1261: Using ChatGPT for Rapid Prototyping of Tilt Five AR Applications with CEO Jeri Ellsworth

I had a chance to catch up with Jeri Ellsworth, co-founder and CEO of Tilt Five again at Augmented World Expo 2023. Be sure to catch my previous conversation in episode #1021 where she gives her full epic backstory in inventing Tilt Five at Valve, getting fire, and eventually being able to re-acquire the IP rights against all odds. At the time of this recording, Tilt Five had been shipping for six months, and I get a bit of an update for how their launch has been going, and their various efforts to promote a physical face-to-face use case for tabletop AR gaming.

Ellsworth also shared with me how she used ChatGPT in order to rapidly prototype an AR demo for Tilt Five that they were showing at AWE. She’s emphasizes the benefits of being able to quickly discover the fun within AR interactions, and how ChatGPT has been revolutionizing her own rapid prototyping processes by feeding in the Tilt Five API code as an input prompt, and then asking it how a senior Unity developer would go about solving how to visualize a specific data set of live satellite data.

This is a listener-supported podcast through the Voices of VR Patreon.

Music: Fatality

Rough Transcript

[00:00:05.452] Kent Bye: The Voices of VR Podcast. Hello, my name is Kent Bye, and welcome to the Voices of VR Podcast. It's a podcast that looks at the future of spatial computing. You can support the podcast at patreon.com slash voicesofvr. So this is episode 9 of 17 of my series of looking at the intersection between XR and artificial intelligence. And today's episode is with Jerry Ellsworth, the co-founder and CEO of Tilt 5. So Tilt 5 is an amazing augmented reality tabletop experience where it uses this type of retro-reflective material where on the Tilt 5 glasses there's actually like a projector that's shooting out these light beams and then as long as you have the retro-reflective material covering your full field of view you can have like a totally immersive augmented reality experience. Really really quite innovative technology and it's quite affordable And so there are signals in into shipping the tilt 5 Units and I had a chance to catch up with Jerry Ellsworth the co-founder and CEO to get an update as to what's happening with tilt 5 But also she was using chat to BT and feeding in the tilt 5 API and starting to prototype Actually rapidly prototype all sorts of different types of applications that she was actually demoing and showing off there at augmented world expo a way to do this kind of data visualization of all these different satellites. And so yeah, quite fascinating to hear how she's been exploring different aspects of these large language models and chat to BT and the context of coding and creating different immersive experiences for tilt five. So that's what we're covering on today's episode of the voices of VR podcast. So this interview with Jerry happened on Thursday, June 1st, 2023 at the Augmented World Expo in Santa Clara, California. So with that, let's go ahead and dive right in.

[00:01:45.763] Jeri Ellsworth: I'm Geri Ellsworth. I'm the co-founder and CEO of Tilt 5. What I do in AR is I invented an optical technique that's used in the Tilt 5 headset. I actually did a lot of the electronics in it as well, and we've been shipping for about six months, and it seems like people really like it. We designed it originally to be consumer-grade, in-home, low-cost, and so that's resonating really well right now.

[00:02:12.583] Kent Bye: And I know we had a chance to talk last year at AWE, where we got the whole long version of your journey into the space. But maybe just to kind of quickly recap for folks a bit more context about your background and your journey into the space.

[00:02:25.339] Jeri Ellsworth: Yeah, I have an interesting background. Most of my life I've been an engineer. Started off building custom race cars, which was fun. And then I opened a chain of retail computer stores. When the computer store market tanked in 2000, I moved to Silicon Valley and brute-forced my way into a bunch of startups and got to work in a bunch of really interesting companies, including toy design, where I learned a lot about how to make consumer products that should and do resonate with particular audiences. Eventually I ended up working at Valve Software, so Gabe Newell hired me to run the hardware department, put it together, and so that's where I got the bug for augmented reality, because Gabe Newell gave me the mission of bringing the entire family together to play games, and as we were prototyping, AR really stood out as the way that you could get people around a table, collaborating in engaging with each other instead of being isolated and alone on a 2D screen or isolated in a VR system. And that was the genesis, the technology I developed there turned into my first startup which as a first-time co-founder of a company I made every mistake you could imagine in a startup and it didn't last too long. But learned my lessons from that and my co-founder at Tilt 5 We made it our mission not to make the same mistakes and get back to our roots of what we understand and know and how to make products. And it's been working out a lot better this time.

