The third annual Virtual Beings & Virtual Societies Summit happened on October 29-30, and I had a chance to attend and unpack it with the co-founder of Fable Studio and instigator of a new Decentralized Autonomous Organization (DAO) called the The Culture DAO, which officially launched as a part of the Summit. They also launched the social token of $CULTUR, which is a token-powered Guild for virtual beings, avatars and metaverse identity, and you can find more info in the “$CULTUR – The Beginning” manifesto published on October 6th.
Saatchi was one of the co-founders of Oculus Story Studio bringing together storytellers and game designers to be on the frontiers of experimenting with immersive storytelling. After Oculus Story Studio was shuttered in May 2017, Fable Studio was able to continue development on the Neil Gaiman-optioned piece of Wolves in Walls, which premiered at Sundance 2018. Wolves in Walls featured an embodied virtual being character named Lucy, that distilled many embodied lessons of communication and body language from the Immersive Theatre company Third Rail Projects.
Wolves in Walls was released on the Rift in 2019 and the Quest in 2020, and Saatchi said that people felt such an embodied connection to the character that they would want to try to communicate with her. This triggered a long journey aspiring towards virtual beings that have a personality, character, incentives, and motivations that make them interesting to talk with in the context of a story world. In other words, the ultimate destination is Artificial General Intelligence. In my previous discussion with Andrew Stern, one of the co-creators of Façade, he said that the constructs of a story world allow you to create a bounded context that allows you have conversations with an AI character that would feel more real and capable than what AI is capable of today giving the limitations of contextual awareness and common sense reasoning.
After many years of experimenting with GPT-3 and one-on-one chatbot types of experiences with Lucy, he’s come to the conclusion that none of the current approaches are getting us any closer to the dream of AGI. This is a big reason why he’s looking for radical approaches of incentive design from the web3 & DeFi worlds of cryptocurrencies, social tokens, DAOs, and NFTs. Rather than data mining the Internet to train massive language models like GPT-3, Saatchi is hoping to create simulation worlds that are dedicated to creating mentorship relationships between humans and AI in a way that could have train-to-earn model of a cryptocurrency token or other gamified incentive structures to move beyond the current limitations of big data scraping and training models that have reached a dead end for now.
I had a chance to talk with Saatchi this past Tuesday after his Virtual Being + Virtual Societies Summit to unpack his quest towards AGI virtual beings, and some of the recent inspirations from the web3 world, multi-modal learning, incentive design, and The Culture DAO Guild to cultivate a cooperative training community to escape the extremes of AI Winters and AI Summers, blitz-scaled companies that lead to bankruptcy, and siloed information within innovation spaces. It’s part of a larger movement towards decentralization, digital ownership, and DAOs that are gaining more leverage through community organizing and mutually financially-incentivized, cooperative action.
I still have lots of cautious skepticism towards these technologies that are always in right relationship to the world around us, but there’s also a lot of exciting potentials to create new social dynamics that just may allow us to start to escape from the more negative aspects of the consolidated power of Big Tech companies their more settler/colonial mindsets of seizing our private data to fund the underlying immersive technologies of whatever may evolve into our ideas of the Metaverse.
LISTEN TO THIS EPISODE OF THE VOICES OF VR PODCAST
After attending the Virtual Beings Summit, I was inspired to buy my first EcoNFT from XR artist Sutu. I’m excited to see where he takes the VR and AR integrations for Neonz, but also have a bit better sense of what Saatchi was talking about in terms of being a part of a community that is mutually financially incentivized to contribute to the project.
After attending the @virtualbeings1 summit today, I decided to buy my first #EcoNFT from @sutu_eats_flies, #NEONZ #5544, which is a part of his @Sutuverse.
