Can Standards & A Spatial Web Browser Deliver The Metaverse We Were Promised Years Ago? Neil Trevett, Khronos Group, Metaverse Standards Forum, NVIDIA

Hello, and welcome to The Tech Glow Up live from Augmented World Expo 2026. Today, I'm talking with Neil Trevett, president of the Khronos Group and the Metaverse Standards Forum. I literally got goosebumps saying that. Neil, you don't know this, but in my first professional role in augmented reality-

Uh-huh

we did a lot of events, and Khronos sponsored those events and was in a large campaign around glTF.

Yes.

And I grabbed a shirt, and my colleague and I vowed that we were ride or die glTF fans- from that moment. And I know that they brought that shirt with them- to this conference. So, like, most of my career in XR, like, your work has literally been in the room with me on my journey.

Aw!

So this is a total treat to have you on The Tech Glow Up. Well,

thank you for saying that. It makes it all worthwhile. Thank you. Well,

For those who maybe aren't as nerdy as I am about 3D standards, and file formats, and interoperability-

Yep.

Can you introduce yourself and the many cool things that you do in spatial innovation?

Sure. My day job is at NVIDIA- Okay uh, where I help develop developer ecosystems, which is helping developers get to GPUs.

Mm-hmm.

And a large part of that ecosystem that I particularly get involved with is open interoperability standards. Mm-hmm. So that's both APIs and file formats that anyone in the industry can implement. And so developers, if they use open standards to develop their, their assets and their content, or they program using open standard APIs, you know, their content and their applications run across m- multiple vendors. Mm-hmm. So it gives content developers bigger reach. And content is king, so everyone benefits, so. And some of the, um... You may know us for standards at Khronos- Mm-hmm such as, um- Vulkan, uh, which is the GPU programming API, the modern GPU programming API that's cross-platform. Uh, OpenXR, which is, like, particularly hot here at-

Some really interesting examples out there.

Yes.

Yeah.

Yeah. And that's for accessing the hardware on XR devices, and it's pretty much, not on every device, but the vast majority of devices now support, uh, OpenXR. And glTF- Mm which is the 3D asset format. And glTF is on an interesting journey. Now, it started out by being a 3D asset format, so you can grab it and display it.

Mm-hmm.

And with Gaussian splats now, which is coming, a hot term of, of, uh, AI- You, you're hitting on

one of my special interests. N-

now glTF has Gaussian splats. It's increasingly, of course, being used for scene acquisition- Mm uh, and, and display. And I think the next step with AI coming online is figuring out how to put scene semantic information into glTF so you can capture the meaning of glTF scenes- Mm-hmm as well as how they look.

Oh, my goodness. You literally hit three topics that were not on our agenda today- that I would die to talk with Neil Trevett about. Oh, my gosh. but I'm gonna keep us to the script because- I love to introduce people, and one of the things that I look to do on the Tech Glow Up is learn from OGs who have been there and seen it and done the hard work to, like, m- make great products, and in your case, like, make standards orgs and global agreement about- Mm-hmm really complicated things. I see creators and developers as, like, a keystone to innovation because they are the ones who translate what is possible and what, like, users need and what a great experience is, and, like, they find that sweet spot-

Yes.

That, like, makes applications and content and stories that people actually wanna use. Absolutely.

Yes.

I'm so interested to get into, like, how you are supporting developers with standards and, like, all of this interoperability. But, like, before we get into that, what is the origin story of your passion for supporting developers and hardware and the very, like, hard process of, like, building and maintaining standards?

