If you've watched Netflix's new sci-fi series 3 Body Problem (and you should, it's excellent!) based on the novel by Liu Cixin, you probably wondered about the amazing virtual world that's a key part of the story. Without spoiling anything, a mysterious alien species communicates with humanity through a multi-user virtual experience with seemingly sentient NPCs and a constantly changing world, conveyed in full sensory VR. (Watch a preview clip above.)
But when can us humans make something like that? I asked a number of top virtual world/VR experts, including Philip Rosedale, Raph Koster, VR pioneer Nonny de la Peña, and Sine Wave CTO Adam Frisby. Their aggregate answer: Both sooner and later than you might think!
"I think we aren't too far from sentient beings in virtual worlds," Philip Rosedale tells me. "AIs that are very interesting to interact with and living in virtual worlds will happen very soon... next few years. We have a lot of compute coming available, and that is the only thing blocking."
However, he's skeptical we're close to full sensory VR technology:
"I think it may never be possible, because the ability to directly stimulate the brain (as compared to reading it, which may be easier) may never be possible, even with invasive devices like stents or brain surgery."
And no, he doesn't think something like Elon Musk's Neuralink is necessarily a first step to that:
"It is unclear how many tiny wires you'd have to stick into your brain and whether you could learn to use them and whether they would change or stop working over time," Philip argues. "Also note that in those cases where it does work (like the paralyzed guy) the subjects have to slowly learn how to control things -- neuroplasticity. This is very different than stimulating a brain in a way that virtualizes sensory input. Making you 'see' your mom standing in front of you -- no one knows how to do that, no matter how many wires are stuck in your neocortex."
Raph Koster, who's developing Playable Worlds, a new metaverse platform, had a similar reaction around the 3 Body headset:
"I mean, how is that thing even powered? We definitely don’t have direct brain stimulation for sensation."
He does think the graphics fidelity is possible soon: "We can make the photoreal imagery. The headset rendering is going to take a while before we can get the illusion of reality. Hardware challenges with the displays."
So his final scorecard:
"Sooner than people think: Worlds that can change and collapse.
"Years: Hyperreal rendering, sentient seeming NPCs that hold up for more than a few minutes.
"Unknown time: Batteries, full sensory support."
Adam Frisby estimates it will take 15 years to get something like the virtual world with sentient AI aspect: "Sensory VR part maybe longer. We don't have much of a pathway to taste and smell at the moment, though I think Neuralink is restoring sight in apes, so ???"
Nonny de la Peña is actively helping develop similar technology now: "We are working on a haptics for Storytelling Lab. We have been investigating a number of different options from suits to gloves to environmental sensors."
As for how many years for when we get something like the virtual world made by aliens: "Hopefully less than 400!" (Which if you know, you know.)
For my part, I'd say it's less about the technology than social adaptation. Since the install base for high quality VR remains so small, and doesn't grow substantially even as the tech improves, there's no reason to think there's a large market for the full VR technology we see in the TV series. (As opposed to being an important but pricey niche for people with, say, physical disabilities.) So that seems over a hundred years away at least.
As for the virtual world/NPC aspect, I agree with Adam that's probably 15 years out. We probably won't use it to save an alien species, though we will have a more enjoyable time playing Grand Theft Auto VII.
Please support posts like these by buying Making a Metaverse That Matters!
> Since the install base for high quality VR remains so small, and doesn't grow substantially even as the tech improves, there's no reason to think there's a large market for the full VR technology we see in the TV series.
There certainly isn't a large consumer market for "high quality VR" at the current price. I don't know how the market for "high quality VR" for professionals (e.g. flight simulators or surgery training) has developed in the last 10 years; you probably know more about that than I do. In any case, all kinds of "high quality VR" will continue to improve and the price will continue to go down. And then there is always this rule of thumb (often attributed to Dornbusch):
> In economics, things take longer to happen than you think they will, and then they happen faster than you thought they could.
Posted by: Martin K. | Saturday, April 20, 2024 at 01:58 AM