere you’re about to touch. Instead of a door there is just a robot arm holding a handle that moves as you push and turn it.
- Proven tricks in VR can get people to turn around without realizing it, letting you create vast virtual spaces in small physical ones. The spaces will be designed to match what the technology can do, of course.
- You will also control the audio and cancel sounds, so your behind-the-scenes manipulations don’t need to be fully silent.
- You do it all with central computers, you don’t try to fit it all inside a robot.
- You can change it all up any time.
In some cases, you need the player to “play along” and remember not to do things that would break the illusion. Don’t try to run into that wall or swing from that light fixture. Most people would play along.
For a lot more money, you might some day be able to do something more like Westworld. That has its advantages too:
- Of course, the player is not wearing any gear, which will improve the reality of the experience. They can touch their faces and ears.
- Superb rendering and matching are not needed, nor the light field or anything else. You just need your robots to get past the uncanny valley
- You can use real settings (like a remote landscape for a western) though you may have a few anachronisms. (Planes flying overhead, houses in the distance.)
- The same transmitted power and laser tricks could work for the robots, but transmitting enough power to power a horse is a great deal more than enough to power a headset. All this must be kept fully hidden.
The latter experience will be made too, but it will be more static and cost a lot more money.
Yes, there will be sex
Warning: We’re going to get a bit squicky here for some folks.
Westworld is on HBO, so of course there is sex, though mostly just a more advanced vision of the classic sex robot idea. I think that VR will change sex much sooner. In fact, there is already a small VR porn industry, and even some primitive haptic devices which tie into what’s going on in the porn. I have not tried them but do not imagine them to be very sophisticated as yet, but that will change. Indeed, it will change to the point where porn of this sort becomes a substitute for prostitution, with some strong advantages over the real thing (including, of course, the questions of legality and exploitation of humans.)
There are already video recording techniques to record human actors in more than 3D — they are recorded from multiple angles so you can move your head and see proper parallax. This will be combined with a simpler “sex robot” which does a decent job of feeling like a person — with warmth and skin textures and more — but doesn’t look like a human at all. Looking human is the job of the video and the actors, real or synthetic. What matters is that the robot touches you in a way to match the scene, and that you don’t move so much that you break this illusion. As such, we’ll first see scripts where the subject stays fairly still, moving only his or her head and perhaps arms, and then advance to more limited movement.
Past history suggests people are quite willing to suspend disbelief in these situations, and stick within the rules to play out their fantasies. Later, the range of what is possible will expand. People will know it’s not real (as they hopefully know the prostitute doesn’t actually have feelings for them) but immerse themselves in the fantasy world. The initial technology might be expensive, suggesting the creation of “VR brothels” where people (let’s face it, primarily men) go to rent a booth, but eventually it could become cheap enough for private ownership.
This leads to a number of interesting social consequences. This technology might well reduce the demand for prostitution, which most people think is a positive thing. Prostitution is one of the leading drivers of demand for modern human slavery. (Slavery has not gone away. In fact, in absolute numbers, there are reportedly more slaves in captivity today than there were in the early 1800s at the height of the classic slave trade period.)
More controversial will be the potential to offer disturbing fantasies. Violent sex and sex with minors. These controversies exist already in the more basic world of porn. Some push to ban even simulations of these acts. Others believe that offering simulated experiences may replace the demand for the vastly more pernicious real experiences. There will also unfortunately be demand for VR video of real, not simulated, child sexual abuse.
It will also be possible to create remote experiences, where a remote sex worker interacts with the customer in live VR, just as is done over videoconferencing today. (Indeed it was 40 years ago that my friend Ted Nelson coined the term teledildonics to refer to these types of interactions, both with sex workers and between couples.) In fact, live streaming VR experiences — sexual and otherwise — may be one of the largest drivers of bandwidth demand in the coming decade. It will become possible in time to alter the experience of the remote person. This might allow a 50 year old woman or man to play a 20 year old woman. And, controversially, it could also allow them to play an underage person.
The world will have much to debate, and hopefully the extreme issues will be a small minority market and not cloud discussion of the bigger picture. Even the existence of relatively tame porn has always been controversial. The latest issue revolves on whether the vastly higher availability of porn for teen boys is clouding their impressions of what women are or should be like and “raising the bar” in a bad way, reducing their ability to be satisfied with their non-fantasy partners. This will be multiplied by VR and VR/robot offerings, though in the early days when it is expensive it will not be readily available to teens. We will also see uses that are more healthy, including interaction between distant partners, or even for couples who are physically together who want to do a bit of consensual make-believe.
A brave new world, is coming, folks.