Yeah. So, I can actually trace it back to, like, a origin event. So, ever since leaving college, I've been involved in 3D graphics. I was initially in a, a startup company in the UK called, you know, 3D Labs. As we were developing silicon to do 3D graphics processing, the industry was kinda searching for what the API would be, and a lot of the silicon vendors were creating proprietary APIs, which was creating a lot of problems for the developers. You'd have to, kind of rewriting- Mm. Your application, uh, many, many times. And then Silicon Graphics, uh, I don't know if you remember Silicon Graphics, of course. Mm-hmm. You know, kind of the, the granddaddy organization of 3D in many ways. They took a very, uh, I think, enlightened and bold decision. They took their proprietary library, which was called IrisGL- Mm-hmm back in the day, and they made it an open standard, OpenGL, and that was controlled by the OpenGL Architecture Review Board. So it was a number of companies, uh, came together to, uh, cooperatively create an interoperability standard that would ship on everyone's hardware, and that was back, I think back in '92.

Mm-hmm.

And I saw firsthand the impact of- How that brought the industry together and how it enabled developers to develop once and deploy on many different, uh, pieces of hardware, and I was hooked. I was hooked. And, uh, I managed... I was in- involved at 3D Labs, but then eventually, you know, Silicon Graphics went away, and the Khronos Group kind of did a diving catch and caught OpenGL and, and made sure that it survives to this day. And OpenGL ES, which brought 3D to mobile phones-

Mm-hmm

uh, was a derivative of- That of OpenGL.

Yeah.

And then WebGL, which brought 3D graphics to the web, was another Khronos standard. Mm-hmm. And that was derived from OpenGL ES, so this b- kind this long history. And then, and now of course Vulkan is essentially the modern version of OpenGL, so.

standards as a product process effectively, right? Like building, updating, I love it. So Neil, we're here at AWE Yes Where I've been talking with a- as many smart glasses manufacturers and brands as possible. And something that you just said really stands out, in that right now, with some exceptions, most smart glasses are developed in their own programming tools, in their own content distribution systems, in their own kind of walled gardens.

Mm-hmm.

And part of that is because the hardware's all different, and part of that is because what they're trying to do is different. But it sounds a lot like this origin story moment of we got a lot of silos, we got a lot of things we're trying to do, and wouldn't it be easy if you didn't have to make it six times?

Yes. And the, the standard that the Khronos Group is working on that is the exact analogy to- Mm-hmm kind of OpenGL for th- for the driving the GPU. Yeah,

yeah. You

know, for driving the displays and the controllers and all the interactions that you have with an XR device. That standard is OpenXR.

Mm-hmm.

And that's actually quite pervasive now. So but OpenXR, you know, is quite low down in the stack. It really- Mm-hmm it's like, like Vulkan, it's really close to the hardware. You know, the Khronos tagline is, "We connect software to silicon," you know? Yeah. So that's, that's kind of where we are in the stack typically.

That's very... Yeah.

Yeah. So, um, OpenXR, h- I think we were fortunate, 'cause making s- standards to get adopted, there's an element- Mm of luck in that you, you n- not only do you need, like, a well-designed spec, you need to do it at the right time. Mm-hmm. Not too early, 'cause if you do research by standards committee, that's death by a thousand cuts. Don't do that. We've learned the hard way, don't do that. Yeah. Uh, but if you're too late, the fragmentation has already set in. So the, there is a window, and we were fortunate, I think. We, we, we did, um, hit the window with OpenXR. So, uh, it's shipping on a bunch of XR devices. So developers, we avoided the, the, the fragmentation phase-

Yes w-

with XR devices. That, at that level.

Yep.

And, and OpenXR continues to evolve. It's also evolving from just the hardware API now with kind of things like spatial entities. Mm-hmm. So bringing in kind of scene understanding and, you know, scene-

That's amazing

um, you know, um, scanning and transmitting that kind of semantic information back to the application. So that's, that's its kind of evolutionary-

Mm-hmm

path. Um, but there is, there is much more that we can do. Um, now we need to th- think about higher levels in the stack. and we really need the, I think, the spatial evolution of the web.

Mm. And this is something that you've been working on The committees have been in process, and just this week you made some really big announcements-

Yes

about the direction of the spatial web. Yes. Can you, uh, can you share a little bit about what th- a spatial web is, what an open spatial web is, and, uh, your announcements, uh- Yes this week?