[00:03:52.669] Kent Bye: Awesome. Well, I know last year when we spoke, you weren't at the point of actually shipping. But now that you've been actually shipping these Tilt 5, they're not dev kits. They're actual just products. What do you call the first version of, because I know for Oculus, they had the DK1. So what's the name that you call these first iteration?

[00:04:07.976] Jeri Ellsworth: We just call it TIL-5. We didn't want to get too complicated on things. You know, someday there's going to be a Gen 2 or some other snazzy name. When we envisioned this product, we knew it was going to be a first-gen product, and we were going to make some mistakes. And so it's been exciting to get these out. So September, around that time, we started shipping a lot of units. We had done a Kickstarter prior. and had like 5,000 or so units that we had to ship out. We got that done, I think by November of last year. And then we had a big backlog from people that were excited about it. And it took us, I don't know, up until a month or two ago to get kind of caught up on the backlog. Now it's exciting that we have the logistics and manufacturing in place that we can mostly keep up with demand. I don't know, after AWE we may be backordered a tiny bit, but we'll see. But it's kind of that next phase of the business where we start to scale a little bit more, we get logistics more regular in the manufacturing side.

[00:05:12.440] Kent Bye: Yeah, from an experiential perspective, when I see the different demos here, I often see a lot of these different AR glasses. And the field of view is very small and tiny. And you have these windows where it gets cut off. But with the TELT 5 experience, it fills up my full field of view as long as the retroreflective material is all in my field of view. So there's still cutoff that happens, but it's a lot more immersive, I'd say. And it feels like that you're able to, have a really high-resolution, immersive experience that you have these wands. And so I imagine that people would be able to start to prototype some of these different interactions with something that feels like a really high-fidelity experience. And maybe sometime in the future, there might be other tech stacks that may be, let's say, more mobile, where you don't need retroreflective material to be able to do that. So I'd love to hear some of your reflections on some of the different types of use cases, or experimentations, or reasons why people are starting to play around with Tilt 5.

[00:06:07.643] Jeri Ellsworth: Yeah, when we decided to go down this path to use this special reflective material, we knew it was going to constrain the design, and we theorized at the time, we didn't know, that this would actually be a big benefit. So, if you're in XR and you start developing, it's really difficult, especially if you're trying to use something with a narrow field of view. or you're trying to insert something into the world because you don't really control your space. And so our thought was like, let's just do something where we can sell a couple different versions of these game boards that are constrained, and then the developer knows their constraints. And that led to a lot of really good things about our system. So our SDK is phenomenal. So inside Unity and Unreal, you know exactly what you're going to see, the end user is going to see, because we have gizmos in there that show you where the rendering is going to happen. It allowed us to make the SDK where everything is game board centric. So if you want to have multiple game boards, you can put multiple game boards. And each person, even though they're looking at the same game board, you can have it attached to game characters. They could be running all over the world. And you know they're going to have a great experience, tuned it for these game boards that are in Unity. And it also allowed us to, because we weren't going for the ultimate in AR of walking around anywhere in the world, we could put all of our focus into really optimizing the performance within this space. And it's turned out to be a really good trade-off. We always have thought and look at it like we're selling the holographic experience. Glasses are secondary. We don't care. So I tell our team sometimes if we could figure out a way to sprinkle magic holographic dust on the table and holograms pop out of it, that's what we're going to be doing for our next generation. Because we're retro-reflective now, we might be retro-reflective for Gen 2 or 3, who knows how long, but as soon as there's something else more interesting, or we develop something more interesting, it gives a better experience, then we'll just do that. I think it's better to get really solid experiences into users' hands, even if they're constrained in some interesting way, than try to boil the ocean and give everyone a mediocre experience. I think that's maybe what you were hinting at, There are some interesting headsets where you can do a lot of stuff like walk around outside but not on a sunny day or not if you want to be immersive. So we just kind of chucked some of that out the back door and we're like okay let's just focus on making a solid experience.

