Yuval, also known at "VRguy" has been doing a series of podcasts with various industry experts and I feel privileged to have had such a wide-ranging conversation with him.
Here's the full transcript with some minor edits for clarity.
Yuval Boger (VRguy): Hello Jason and welcome to the program.
Jason Perlow: Hi, Yuval. How are you doing?
VRguy: Good. Who are you and what are you do?
Jason: I'm Senior Technology Editor of ZDNet and I've been blogging and writing about the technology space for over 20 years.
VRguy: Excellent, so we're here to talk about virtual reality and I saw a piece that you wrote a little while ago with some concerns about social impacts of virtual reality. Could you explain what you meant in that piece?
Jason: I believe that we don't have enough data, as far as how the brain reacts to stimuli in a virtual reality or augmented reality form. Particularly we don't know, there probably has not been enough study to see how it affects the brain in developing minds, such as children, or how it affects adults that may have addictive personalities in general.
We do know how addictive personalities respond to other technologies, such as smart phones, texting, even video games. In fact, there have been studies that have been done in South Korea, which is one of the most technologically advanced countries in the world, where they've discovered there are very high rates of addiction to video games from not just young people, but also from adults, where people will sit in immersive video games, whether it's on PCs or on consoles for hours and hours at a time, neglecting their life.
This is, of course, video games, which are not nearly as immersive experiences as virtual reality.
VRguy: Okay, so the concern is sort of a super-sized effect on people when using virtual reality, so what do you propose we do, other than not use VR devices?
Jason: I think VR devices, there's going to be different types of applications. In the vertical space, we know that virtual reality and augmented reality is extremely powerful and extremely useful. Whether it's medical imaging, if it's engineering, military, law enforcement, any of those things where we have a specific application in mind, I think augmented reality and virtual reality is great.
I think we need to be very careful how we use virtual reality in recreational settings and by recreation, I mean as a primary form of entertainment. I think there are people that will be able to very easily detach themselves and to regulate their usage of virtual reality on a recreational basis, but I think we're probably going to need some form of, I would say, reminder or some type of check built into the software that's being used to inform the user that maybe you should be taking a break.
That's by the way, the kind of thing that we're starting to introduce into smartphones right now. There are apps that you can download that tell you maybe it's time to stop texting or maybe it's time to stop browsing or whatever it is. Just because of the way that the brain responds to this type of stimuli in general. This feels also a lot more real than video games. Not just from an addiction perspective, but virtual reality affects centers of the brain that are completely different than in how we respond to video games.
There's more of a sense of you are experiencing the real thing. For example, if you've seen any of the third person videos that have shown people responding to some of the Oculus demos, such as the roller coaster one. I mean, people gasp and things like that. They literally feel they're experiencing it. The same could be said for the type of games that are violent, such as a simulation of a combat situation. We don't know how it's affecting people that might otherwise respond to traumatic situations differently from others.
We just don't know enough information yet, so we need to be a lot more conscious about how this technology is used, I think.
VRguy: There could be many effects, right? One is the addictive effect. The other one is I could become or someone would become seasick just, vestibular mismatch. You could run into your sofa and hurt yourself just like people were hurling the Wii remotes at the TV and getting hurt this way. You could lose balance and maybe there's even epilepsy, right? Certain individual stimuli.
Jason: Potentially, yes.
VRguy: Is the solution an age limit? Is it a rating scale for sort of a level of immersion? Is it just gradual getting used to VR or just simply let's watch and see what happens over the next year and then adjust?
Jason: Again, we don't have a lot of precedent for a technology like this. We've seen what has happened when you introduce smartphone technology to very young people. I'm sure you've -- I don't know if you have kids or if you know people with kids, but if you've seen kids use smartphones in social situations, you go out to dinner and you'll see some kids playing with their smartphone at the table and they're not engaging with other people. That type of behavior becomes highly reinforced. It's very difficult at that point for them to remove themselves from that technology.
Certainly, I've gone on vacation and I've seen children on cruises and on islands in the Caribbean where they could otherwise be having a good time sitting on the beach and instead they'll be looking for Wi-Fi. Look to do texting. They have no other way to have a good time unless they're using their cell phone. That's just one example. It's very hard to say when you start getting people used to virtual reality, especially if they grow up with the technology, what happens if they're withdrawn from that technology.
I'm 46 years old, I'm Gen X, we grew up in a time when the technology was being developed but we also remember a time when we didn't have this technology, so I can just as easily put my smart phone or tablet down and walk away for a couple of hours and have a perfectly good time. I also can get heavily immersed in it for several hours at a time, but I can survive without it.
We don't know what's going to happen to these kids with smartphones and tablets if they become detached from it. The same could be said for virtual reality. The technology is advancing so rapidly, we don't know how exposed people are going to be to it over the next period of time. Right now a headset costs, for Oculus, I believe they're in the thousand dollar range.
As like anyone else, I remember when smartphones cost a thousand dollars apiece or when PCs cost 4, 5,000 dollars apiece. Now PCs cost, you can buy one for 3, 4, 500 dollars. A fairly decent PC, a laptop. It's common technology. Smart phones are getting down to 99 and 200 dollars. Everybody's going to have one. Currently, the VR technology is something that only fairly wealthy people can afford. The prices are coming down. Within five years, it'll probably be down to half or less than that.
As the technology proliferates, it's going to have greater, greater impact on our society and we just need to understand, what are the risks that we're taking on?
VRguy: Of course, what are the benefits, right, because there are plenty of educational benefits, therapeutic benefits, so we just need to be cognizant of the addiction side. Switching gears, I know that the other thing I saw you write about is Linux and basically non-Windows environments. Do you see that becoming significant market for VR? What's the use case for VR on a Linux machine, for instance?
