Stanford professor Jeremy Bailenson has been thinking about virtual reality and the metaverse for decades. As of 2020, he even teaches in it (more on that in a moment).
For all of the chatter from Facebook/Meta, Nvidia, and other companies about building the metaverse, though, he thinks the metaverse will be mostly empty. That is to say, there won’t necessarily be a lot of things to do in this immersive version of the internet.
While social experiences and games could come to define the space, Bailenson, who founded Stanford University’s Virtual Human Interaction Lab, is betting that education and work will remain the “killer apps” of virtual reality (VR) in the years to come. Fittingly, the VR sports training company he co-founded, Strivr, has since shifted its focus to business training broadly.
Bailenson recently spoke with Quartz about what the metaverse is, the state of metaverse technology, and why the developers of the metaverse can be conscientious about their carbon footprint. The interview has been lightly edited for length and clarity.
Quartz: What is the metaverse, exactly?
Bailenson: The term “metaverse” comes from Neal Stephenson’s 1996 book Snow Crash. Stephenson defines the metaverse as basically the internet, but immersive. Imagine the internet skipped the 2D version and went right into VR. That’s the notion of the metaverse.
Is VR technology ready for this?
In March 2020, when covid-19 hit, Stanford asked professors to volunteer to move their normal teaching load to summer 2021. I volunteered because I’ve been teaching a class since 2003 called Virtual People and I wanted to, given that we were remote anyway, try to do it in immersive VR in the metaverse. So in June 2021, 101 Stanford students all had their own headsets at home and we networked via avatars in the metaverse using a platform called Engage.
We talked and we learned and we experimented and we traveled. We spent about 60,000 shared minutes inside virtual reality and did all the things that Stephenson wrote about way back then. This fall, I’ve got 178 Stanford students and we are doing incredible things. We’re building things in the metaverse. We are having small group discussions. We’re doing travel and meditation and medical classes. I don’t think the hardware was ready nine or ten months ago. In the last six or seven months, the hardware has gotten good enough where I can do the class I’ve always dreamed of since the late ’90s.
How does the metaverse change what’s possible for teaching?
We’re learning a ton about how to teach and learn in VR. Our class was a magnitude of order larger than anything anyone has ever tried. When you combine the two classes, you’ve got 250 students-plus in VR with a group for 10 weeks in a row. The lessons that we’ve learned ranged from what size group is the best for small-group discussion to how many avatars can you render in the same scene before the whole system crashes.
More importantly, we’ve been developing a curriculum that leverages what the metaverse is [and isn’t] good for. Once a week, we do kind of a lecture where we talk about readings and we have guest lecturers come in and we do that over Zoom. If you’re looking at someone talking, you don’t need to be in VR, right? One reason why I want my smaller discussions to be in VR is that it preserves the spatial coherence of the conversation. And I’m very strict—I don’t want to get anyone dizzy ever. We have short bursts with 30- to 40-minute experiences.
You think work and enterprise are going to be huge parts of the metaverse. Why are they the main use cases?
When you go through the history of VR, it’s all about training, starting with the Flight Simulator in 1929. It has been a killer app of VR since there’s been VR and that remains the case today. Video games are doing okay. And you’re getting thousands of people per day going to places like AltspaceVR or VRChat, but not hundreds of thousands.
Ultimately, what’s going to drive VR is that it’s really good for training. What Strivr does is we put you in this incredibly immersive scene. We began as a football training, training quarterbacks and other players, and then we went to other sports—US Olympic skiing, NBA free-throw shooting. In 2016, we pivoted to enterprise, and so the largest client remains Walmart. Currently, we do lots of things where there are other people around, but those people are either controlled by A.I. or they’re recorded people that we’re beaming in via video capture. What the metaverse is going to do is that, with all the Strivr training scenarios, we’re now going to be able to do it in teams.
One of the most successful trainings for Strivr is active shooter training. The CEO of Walmart has publicly discussed this. [In 2019, a gunman shot and killed 22 people and wounded 23 others at a Walmart in El Paso, Texas.] Many of the employees were working that had already trained with Strivr’s active shooter preparation tool. And so they were prepared in a way that they wouldn’t have been had they not done that training. It’s one of the most incredible success stories of VR, as horrible as that day was. The CEO says that lives are saved because decisions are made faster because of the practice from employees who use VR.