[00:08:45.600] Kent Bye: Yeah, I was roaming around the hallways here at AWE and popped in and saw you giving a presentation and heard you say in that moment that you've had a lot of companies that have been buying Tilt 5 in bulk because it's a lot cheaper than a lot of these other options that are out there, but it allows them to start to do this type of rapid prototyping and iteration in a way that is not only affordable, but also giving a really compelling experience. And like you said, you have this retroreflective material and more of these tabletop contexts. And having a tabletop game interaction is something that is like a portal into these holographic worlds that's super compelling. And there's a lot of use cases across many different industry verticals. So I'd love to hear some of your reflections on what types of things that you're seeing with as things go out there. I'm sure you don't always get a sense of what people are working on. But what are the types of use cases that you've seen people use this for?

[00:09:36.339] Jeri Ellsworth: So I obsess over our sales dashboard. And I love when I see like just some random spike, like someone just bought 12 units in an hour. Like what's going on? And I look it up and like, Oh, it's some big aerospace company. So I, one of us goes and reach out like, Hey, what are you doing? Or, you know, maybe it's, you know, some other school buys a bunch of them. Like, what are you intending to do? And what's really cool is we suspected this might happen. is if we make it really good for in-home use, that there's going to be certain applications outside of gaming that it's going to solve problems, and we're seeing that. So when we reach out to these different companies, so, why'd you buy a dozen of them? It's like, well, I could buy a dozen of them instead of a HoloLens or something. We just needed people sitting around the table collaborating on this thing that we're doing. And as I walk around AWE, it's kind of cool. We don't always identify non-gaming developers that have bought our system, like just random sales. People come up and like, hey, I'm building this thing. I'm using it as a sales tool to sell like, I'm going to invent this, like a caterpillar tractor to farmers. And it's a great tool for this because now I can sit there and show them all the features. And we've never been able to do that before. Obviously, you can't drive a caterpillar tractor out, but we can take it there and show them like the features of it. We're seeing a lot of that. We're seeing medical imaging, computer-aided design and collaboration, digital twins, scientific visualization. It's just across the board. Universities are buying these in bulk. It's great for students. And it comes back to these constraints, too. It's something like our tools are super easy for beginners to even use. So a student can pick it up and do a collaborative AR experience without having all of this other considerations that you have to deal with other systems. I can't even describe everything that we see. It just shocks me every day when there's these little verticals that are just popping up.

[00:11:40.162] Kent Bye: I know here at Augmented World Expo, I know there's Bad VR. They're doing some data visualizations. There's also UltraHaptics doing some more of a haptic integrations with Tilt 5. And when I was at Laval Virtual, there was architectural visualizations and interacting with these CAD models in a certain way. networked social environment where we're able to have someone in VR and then I'm in the tilt 5 and being able to interact with those two so yeah like a portal into these immersive worlds that then are able to have these use cases across anybody that's doing anything with a spatial context but with the demos that you're showing here is very much into that enterprise vein of doing visualizations of bodies and 3d objects and some of these other demos that you're showing here, but I wanted to ask about the gaming aspect because I That's where this all began. And so I'd love to hear now that it's been out what are some of the most compelling games or use cases that you've seen in terms of folks that have either taking existing games and ported it over into using it within Tilt 5 or something that has been created from scratch to be able to use all the new affordances of this platform.