Jason: I think the big thing about Linux is that obviously, it's open source. A lot of the tools are open source. All of the projects are open source. It's an open environment where basically you can sort of roll your own tool sets. Of course, you have all these communities that are kind of centered around open source development. I think ideally an OS like Android is going to be a popular environment for using virtual reality hardware, in general.
That's not to say that Microsoft stuff isn't useful, certainly there are a lot of tools that, even Microsoft is doing on the open source side, so I don't think open source is necessarily endemic or mutually exclusive to Linux. I certainly think that there's a tradition with open source and Linux. It'll always be probably the strongest open source environment for programming and developing these type of tools an environments.
Certainly, there's going to be a lot of stuff coming out from Microsoft, especially if you look at Hololens and it being an integrated hardware and software environment, as well. I think there's room for all these communities, all these software development models and some may be better suited to commercial, military, aerospace, and some may be better suited toward entertainment and video games and things like that. Just as those things are suited to those types of environments and communities today on PCs and tablets.
VRguy: I guess the other point is that, if I had to choose a real-time operating system, Windows might not have been my first choice, right? A lot of VR interaction is time sensitive.
VRguy: If you want to control latency and you want to make sure you render on time and so on, you usually want to have sort of deeper and deeper insight into what's going on under the hood, so perhaps an open source operating system gives you the feeling that you have more control over what's going on.
Jason: If you're going to integrate the micro controller for the entire operating environment within the headset itself, I'd probably tend to agree with that, but if you're talking about... It's interesting when you think about real-time operating systems and what our preconceptions are about Unix and Linux versus Windows. A lot of that stuff is really changing now because you start thinking about containerization and hypervisors and things like that which have gone a lot thinner now than you used to be.
Developers don't necessarily need direct access to the hardware like they used to, especially when you can modularize components of operating systems and applications much easier than you used to be able to do. Those RTOS rules may not still apply, especially when the hardware's gotten more and more powerful.
For example, we look at what an ARM processor can do now. Just looking at something like an iPhone, A9, A9X processor today and what that's going to look like 3, 4, 5 years from now when we start to see the inexpensive ARM-type processors coming out of Mediatek in China or whatever see coming out of Qualcomm or Samsung in the next few years. With the number of cores and the hyper threading capability that's in them, I think the base operating system is going to be less important than just the development environment that you can run on top of it.
VRguy: On the subject of the architecture, today a lot of VR is local in the sense of local to the machine. Do you see that evolving and getting more and more stuff pushed into the cloud for multi-player or even some people are talking about doing rendering in a distributed fashion?
Jason: Sure. I think that's definitely something that's going to be realized, I think within the next 3, 4, 5 years. There's no reason why you can't do it. One thing I've seen, at least as it pertains to cloud architectures is there's a challenge in putting GPUs in the cloud, simply from a price/subscriber standpoint. You see providers like Azure putting GPUs in the cloud today for computational tasks, and it's easier for them to put up a cluster of those things and then to allocate compute time and provide costing for that.
When you use GPUs literally for rendering type of applications, right now the GPUs tend to be very expensive. It's hard to virtualize those cores in such a way so that you have better density than a one user to one core relationship right now. There are certain things that are being done by NVIDIA and companies like that and AMD that to improve that, I think that we're going to see some improvements along those lines over the next year or so with these public clouds as they work tighter relationships with companies like NVIDIA and AMD and figure out how you can get better density on those GPUs in the same way that, it was a challenge with just regular CPU-based virtualization in the last 10 years to get core to process density higher which was the trick in getting the cost of cloud computing down in general.
Once that hurdle is overcome, I think we'll be able to do cloud-based rendering, not just for VR but also for things like virtual desktops and being able to run things like Adobe Photoshop in the cloud without having to do any kind of localized processing at the client.
VRguy: In VR, there's the issue of near field and far field. Near field, if I need to see my hand in my personal view, that is probably going to be rendered locally.
VRguy: If I want to see sort of the skybox, the background with multiple participants, that perhaps could be rendered in the cloud if compression is good enough and if bandwidth is fast enough.
Jason: Yeah, then you'd have some type of compositing software that would be able to do that. I think the challenge is really, as you said, with bandwidth, especially in the United States, we have such a large variance in bandwidth in different metropolitan areas and the last mile is still a problem.
For example, where I live in South Florida, in the community of maybe 200 houses that I live in, the fastest I can get is about 50 megabits, 100 megabits a second. Very frequently, in the evenings when people are using their internet for entertainment purposes rather than business, I'll be lucky to get 20 megabits per second download and then like a 5 megabits per second uplink. Whereas in the daytime, during business hours, I can get pretty much my full 100 megabits down and 10 megabits up.
Things like compression, quality of service, all those things are going to have to be taken into account with our infrastructure and probably we are going to have to look at whether or not some of these Wi-Fi technologies, such as 802.11ax or some of these metropolitan Wi-Fi standards for like 10 gigabit access need to be built out more to handle this sort of thing. Certainly, if there is an increased demand for 4K video content from things like Netflix and Amazon, we're going to need to see substantial infrastructure improvements, let alone these things for VR, which we know are going to have more, not just straight up streaming of a video, but interactive types of protocol activity.
VRguy: Reducing the bandwidth is maybe the way to combat addiction.
Jason: Certainly, if you don't have the bandwidth, you're not going to be able to watch, you're not going to be able to play your game.
VRguy: Jason, I thank you, you get the award for the most wide-ranging interview in a while. We started with addiction and we ended with cloud-based rendering, so this has been great. Thank you very much. Where could people connect with you online?
VRguy: Excellent. Thank you so much for coming onto the program.
Jason: Thanks, Yuval.