So, so the announcement is, um, at the Metaverse Standards Forum. Yes. So maybe just a little bit of context about the Metaverse Standards Forum- Perfect. Absolutely 'cause the, um, Khronos is a standards organization, and it's, of course, it's not the only standards organization. Th- there's many standards organizations that, like the World Wide Web Consortium, the W3C. Mm-hmm. There is the Open Geospatial Consortium that does all the geolocation, uh, type stuff, and dozens and dozens more. You know, wireless standards, privacy standards, and y- you name it, there's, there's a standard for it. what we found when the metaverse first became a thing a few years ago- we quickly realized that no one standards organization could, could bring all of the pieces together that we were gonna need to build an interoperable open standards-based metaverse. And so the Metaverse Standards Forum was specifically set up to let all of the standards organizations come and cooperate, and to engage with industry to figure out how all the pieces are gonna come together to, you know, create this thing we call the metaverse. Mm-hmm. Now, the first question, of course, as soon as people hear the M word is, what is the metaverse? Yeah. Great question. And I think- the, the metaverse has gone through some ups and downs in terms of its, its public image.

Mm-hmm.

Often when I talk metaverse, you know, people laugh. Mm-hmm. And I think it is actually, it goes back to the first en- envisioned form of the metaverse a few years ago. It actually took a wrong path, and the path was we're all going to be in virtual reality for 16 hours a day. Mm-hmm. Like, we're all gonna live in virtual spaces, but that means we're gonna be disconnected from our friends and family, you know? And that is actually quite a dystopian view of what the metaverse could be, and I think society in general got rightly rejected that- Mm-hmm dystopian view. But now I think with smart glasses, that use augmented reality rather than virtual reality, augmented reality brings us into the world rather than isolates us from the world. So my definition of the metaverse is you're using XR- Mm-hmm you know, ex- OpenXR for the sensors and the, the bits and bytes.

Mm-hmm.

And then if you let your devices understand the environment, that's spatial computing. Mm-hmm. And if you have multiple users using things like smart glasses, augmented reality bringing them into the real world, but they are connecting together as users-

Mm-hmm

then that is the metaverse. So a, a metaverse is a connection of multiple people doing spatial computing on their smart glasses with the help of AI. Mm. And I think that we have a chance to rehabilitate the M-word. The, the metaverse doesn't need to be a joke anymore.

Oh, my

gosh. Um, we can actually make it, you know, the, the, this thing. And- If we're going to realize that vision of connecting anyone who wants to connect to the metaverse and all the spatial services that are gonna be available, then we need a platform to do it. Mm-hmm. And the only platform that possibly has the sufficient reach-

Mm-hmm

and is, and is totally based on open standards is the web. And so our mission is to bring spatial, true spatial computing capabilities to the web.

Mm-hmm. Oh my goodness. So many things. I joke, Neil, that I've been in XR so much that, like, my least favorite conversation is the definitions, like continuum. but I actually really love, this posit that the metaverse is people basically connected to metadata about the world- Yes in their own personal compute devices that are very likely wearable. Yes. Right? Exactly. And like- Exactly that's way less dystopian than, "Hey, let's hang out in gigantic concert spaces in VR where there's no one there because you missed the event."

Right. Right. We, um- And but, but to be clear, I mean- Yeah VR is- Yeah is cool. I'm not saying VR is bad. Yeah, yeah. And VR definitely has awesome use cases.

Absolutely.

Um, uh, you know, and- No,

no,

no but, but it's just not the thing we're gonna be wearing 16 hours a day.

No. You

know? Right.

Uh, not at this weight, right? Like, the, the, the rally toward 49 grams I think is, like, a, a real earnest pursuit- Yes that, like, this industry should really be-

Yes,

yes focused on.

Yes. So the- Yes the next step in this whole thread we're pulling on here is, okay, well, spatial web, that sounds cool. Mm-hmm. But what's the difference between spatial webbing-

Yes

and today's There's a couple things. On the 2D web today, you know, you typically go to s- somewhere on the web by typing in a URL.