[00:12:42.920] Jeri Ellsworth: Yeah, it's pretty cool. We get to dog food a lot of the games. I don't get to play every single game that's on our system now. They're starting to come in faster than I get to play them all. But there's some that stand out, like Headlands came out recently. It feels like Zelda. I really had a good time with that. It was a single player experience, really didn't take advantage of our multiplayer aspect. Because it was such a big experience on the table, really colorful, Zelda-like, I'm like, oh, this is really resonating with me. It may not be for everybody. There's other standouts. Our addiction around the office right now is there was a developer did a thing on a weekend. He calls it Go Monster Go. It's like Fall Guys. It was a single-player experience. And I did download that and try it out. And I'm like, oh, I feel like there's something fun here. super compelling, yeah, and I asked the developer, like, hey, do you mind just sharing that with me? And I love using our system, so I use our system on the weekends when I'm not doing business-y stuff, to actually make stuff. And so I took Go Monster Go and made it four-player, changed some of the visuals on it, so it was very close to Fall Guys, and now, like, we're always like, let's go play some Go Monster Go. Even though it only has three levels in it right now, but it's like totally feels like Fall Guys and it's when you're sitting in the round with each other and being able to look at each other and snicker at each other when you like cause them to fall off or you tip a platform and they can't get through. It's super fun. Some of the most simplistic game experiences are really heightened when you do it together. I made a really silly game just to test a feature of our system in our SDK. It was Desert Buses. Desert Bus is a famous single-player game by Penn & Teller, which is the world's most realistic bus driving simulator where you drive from Phoenix to Las Vegas, which is a straight shot at something like eight hours, and your first point is when you get to Vegas. And the reason they made this is back in the 90s, like Congress or someone was like, games are too realistic and making our kids violent. So they made the world's most realistic game as a joke. So I made a three player version of that. And we found ourselves playing that for like hours because now you had three buses and you're griefing each other and you're trying to run people off the road. And if you run off the road, it's game over and zero points to anybody. Then we weren't going to sit there for eight hours to get our first point, but It was super fun. What else is really interesting? I see a lot of people using Figman. Figman is really cool. It's got Tilt Brush built into it, a lot of ability to make little simple games and bring in physics and it's just a sandbox where you can just play and that one really spans a wide audience and I love seeing that like you could have a a young kid and someone who's never played a game before that's an adult in there just kind of playing and that's cool to see and that's the whole goal. I still believe in what Gabe wanted us to do is bring the whole family together gamers and non-gamers and to see hints of that happening is really really awesome.

[00:15:56.373] Kent Bye: And you also mentioned that you'd created a demo that's being shown here at AWE, which is an Earth that has these satellites that are moving around. And you can use the wand to turn it into a torch and start to destroy some of those different satellites. But you'd mentioned that you actually used ChatGBT to create this game, where you had fed in the SDK into ChatGBT. So maybe you could walk through that process of some of these experiments that you've been starting to do with something like ChatGBT and creating some of these different experiences.