Looking forward to seeing what VR & @eyejack AR integrations he has planned for them.https://t.co/Vc6Ykv0FnI pic.twitter.com/hsvgutlIP0— Kent Bye (Voices of VR) (@kentbye) October 31, 2021
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. Today I have a conversation with Edward Asachi who is on this mission to create a artificially general intelligent virtual being and that all the existing ways that are out there that are maybe leading us towards them are not viable paths that he sees and so he's really trying to look for Some radical new ways of thinking about how to incentivize and create these Virtual beings that are eventually going to translate into these non-player characters and AGI's that we're interacting with in these story worlds So on October 29th and 30th was the virtual being summit. This is the third virtual being summit that Sachi has run. He's one of the co-founders the oculus story studio which eventually got shuttered but then some of the IP that they were working on with walls in the walls and actually got passed along to the Fable Studio, which continued to then work on Wolves on the Walls, which got released back in 2019. And in that, there's a character named Lucy, which is like this virtual being. So they continue to work on Lucy as this virtual being character. Edward's been in conversations with the larger virtual being communities, everything from Little Michaela, Code Miko, and these virtual influencers, all the way to these frontiers of multimodal learning and artificial intelligence and machine learning and all that. And also on the 29th was the beginning of the CultureDao, which is a guild for virtual beings, avatars, and digital identity. It's a decentralized, autonomous organization that was originally announced back on October 6, 2021, but officially got its start during the Virtual Beings Summit, which is kind of a social token. It's a community-driven organization that's trying to create new incentive structures to be able to facilitate collaboration around virtual beings, avatars, and digital identity. So that's what we're covering on today's episode of the Voices of VR podcast. So this interview with Edward happened on Tuesday, November 2nd, 2021. So with that, let's go ahead and dive right in.
[00:02:03.414] Edward Saatchi: I'm Edward Saatchi, co-founder of Fable Studio and instigator for the CultureDAO. And Fable Studio is building out a simulation for AIs to live in, including our brand characters like Lucy, but also characters that anyone can create and power them with AI. And the CultureDAO is a decentralized autonomous organization for the whole virtual beings and AI community, and it's powered by the culture token. And the theory of an industry token is that if everyone in a high innovation industry is mutually financially incentivized, then maybe we can avoid some of the summer and winter, blitzscaling and bankruptcy, siloing and reinvention of the wheel that I've seen again and again in VR, in AR, in AI. in XR, in virtual beings, in all of these high innovation industries. And I don't like it, and I'd like us to move much, much faster forward. And so one idea that I'm exploring, and hopefully it seems as though some people in the community think it's cool too, is of an industry token for the whole industry that we live in.
[00:03:27.553] Kent Bye: Okay. Yeah. And so just as a bit of context, I think I first met you when you were still at Oculus Story Studios. Yeah. And you since went on and started Fable Studio and Wolves on the Wall as a VR immersive storytelling experience working with Third Rail Projects and Immersive Theater to be able to have these virtual beings of Lucy. So maybe if you could catch us up and to, you know, the continued work that you've been doing with Lucy and maybe just some general context as your background and your journey into this whole immersive storytelling, VR and virtual being and Web3 experiences that you're doing now.
[00:03:59.438] Edward Saatchi: Yeah. So I am from the UK. I came to the US to join the Obama campaign in 2007 and created data and social tech for the campaign. And then after we'd won, moved to DC and started working with the administration around the same technology, trying to help people sign up for Obamacare. and working on issues that were important to me and to a lot of the people that we worked with at the time and worked on the re-election. And after the re-election felt as though that was the only person that I was passionate about and sold that company, that startup. that created that technology to the main big data provider for the whole democrat and progressive technology space and moved to SF and found two friends from Pixar. We created Story Studio that was then acquired by Oculus and turned into Oculus Story Studio. And we created the first real-time VR movie and then created Quill, which we used to hand-paint our third virtual reality movie, Dear Angelica. And then we Created Wolves in the Walls, which was our fourth virtual reality movie between Oculus Story Studio and Fable. And that one had Oculus Touch. Because we were at Oculus, we got to try out and use the tools early. And we liked the idea of a story that you can interact with, characters that you can interact with. And that was back in 2016, 17, that we were exploring those. And I think we were probably the first team that created a 50-50 blend between Pixar folks and games folks in the VR story worlds. That's now pretty standard, but at the time figuring out how those two groups should work together was definitely a challenge we had there. creative director for Bioshock 2. We had people from a lot of immersive simulations that I really respected, like Doug Church, who worked on different Looking Glass games, which is probably the father games company for all of that. And bringing together immersive theatre people like you were mentioning with Third Rail. So we were very passionate about that. I think we, with Wolves, created a summation project for what we believed VR movies should be and started to feel that we kind of explored what we felt was important to explore in that space. And also that having a studio inside of Facebook or Meta or whatever that was there to inspire and educate had kind of done what it was there to do. and started to think about what was next. And we felt that virtual beings was what was next. And so we focused on AI, how AI can bring characters to life, how simulations can bring characters to life, how we can connect with characters. And now that's what we're doing. at Fable is trying to build simulated worlds that inspire characters to have rich daily lives without a single person having to kind of manage everything. So you could sort of think about The Sims meets GPT-3. a 3D simulation with complex lives and routines and rhythms that gradually becomes emergent as those simulation loops conflict with each other, and individual humans that are overseeing it, making the characters interact with each other. We believe being compensated for that work, sort of train-to-earn model, where people have a sense of ownership over the character they're creating and they are earning for that work. So it's hopefully a vision that takes into account the huge amount of work that's going to require to actually create AI characters. If you've talked to an AI character, they're terrible. It doesn't work right now, the AI. It works for 20 minutes and then you just get very bored and kind of goes off the rails. We want to have a linear path for for how, by harnessing game mechanics and harnessing economics, we actually can incentivize people to take part in this big, big mission of how do you actually make AI characters work as believable, rich, deep characters. It's going to take a lot of time, and it's going to take a lot of humans caring about those characters, feeling a sense of ownership in order to actually achieve it. And hopefully by taking a humble approach that accepts where AI is today, we can go a lot further than if we were kind of kidding ourselves that AI characters are just on the verge and all of that. We have to create systems and incentives that are going to get us to a big long range goal of AI characters that actually work, fully work. And that's what we're working on at Fable.