Mm-hmm.

You know, or clicking, you know, clicking the URL chain.

Yeah.

In, in the spatial web using your wearables, you're gonna be literally walking around.

Mm-hmm.

And you're gonna want to be discovering services, by your location. and that's a, that's such a fundamental difference. Mm-hmm. Then the next difference is you are in- the real world, not staring at a 2D screen. And so very often you want the services, and it's by service, and what do you mean by service? Mm-hmm. Right. It can, it could be anything. Mm-hmm. It can be, you know, a, a menu that pops up when you w- wander into a coffee shop, right? Mm-hmm. It, it can be the navigation path to your gate at an airport when, you know, you, you're late for your flight and you're trying to select your seats- Mm-hmm as, as you go, you know? Oh, goodness. Can I get that upgrade? Yeah. Uh, it's any information very often graphically represented-

Mm-hmm

in the space around you. And because you are in the space and things are being overlaid the real world, and you have all the space to place information, you have the chance to bring in multiple services at the same time. So-

Mm.

You know, some of the neat pictures we have for the OMVI launch is, you know, you're navigating to a particular place and your Uber arrives, you know, and so that's the navigation app and the, the, the rideshare app.

Mm-hmm.

You know, two different services, completely different companies in your view at the same time, in your little infosphere.

I've encountered this idea, and I'm, I've been loving this idea of a spatial interface isn't locked to, like, a screen. Mm-hmm. And so you can... And if it's, like, relational to you as a user and what you're doing, right, having multiple different services, like providers, softwares in the world is, like, a lovely idea. It's the first time this idea of incredibly personalized, like, contextually aware, software and service has... Like, I've heard that idea-

Mm-hmm

in the last, like, three years without it being directly tied to an idea of an agent. Right. And this is, this is literally not focused on a technology, but what is a user doing as they move about their day-

Mm-hmm

trying to accomplish tasks.

Yeah.

That feels really good.

It, it, it, it is good.

It's centered on the right things.

Yes. And, you know, AI-

Mm-hmm

is gonna be- in this mix. Yeah. Absolutely. both on the, I call it, both on the input side and on the output side.

Yeah.

On the input side, it's gonna be kind of under the covers, you know, l- looking at the new Snapdragon announcements, you know, the, that lot of emphasis on how fast they can do inferencing on device.

Mm-hmm. Yeah.

Because that's gonna be the engine- That context that, that lets you understand where you are and what, what you're doing- Mm-hmm um, and your gestures and everything. I call it the magic pic- pixie dust, because it's really enabling stuff that wasn't possible three years ago. Mm-hmm. Right? It's kind of magic.

Oh.

Kind of magic. So, but, but then on, on the output side, agents, yes. Mm-hmm. You know, so your own private agent that understands what you're trying to do today, your context, and can help you navigate all your devices and, you know, your interactions- Mm-hmm with, with people and the, the, the, the scene around you. But we need all this in the web. And I know I love the web.

Mm-hmm.

And so we're not replacing the web. What we're trying to do is to find the minimum amount of work we can do to bring spatial computing into the existing web.

Mm-hmm.

And the, the, the key thing that... And, and shout out to RP1, you know, because they've been, building this kind of concept as a kind of- Mm-hmm a prototype for several years. And, you know, they are the ones that have found the, kind of the bottlenecks. The current way of doing immersive, web is WebXR.

Mm-hmm.

And I'm sure many people l- love WebXR. I love WebXR. And WebXR is not going away.

No.

You know, and for many types... So if you're doing a, an application and you want programming control, and you want just to do everything, WebXR is perfect.

Mm-hmm.