[00:16:24.222] Jeri Ellsworth: Yeah, my background coming from toy design, like, they beat into your head. Like, identify the fun gameplay loop first and then worry about the materials or the techniques to make it happen. And so, it's kind of backwards from a lot of other engineering where sometimes you engineer it from the technical aspect up and then you have a hammer and you're trying to find the nail to hit with it. And so, I really buy into, like, find the fun and then back up from there. So, I don't mind using chat GPT. That's not going to make the most ideal code to try some things and see if it feels good. And you know, there's probably 20 things that didn't make the cut that I've done and I can do it really fast. And so when chat GPT came out, I was like, okay, this seems pretty powerful cause it can write some code. And I did some testing. It's like, this looks like legit code. And I wonder if I can constrain it and actually get good code out of it. And so, I did some experimenting feeding it pieces of the Tilt 5 SDK to like prime it and then I could ask it to do something really complicated. So it's like I feed it some of the SDK public functions and some public enums and then I can be like okay now I want to game objects that is spawned when I pull the trigger and I want a sound that's played and it'll generate perfect code. And that's perfect for me because I want to do rapid prototyping and just see if it's fun. And sometimes you look at the code, it's like, that's not the way I would have done it. Not that I'm a good programmer. I probably would have done it equally bad anyway. But I use it a lot. And so the satellite viewer actually was some utility that I was trying to put together for myself. I'm an amateur radio operator. I like amateur radio satellites. I have this automatic antenna positioner that will follow satellites across the sky when they pass over. But it's a really pain in the butt to use, because it's 2D, and you know, these weird orbits, these elliptical orbits some of these satellites have just look like weird squiggles. You can't tell if they're even in range, you can reach them. They're like, I just want to visualize it. So I went to chat GPT, primed it up on Tilt 5. I now keep it in just kind of a little Word document there, where I'm like, I pull things over, like you're going to think like a senior Unity developer. And I'm like, here is like these public enums and public functions. And I just kind of get my chat primed and it doesn't take much time. And like, okay, it's going to probably generate good till five code. Then I ask it things like, OK, I knew what the file format was from NORAD that tracks all these satellites. I'm like, do you understand what it is? And it's like, yes. And a lot of times it lies to you. And I knew one thing with like chat GPT, you can't say, like, make me a satellite tracker. It'll generate something, but it's not even close. But I asked it, like, are there libraries out there that can parse these special files? And it's like, yes, go to this GitHub. I'm like, great. So I went to GitHub and got the libraries and got them into Unity. And then I'm like, OK, now hook it up, whatever. I prompt it, and it broke. It had lots of errors and stuff in it. So I just did the same thing I do with the TIL 5 functions. I cut and paste the function I think it should use in there. I'm like, well, here's the function. Why does it not work? And then, boom, you get good code that works. So this entire satellite viewer plus the ability to operate my antenna positioner through a serial port was put together in two or three hours. I taught it how to code for this positioner. It parsed the files. I even went an extra mile of asking it to just automatically update these files because they need to be updated because these satellite orbits change over time. and it's like okay it just goes out and pulls it down from the proper website and it worked out beautifully oh and then i added this feature which is just for fun is that when you pull the trigger on the wand it makes a vacuum sound so it sounds like a home vacuum cleaner going and then as it sucks the satellites they just kind of move towards your wand and it sounds like you're sucking up pennies in a in a vacuum, just for fun. And I brought it back to work, and everyone was enamored with it. But of course, our team, when they look at anything I code, they're like, oh god, Jerry, what did you do? Because they know they have to go and clean it up. And actually, this particular one, people were like, this probably would be a good thing for AWE to show. So my co-founder went through and optimized some of the janky Chatsheep-y tea. There's actually another one down there that I put together, which is, we call it Orbit Playground, which is more kid-centric, teach you how celestial bodies interact with each other with gravity. So you can launch meteorites into the solar system so you see all the planets orbiting around the sun and then gravity will pull on them and you can get these asteroids to skim by Earth in orbit and you get like play some pretty sounds and get bonus points if you get really close to the planets without crashing into them and then it made it turned into a really nice educational tool. It gives you kind of a sense of how these different bodies interact with each other. And that was put together in almost no time. And that's really exciting. I think that's a powerful thing with AI is especially someone like me, that's an appliance operator who doesn't want to be like an amazing programmer can like stand stuff up prototypes and see like, is this a good idea before we proceed? Powerful.

[00:21:58.640] Kent Bye: That's really cool to hear. And it sounds like you also have to have maybe a certain level of expertise of being able to discern whether or not, like you said, it's good or bad code. And I'm excited to play a little bit more around with some of those more coding aspects. Eventually, at some point right now, I'm just trying to get transcripts for my podcast, which is much more a clear input or output. But have there been other types of generative AI or other AI that you've been playing around with, either personally or in the context of working with Tilt 5?