[00:08:43.014] Kent Bye: Okay, yeah, I can definitely see the progression from my direct embodied experience of Wolves on the Walls, of having my embodied interactions with Lucy. To me, it's like the bar that's set in terms of the most engaging embodied interactions I've had with an AI character, just really nuanced things that you're distilling that wisdom from Third Real Project's immersive theater, too. really was digesting it into the timing and ways that she's moving her body, the way she's holding herself. So a lot of subtle things there, I think that are kind of distilling from immersive theater. But after that, you've had different aspects of like chatbots where I've had discussions on Instagram with Lucy, and then you've had other things with Twitch streams. And so just this past weekend was the third virtual being summit that you've held. And I think you've probably had an evolution from each of those summits in terms of your thinking about what is interesting. So maybe you could take me back to the point that you realized that the AI was going to be the thing and that you wanted to start to bring a larger community together to have these larger discussions about virtual beings and the full range of what virtual beings are.
[00:09:47.710] Edward Saatchi: Yeah, no, I think it might be good to almost level it so that people can follow along how to think about it. But you're totally right about embodiment. I think that's a great word. And you're right about Third Rail, this immersive theater troupe that really taught us a lot about how a person reacts to another person being in the room and acknowledges their presence and reacts to that presence. So that's a great thing to kind of ground us. Like that's where we started. Lucy couldn't talk to you in wolves, but she could react to you. And she validated your presence in that space. And for that reason, we started to notice people wanting to talk to Lucy and talking to her because they just felt she was very real. And so we started thinking, okay, well, that's interesting. And it seems like a real path to explore and AI seems to be growing. So you're right, we've been exploring a couple things publicly. So chatbots, so how do you talk to the character, whether that's texting or video chatting with the character. And then more recently, we've also been exploring the character playing video games, jumping into Minecraft and being an embodied character, but embodied in a virtual world. which we think is fascinating because then you're both on the same plane. But I think more recently, so in the last sort of nine months, now that we've thoroughly looked at chatbots and looked at GPT-3 and looked at all these technologies, we're convinced that this isn't the right path to actually get to AI characters. And if I were to quickly say what that path is that we don't think is the right one, it would be having these one-on-one conversations with characters and powering all of that by a model that's kind of scraped from the internet and all of that. And I think that does work for certain situations, especially ones where the character isn't so important. So a therapy bot that kind of redirects things onto you. I think that could work because they're just in each moment trying to, like, come up with the appropriate sentence to say back to you. And, you know, scraping the Internet to know exactly what to say next just sort of works. But if you want to really have a rich character that has a life, knows what it was doing 10 minutes ago, an hour ago, a day ago, a week ago, a month ago, and can talk to you about that, knows what they want to do in a month or a week or a year, and that can change over time. GBD3 and those technologies aren't really going to get us there alone. So that's why we started to think about the simulation and having characters live in a simulation where they are together, where they're interacting with each other, where emergence can come from the different simulation loops the characters are under, and also making it more of a game because we don't believe that any technology is going to really work for an AI character without human intervention. through the life of the character. So giving you a sense of game mechanic for like how to control the character and grow it, as well as train to earn mechanics such that you're financially incentivized to contribute to the character. I think that's both more realistic to what is actually achievable, maybe could lead to more beauty and more art in the world. but also could be morally better because I don't think it's a great path that we're just constantly scanning all the internet to like come up with these models or something a bit creepy about it. And if instead we could supplement that work with humans actually really caring for these characters and trying to grow them, then I think you're onto something that's both artistically better, maybe, I believe, innovation-wise better, and maybe just better morally as well for how we actually create these AI characters. So that's kind of our journey from how we went from making a character in a movie to how we are now advocating for train-to-earn, for a simulation that the characters live in, and for people to have a sense of ownership over the characters that they're creating.