And that, that will continue. But this extra new mode where you're wandering through the world and you want to have multiple services being composited- Mm-hmm into your scene from multiple different companies, that's not possible with WebXR in the general sense. You could imagine- Mm-hmm you know, someone coding a very nice WebXR app. and with permission, it brings data from one company and data from another company, and it puts it all together in kind of like an infosphere- Mm-hmm in a very nice way. But that means that the data from those two companies is being shared with the application developer or the platform in the middle. And if you talk to enterprises-

Mm-hmm

who want to use, spatial computing to engage both their internal employees and their ex- external customers, they do not want to share all their proprietary data. So we're blocked at step one, and that's why I think you're not seeing a lot of enterprise deployment of spatial computing yet.

Mm-hmm.

There's simply no way to do this without being locked into a proprietary platform. So being locked into a proprietary platform is bad, and being locked into not being able to do this composition is bad. So the way to solve it is to bring composition into the browser. So the browser does all the necessary composition. You don't have to go through any third-party intermediary. And that is what we're doing with the Open Metaverse Browser Initiative.

Oh my gosh. My brain is so aflutter with this idea that the metaverse was, like, omnipresent and, like, imminent, And so much of the conversation was on technical capabilities and, like, platform level APIs, and, like, having spatial kits. And this data bottleneck that allows for a user-first spatial, understanding of, like, what somebody's trying to do, and to bring in, we're, two and three-screen humans now. Mm-hmm. Partly because, like, I need the big screen for certain things- Mm-hmm the pocket screen for others, and the mediums. Mm-hmm. Right? And so, um- Neil, I'm having too much fun with you. We're, we're running out of time. Uh, I just absolutely love, and there's, there's some- there's this gorgeous, like, elegance, uh, to the way that you talk about approaching problems. Like, you're like, "Hey, we have this 2D web, but, like, it needs another dimension." Right. And it's just, like, sort of taking where we are, the problems that we have, and just expanding one step further. Again, it's, like, connected so much to that product iteration- Right and growth.

Right.

And, like, solving the next most urgent-

Right

problem.

Right.

I'm curious. The show is called The Tech Glow Up.

Yeah.

And I love to ask, people what their glow up is for the industry. So, like, a glow up is a notable transformation, a rebirth of sorts. In this, like, coming age of smart glasses, personalized agents, compute, et cetera, like, what do you wanna see the industry move toward with earnestness in the next six to 12 months? Where does this need, all need to be going?

Well, I, I think the, the conversation that we've just had kind of, is kind of leads us to that- Yeah obvious kind of concl- So I think the... Well, there are a number of threads. I, I think, you know, the, the, the way that we can see to help the web upgrade to a spatial capability is not to reinvent the whole web stack, but just put a new spatial engine that sits alongside the current 2D engines that are already in the web stack. That's going to bring us the capability. We We'll be able to compose these services into a single scene graph, and that kind of opens up the whole world.

Mm-hmm.

And AI is gonna be a vital part of that, you know, as we said, both on the input and the output side. But it, it all does come together. For better or for worse, we did make the decision to call the Metaverse Standards Forum, the Metaverse Standards Forum with the M word in. So I think it's, you know, it's our responsibility to own the M word, and we haven't come up with a better term yet. Mm-hmm. Uh, but if we need to, we need to make the Metaverse be a positive term- Mm-hmm uh, that people can under- actually understand what it is and what we're trying to achieve. And the spatial in- evolution of the web, I think is with smart glasses as the primary target device- Mm-hmm is, is the way forward. And I think we can actually make really rapid progress in the next 12 months to actually making that vision come true.

I love it so much. Neil Trevett. leader of so many, uh, important initiatives in XR. I got anxious just trying to start to name them all. But president of the Khronos Group and Metaverse Standards Forum, leading the charge on the Open Metaverse Browser Initiative and advocating for spatial creators and developers by making it easy for them to build impactful tools, ship them once, and serve actual customers and real users' needs. such a treat. I've been a fan for over a decade. Thank you so much for joining me on the Tech Glow Up.

Oh, it's a pleasure.

Amazing. Yeah. Until next time. Next time. I can't wait to see- Absolutely. uh, how we're doing in 27.