[00:22:24.708] Jeri Ellsworth: Yeah, I mean, some of our creative folks at the company are generating textures and models and they're using more of the art tools. And, you know, going back to your point, like a good architect, like I think chat GPT is a good thing for like someone that works at the architecture level. Like you kind of need to know what is good underlying technology to know. And probably the same is true for our folks using it for art. They kind of need to know like what are the mechanics of like when this thing spits out a texture, is it like, Too low-res, too high-res is mapping, whatever. I don't know all the lingo. So that is critical. I'm interested to see where it goes as far as gameplay elements when you actually have some of these AI elements right in the game. And I'm a little bit undecided. The time I spent at Valve, I enjoyed going and talking to the folks that worked on NPC AI or enemy AI for games. And I went into it naive, thinking like, hey, the smarter the AI for the enemies, the more fun the game would be. And they quickly corrected me, like, no. To set up a good gameplay loop, a lot of times the enemies are very predictable. They make the same sound when you approach them. They move in the same way. And then you can attack them or interact with them the same way. And that makes a good dopamine hit for you, which makes a good gameplay loop. too smart and always winning, or completely random, you don't get that same interaction. So I don't know where it's going to go as far as in-game interactions. I think outside of games, it's going to be super handy. We're showing off a CAD tool where you can import CAD and then interact with it. Also showing off a medical imager that can bring in these standard medical imaging files for CT scans. My mind immediately goes to, like, can I, like, take that geometry that's been loaded into the system and then use our wand to, like, point to part of, say, a 3D model for, like, industrial design? Like, I want it smoother in this area, and, like, instead of trying to text, you know, prompt this thing, actually draw. Like, no, I'm thinking more here, and then it can take the geometry, crunch it, come back, and present it to you. Is that what you want? So I think XR is going to be a very interesting interface to AI because it's not very efficient to be doing text prompts. Same thing for medical imaging. I've already heard of AI is better at diagnosing problems. So can we take a CT scan of a human body part and have it analyze it and give hints of what could be wrong with a patient?

[00:25:19.682] Kent Bye: Yeah, well, I know that we're in a bit of a liminal time right now, where the big buzz right now is that, from all the reports that I can see, that Apple's expected to be announcing their first mixed reality device next Monday. And we have a lot of movements into this mixed reality device. The Quest 3 was just announced this morning at 7.05 AM, where you have these pass-through cameras. But it's mostly in the context of, like, say, a virtual reality headset with these pass-through cameras with high-resolution color. Varyu has that. We have the Quest 3. Lenovo and other companies that are starting to have this more mixed reality Oppo was also announcing it and then as Apple comes into the fray It looks like they're potentially taking a very similar approach having up to like a 4k per eye resolution and the pass-through and so it's you know, maybe getting away from a lot of the limitations from the waveguide approach from more of the augmented reality, pure approach, although there's certainly still a lot of folks that are working on that here at AWE. But I'd love to hear any of your reflections. I know that, you know, with Tilt 5, you're taking a completely different path, both philosophically and much more constraint in its own way. But as we're in this liminal space, I'd love to hear any of your thoughts and reflections of some of these other trends that you're seeing with mixed reality, and if you have any reflections on what Apple might be doing and how that may be changing the ecosystem.

[00:26:31.752] Jeri Ellsworth: I think it's really exciting right now. I've been following all the rumors. I don't know what to believe. You know, some of the things that I'm seeing like, oh, that makes sense, like health, fitness, you know, stuff like that. Oh, yeah, that could be like probably a pretty cool use for a high end headset from Apple and then other ones. I'm like, hmm, you know, like desktop replacements. I've seen that a lot in the past and like, is that real? Is that going to be productive? Like didn't really work for Mark Zuckerberg and the millions of other AR headsets, air quote, AR headsets that are just video players and stuff like that.

[00:27:06.637] Kent Bye: I don't know on those like maybe they cracked it like well They say that it's gonna be 4k per eye which is gonna be you know The highest that we've seen up to that point was maybe 2.5 K per eye But this is like maybe at a critical mass where it's like the right size to be able to actually see the text So that's what I've heard at least that it would be that screen replacement and that at that resolution But I mean by the time this podcast airs we'll all know what it is But just curious we're in this illuminal space as we're sort of speculating

[00:27:31.158] Jeri Ellsworth: Yeah, I mean, it's like, I'm pretty sure they haven't solved Virgin's accommodation, like, so resolution doesn't fix that. So if it's better for text, that's good. Like, is their intention that you use it for short periods of time? Maybe. I mean I think it could be useful for me like I'm on the road or something I need like a big monitor for 30 minutes or something but I don't know if it's the price point I'm hearing I'm like hmm I need some other really cool things in there to justify that. It's exciting I can't wait to see you know there's the saying all tide or when tides rise all ships forever.