[00:13:44.058] Kent Bye: Yeah, well, I want to get back to the incentives here in a moment. Before we do that, I want to dive into the spectrum, at least how you understand a virtual being, because I know that you've featured on your summit, everything from little Michaela, who's like a digital representation of a virtual influencer, who's being run by an entire team of people that's architecting this, but using the conceits of social media that we are all familiar with to be able to tell a story. everything to like these autonomous virtual beings that are AI driven, but are in the context of a story world. And just to kind of the spectrum of how you start to see the range of what virtual beings all encompass.
[00:14:22.418] Edward Saatchi: Yeah. So yes, it definitely includes synthetic media that has no innovative AI, let's say. So that could be Code Miko, who is a character that a brilliant woman has created over the last year, where you kind of see the virtual avatar of this character, Code Miko, who lives in her apartment and wanders around. Little Michaela, who was created by a brilliant team called Brad. Those are important and connected because they are teaching us how to get humans to accept synthetic reality, just as bringing immersive theatre people in was important to how to get people to accept a computer-generated character. So some people have said, oh, well, God, but these people aren't really AI. That doesn't matter. The business that we're all in with virtual beings is how do we make you feel that this character is real? Feel and real being the most important words. So they know how to do that, in many cases, way better than the people who are trying to create AI powered avatars. And yes, it goes all the way to AI-powered chatbots and all the rest. And I think we at Fable would like to significantly expand that spectrum because I think even that spectrum is too narrow in that it should also include characters and simulations. And that's the work that we're doing. But yes, it's the character that you are building a two-way relationship with that's interactive and in some ways AI-powered. And I don't think any virtual being exists today. So what we're trying to do at the summit is to inspire the person who's maybe 18 years old today, who is going to build a true virtual being, because I don't think there really exists one yet. But we should be neutral about where we take ideas from.
[00:16:13.856] Kent Bye: And I think in the AI community, that's often referred to as AGI, or Artificially General Intelligence, which would mean that you have AI that is just pretty much like a human consciousness to being able to have more sensemaking capabilities, common sense reasoning, all sorts of things that are probably the Turing tests of the next century in terms of replicating different aspects of intelligence. But in the context of a story world, I think you probably have more leverage to be able to cut some corners. I think that's probably where facade which is one of the most innovative ways of integrating AI, but in the context of the story world, they're able to kind of cheat that a lot by having the construct of that story so that even though the AI is not as robust as that, you can use the context of that story world to cut a lot of corners and make it feel like it's real, almost like Westworld in some ways.
[00:16:59.634] Edward Saatchi: Yeah, we think a lot about Westworld because of the simulation aspect. You could almost think of, let's say, on one side, these one-to-one relationships that movies give us examples of, like the movie Her and Blade Runner, and then these society examples that movies give us of Matrix and Truman Show and Westworld, where it's more about a society of synthetic characters. I include Truman Show because of the synthetic side. And we think at Fable that that second one, the society, is actually the right path to go on, and that just having this one-to-one dynamic isn't going to be the right one. But I also, just on your AGI point, would just say, you know, I think there's a sense that the, which I think you're correct to kind of point out as maybe flawed, that the story people need the AI folks, the AI people don't need the story people. They need us just as much as we need them, if I were to put myself in more of the story and emotional side. even though I'd love to be 50-50. For AIs to progress, it seems like they need to be interacting with people. And if you can't create an AI tool that people enjoy interacting with, then that tool isn't going to learn and get better. So it's fine when it's just like photos on Facebook and we're just like, yeah, you tagged it correctly. But once you're getting to a general intelligence that's multimodal, then you're creating something that's very similar to a human. And our expectations are actually instructive for how a human would interact by combining together computer vision, natural language, and synthetic speech, and natural language generation, and like all of those things. So they kind of need us just as much as we need them. And I think there's almost a sense that they're just on this like linear track to AGI, which seems completely off to me. The idea that they just need more data and more compute power doesn't seem like it's accurate in my personal view from what I'm seeing as the progress that's being made. So there needs to be a sort of X factor. And I think the X factor is actually building AIs in a multimodal that people enjoy interacting with and therefore are willing to spend hours knowingly training instead of us just unknowingly training AIs through Facebook saying, we think this is the right person. And the unknowingly stuff, I don't know if you saw a few hours ago, Facebook announced they're going to turn off all that facial recognition for the photos. So unknowingly training AIs, whether that's just by writing messages on the internet or by tagging friends or whatever, like the idea that that is going to get us to AGI just seems completely out of touch with the reality of like how AI is moving forward.