Google Glass went from a failed consumer product to more of a workplace apparatus. Is that what we could see with the metaverse? And does that point to bigger consumer reluctance to use VR and AR?
I’ve dedicated my career to VR—I’ve done nothing but think about it during my work life since the late ’90s. That being said, I don’t use VR recreationally. It’s not something that you do for fun yet. VR has always been about solving hard problems. You’re putting something on your head. You can’t see the real world. You can only do it for a certain amount of time because it is different, perceptually, than the real world.
What Google Glass and others learned the hard way is that if you’re not solving a problem, people don’t really want to be having these wearables on their faces. The problem with Google Glass is that there wasn’t really much to do there. It was a small visual field. There was no way to track your input properly and it just didn’t actually do that much. VR eventually is going to be super fun for the consumer, but until we get there, it’s going to continue to do what it’s always done, which is solve really hard problems.
Do you think that the pandemic has sped up our path to the metaverse while breaking down important barriers around remote work?
I certainly think it’s broken down barriers to a ton of areas of remote work and about, you know, the need to have a physical handshake in order to seal the deal. On the other hand, the pandemic caused delays in production channels that led to made it harder to get hardware. And the pandemic caused the economy to struggle, so a lot of startups didn’t make it through.
Also, VR is still fairly new—the way that a lot of people get exposed to it is that they go to someone’s place of work or their lab or their house, and they try on VR and say, “Oh, this is awesome.” It kind of spreads that way. And that was lost during the pandemic. So I say there are two sides of the coin, but I agree with you that the loss of the stigma of remote work is a huge one.
What do you think about social media companies like Facebook becoming interested in building the metaverse?
What you’re going to see in VR is no different than what you’re seeing on different social media. There are dozens and dozens of platforms. Some of them are designed for small-group interactions. Some are good for large-group interaction. Some of them are designed to be places where you’re respectful to one another and a place for work. Some of them are designed specifically to be the Wild Wild West where anything goes.
Every company wants to be a platform. And they want to be the platform.
A lot of us in the field of VR are very surprised that of the top 10 head-mounted displays that exist, two of the more portable ones, the Oculus Quest 2 and the Pico Neo 2, both have been bought by social media companies—Oculus by Facebook and now TikTok buying Pico. It’s very strange for veterans and pioneers in the field to think about why is a piece of hardware tied to an account. It was never that way before. VR was always hardware. It wasn’t part of a personal account.
I work with a lot of different tech companies in this area. And one of my jobs in working with them is to jump up and down all the time about policy issues and things like privacy. So we’ve got a lot of work to do on that front. But that’s just where we are.
The tension is about when users should be anonymous and when they should be identified, right?
There are about a dozen prominent VR social platforms. The ones we’re using from my class mostly are Engage, which is a small company out of Ireland, and AltspaceVR, which is a startup that I worked with early on, and then Microsoft bought them. Neither of them requires you to have a social media account, but you do have to make an Engage account and an Altspace account. So I don’t think any of them are going to let you just go in there without having a username and password. That certainly hasn’t been the case yet.
Now there is a growing movement around WebVR, which is a more flexible way of entering VR. In the VR scene, there are those that want pure bottom-up stuff that’s not top-down from the big companies and you’re seeing some energy there.
Is there an environmentally responsible way to build the metaverse?
I love that question. I’ve got two answers for you in terms of how to design it. Well, one’s going to be obvious, which is that we should use servers that are green in order to power the metaverse.
Here’s one that’s not too obvious: In the early days of Second Life, if you had some kind of interactive algorithm such that a sun was going around a scene and casting a shadow, it didn’t matter if there were 200 avatars visiting or if no one had visited it for four months—the processing in order to make a sun go around was happening. In other words, the tree in the woods was constantly falling, even if nobody was there.
That sounds like a simple thing [to stop doing], but if you’re going to have a persistent world that’s the same for everyone, it’s actually hard to solve that. The metaverse needs to be the same anytime anyone comes in. And one of the design principles to be green is that we need to make sure that there aren’t things happening when nobody’s there. You don’t need the tree falling if nobody is there.