[00:28:09.158] Kent Bye: The tide rises, all boats.

[00:28:10.399] Jeri Ellsworth: Yes, that's it. Thank you. My brain's mushed from all day at AWE. No, I mean, it's what we need right now. Like, Horizon Worlds and the metaverse, like, kind of disillusionment right now is, like, not been great, you know? I mean, this time last year, I think there was a lot of hype around Metaverse, and that feels really flat right now. And I think there's some folks out there very pessimistic, like, oh, it's not going to happen this time, but maybe Apple can change that. Because they really are good at making products that have a really good bundle of functions. I'm excited to see what they are. I'm scratching my head on a few, and I'm agreeing, like, yeah, that's probably pretty cool on others.

[00:28:55.058] Kent Bye: Yeah, and as you look forward for what's next for Tilt 5, I imagine you keep iterating and having hardware increases. But what's the main priority as you move forward in getting it out to new demographics and continuing to outreach and expand the ecosystem from software or these next generation hardware? So I'd love to hear some of your thoughts of what are the big things that you're looking at right now for the next phase of Tilt 5.

[00:29:18.895] Jeri Ellsworth: Oh, it's content, content, content. You know, that's all we're thinking about. Well, not all we're thinking about, but we're thinking about that a lot. It's like, what can we learn from what we're seeing? Like, where do we need to double down on content and get more of it? And where did we make mistakes? We're doing, you know, NPS scores and surveys. Now we have some critical mass out there on headsets. And it's like, oh yeah, we see people are like, we like this part of it. Like, okay, we figured you would like that part of it. And like, oh, we didn't like this part of it. It's like, oh, that's a bit of a surprise. Or, oh, not a surprise. We definitely, now that's prioritized in the future, we got to fix this problem. It was fun. We ran our NPS surveys and it came back and the team, it's a bitter pill to take sometimes because people will be critical. I just asked folks, are you surprised at all by this complaint? And some of them were just like our lack of execution on doing things that should have been kind of simple and it just woke us up. Some of the complaints were like, I bought the TIL-5 because I wanted to play XYZ type of game. And we're like, we got five of those. Why is this person upset? And then we go back and we look, oh, this person bought their system six months ago and we forgot to communicate very clearly to our customers that now there's five versions of this game that probably would satisfy them. So it's been really good for us. And so that's going to lead us into where we go with content, where do we go into other vertical markets, and what do we do for second gen products somewhere down the road.

[00:30:52.642] Kent Bye: Awesome. And finally, what do you think the ultimate potential of augmented reality and these sort of XR mixed reality systems might be and what they might be able to enable?

[00:31:03.207] Jeri Ellsworth: I think it's huge. I think it's on par with a lot of other introduction of technology. Home computers was a game changer, but it took a while, right? People forget that it took a long time. Mobile phones was a game changer, but it took a while before it really took off. And I think, you know, XR devices of all different flavors It's going to take a little while before everyone's using it. I can't even think of any industry that happened overnight. And sometimes we forget about that. We're so excited. We see the potential as we squint really hard and we're like, why doesn't everyone want to have one of these in their home? And it's like, oh, yeah. It took about 30 years to get cell phones in our pocket and 20, 30 years for everyone to have basically a computer in their home and stuff like that. We probably can't even imagine what impact it's going to have. As much as we try to use our crystal ball, now we have AI in there, which really seems like a powerful tool. How is that going to apply with XR? It's hard to imagine.

[00:32:08.120] Kent Bye: Awesome. Is there anything else that's left unsaid that you'd like to say to the broader Immersive community?

[00:32:13.562] Jeri Ellsworth: Buy a Tilt 5 system. $3.59 for your base kit. And then if you want another headset, it's only $300 more. And typically, people are buying two and three headsets at a time to get that group experience. So a shameless plug.