[00:19:39.935] Kent Bye: Yeah, well, over the weekend, I got a chance to attend the Virtual Beings Summit, and the programming was, let's say, a lot more expansive in terms of what's happening with Web3, NFTs, and with Intelligent NFTs, with iNFTs, with Aletheia AI, and someone over the course of the weekend quoted the Charlie Munger quote, which is, if you show me the incentives, I'll show you the outcome. And that seemed to be a big theme of having ownership or value exchange in a way that is going to maybe cultivate these AIs where you treat it as almost like a pet or a parent child relationship, or there's the virtual pets, which are the first AIs that are very simple in the sense of that you just need to take care of their basic needs. But thinking about this nurturing relationship between investing your own emotional labor into educating an AI, and sort of building up a relationship over time. And what's that mean to kind of invest yourself and your time into something that's going to grow value over time. Those are some of the interesting confluences that were surprising to me, but it kind of felt like you are going all in into this Web3 crypto NFT realm. And I'm really curious to hear like when that turning point was for you personally to realize that there was something here that was going to be feeding into this larger movement with virtual beings.
[00:20:52.264] Edward Saatchi: Yeah. Well, I mean, I've always felt the summits And whether that week also create a story studio university, whilst you're at story studio, like I've always cared about, inspire and educate as a mission, whether it's a story studio or fable or culture. So I do want the summit to be a campaigning summit and not just a place for us all to hang out and be like, oh my God, it's so great to see you again. Like, that's fine. But we're actually trying to do something quite important and quite difficult. And so we should be really serious about what's working and what's not and take new ideas because we are not on a trajectory to building an AI character, to really building a virtual being. We're not on a trajectory to do that. We're on a trajectory to have brilliant demonstrations and some great conversations that you have with these characters, but we're not on a trajectory that we should be on to creating virtual beings. And that's true of every summit, that we're not on that trajectory and that we have to have radical thinking each time until, you know, I mean, I guess we're really convinced that we're on that trajectory. I don't think anyone honestly is convinced that we're on that trajectory if we're really being honest and humble about it. So that's why I do want new ideas in each summit and to be provocative and to make people question the path that they're on, because that's where innovation comes from. So the first summit, we brought together the founding thinker of multimodal learning, Lukasz Kaiser. We brought together a lot of people that I really admired around multimodal learning, around AI, open AI. and tried to provoke the people in XR and the people in movies and games to think that this could really change everything. And I think they do know that now, so it would be no point having another summit that was restating things or like, let's check in with the people who were here three years ago. It's not relevant. People have been exposed to that and they can learn about it. So I mean, obviously, I suppose I selfishly did include Fable, but I mean, that's terrible. I shouldn't have done that. But you know what I mean? I just mean, I like it when it's not repeating year by year. Fable will always be part of it and Lucy, but I would never want to abuse any trust of like using the summits to really expose people's new ideas. So a big new idea this year has been Web3. And I think you can see that I don't have faith that we are just, month by month, we get closer to these AI virtual beings. I don't think so. We have to be radical in our thinking and really embrace new ideas. One new idea that the community is really curious about and learning about is incentive design and is how we get people to help be part of the big, big mission of how you make AI smarter and that we need people. We need people's help that just scraping data surreptitiously to get better models and getting more compute is not the path. I just don't believe that. We've all had enough time with these products. I don't think it's the right path. So the revelation around the potential incentive design of train to earn has been something that for me has happened this year. I mean, maybe many of your community like me might associate blockchain with things like privacy, might associate it with decentralization and might be like, okay, what does this really have to do with like, why is it connected to AI? And there really has been very little intersection between AI and Web3, very, very little. I think Arif from AletheAI, who is working on intelligent NFTs is one of the few touch points. Numeri, a friend of mine, Richard Crabe is another touch point between these two worlds, but there are very few touch points between those two worlds. And so I was trying to communicate to people that maybe these two worlds have things to learn from each other. And there were several, obviously DAOs is one that I'm really exploring around industry