[00:32:29.988] Kent Bye: Awesome. Perfect. Yeah, always excited to see what the latest innovations are. And also, we just really appreciate your perspectives and reflections on where we're at in the industry and where we might be going in the future. Thanks for taking the time again to sit down and help unpack both your journey and what's happening at Tilt 5. So thank you.

[00:32:44.662] Jeri Ellsworth: Thank you very much. I love chatting with you.

[00:32:47.524] Kent Bye: So that was Geri Ellsworth. She's the co-founder and CEO of Tilt 5. And we're talking about the latest updates with Tilt 5 and also how she's been using ChatGBT in order to develop some of these different prototypes for augmented reality. So I have a number of different takeaways about this interview is that, first of all, Well, I highly recommend you go back and listen to my interview that I did with Jerry Ellsworth last year at Augmented World Expo, in episode 1021, where we do a deep, deep dive into her journey into augmented reality, and also just how mind-blowing the Tilt 5 technology is. It's just So cool and I feel like there's a lot of different companies that can start to buy like literally like a dozen of the tilt 5 for the price of one head mounted augmented reality headset like the HoloLens and so if you're thinking about exploring the potentials of Interactive collaboration or even the potentials of gaming then they've been certainly pushing limits for the different types of tabletop augmented reality human-computer interactions So definitely go check out the previous interview and if you have a chance to check it out definitely check out the technology I was so blown away with how good it looks and it's surprising because it's sort of like a augmented reality hack but actually Overcomes a lot of the different limitations when it comes to price and field of view and the quality of the experience It's just such an immersive experience So definitely check out the technology and support whatever they're doing because it's so cool to hear they're kind of taking this Indie take on all this and also they just have a really quality product So yeah Just super fascinating to hear the updates and what's happening in six months that they've been shipping out the top five units and yeah just the challenges of communicating to folks and the spirit of trying to bring people back together into a face-to-face context. And I think that's where the TILT 5 is really exceeding. They just released a really humorous promo video calling for people to come back into, like, the old days of LAN parties where, why don't we start to have this type of physical gathering face-to-face? And I think that's where a lot of what TILT 5 is doing is trying to bring people into a face-to-face context and have these different games that are happening within augmented reality. but this is part of the xr and ai series and so i also just wanted to highlight how jerry as the ceo and working with all of her background and a rapid iteration for working in the toys industry just how she has an idea and she wants to rapidly iterate on that idea get something out very quickly so she's able to feed in the api for tilt 5 into chat tbt and there's certain limits of the tokens for how big of the tokens that you can put in at a time and so it also has a bit of a memory so you have to create these different prompts and she's saying you know as a senior unity engineer how would you solve this or that and use natural language to describe what you're wanting and then input that and then say hey is there different libraries that are out there and just the way that she's able to do this type of rapid prototyping a conversational interface just to see if it's fun, you know She's saying that you want to design around the fun and you want to see if it's a compelling a fun interaction And the quicker that you can get to that fun Then you can start to flesh out all the different software architecture and software development or whatnot but as somebody who's a CEO and not necessarily a professional coder she's able to use the chat GPT and and AI to get it good enough and to also help fix the different errors and to get different libraries from GitHub. It's just really cool to hear that creative process and to see that she's able to create this demo that was really quite compelling to visualize all the different satellites that are flying around the Earth and have different ways that you can highlight them or interact with them in a more gamified way. But yeah, super cool to hear how she's starting to use these tools to rapidly develop some of these different prototypes that ended up showing there at Augmented World Expo. So, that's all that I have for today, and I just wanted to thank you for listening to the Voices of VR podcast. And if you enjoyed the podcast, then please do spread the word, tell your friends, and consider becoming a member of the Patreon. This is a listen-supported podcast, and so I do rely upon donations from people like yourself in order to continue to bring you this coverage. So you can become a member and donate today at patreon.com slash voicesofvr. Thanks for listening.

More from this show