tokens with the culture DAO and like maybe incentives for a whole industry could be good to break down silos. And then specifically within our industry, working on incentives that would get you to feel a sense of ownership around the virtual being and want to contribute to making it a better fit for its character. We try not to just say smarter, because that could mean anything. Maybe making it a better fit for its character would be more precise, the way that in Westworld, they would care about making the AI a better fit for the character that's playing at that moment, not just generally is the character smarter, because that would not have clear meaning. So yeah, incentive design, I think, and that Charlie Munger quote of, show me the incentives and I'll show you the outcome. That's what I think. Right now, the incentives for virtual beings, AI virtual beings, are not there to get to an outcome of actual AI virtual beings. The commonly understood AGI level virtual beings, the incentives are not there. Because so much of AI is about just scraping the data at mass scale and not about how do we get users involved, enjoyably interacting with these characters and like that parent-child relationship that you're talking about. If we could shift to that, then I think that's a good shot at how we make the characters genuinely get to those AGI level intelligence.
[00:26:10.965] Kent Bye: Yeah, there's a number of different words.
[00:26:12.789] Edward Saatchi: I didn't pull this stuff up and try to figure it out myself, but it makes sense to me. What do you think?
[00:26:19.515] Kent Bye: Well, so there's a number of different concepts that I see that are happening. And what I see happening with, say, Decentraland or CryptoVoxels are the two most popular crypto-based virtual worlds that I know. There's some other ones like Sandbox and Webiverse that are still coming up. Somnium Space is another virtual-based crypto world. There's things like VRChat and RecRoom that I'd say start with the experience first. with like trying to build an awesome game and the economy comes later. Like Rec Room, it was a few years before they added an economy and VRChat's still in the process of adding the economy. Whereas Decentraland, they start with the economy and then they add in the virtual experiences later. So, you know, they have plots of land they're selling for $20,000 to $30,000 and probably even more now. But that was when there was nothing on the land, it was just speculative. And the actual experience that I've had within those worlds have been almost like a lot of speculation. Like this is people thinking that there's value there, but don't find those types of experiences as compelling as say Rec Room or VRChat or other virtual worlds. But one of the things that I noticed with your guided tours through all these different crypto voxels and Decentraland was that there's a lot more worlds that are being created that are showing NFT art. and how there's ways in which that there's ownership. And the theory, at least, is that if that people own things, then they're going to improve it. And then over time, the value gets increased. I think there's been a lot of that value exchange just upon no basis of what the experiential aspect of it is. It's just based on this crypto hype or tulip bulb perceived value that's there that then you can swap and trade these plots of virtual land for tens of thousands of dollars, even if there isn't actually anything that was from an experiential perspective. So I think that they've been bootstrapped on that economy, but now they're starting to get to the point where they're now starting to take seriously the actual experience of some of these worlds, like Decentraland just had their Metaverse festival that I was able to hang out a little bit more, bringing lots of people. But, you know, they advertised that Paris Hilton was going to have a big set, and she literally played for like eight minutes. And then she was gone. And so there's this big money that is bringing a lot of this stuff in. And what is the actual direct embodied experience of those worlds? And the number of concurrent users over the daily or a month, there's a lot smaller than these other virtual worlds that may be valued less because of the economy isn't as strong. I mean, we know that a Rec Room is like a billion dollar company. However, the thing that I would say is that there's something that is important about ownership and value. And being able to own something and increase the value. And so something that'd be opposite than say the Rec Room or VRChat, which is relying upon user generated creation. But even like Roblox, they may have tens of thousands or millions of people creating worlds, but there's like a very small number of people that really are sustainable in a way that are able to make a living off of that. And there may still be that existential power law dynamics for anybody who's creating these virtual worlds that there are really only still a few handful of big winners. And for everybody else, they're not really making a good middle-class living off of it. But at least with the crypto, it seems like it's theoretically that they could do that, but it's also such a highly volatile area that it's. difficult to know whether or not this is something that's going to be here or whether or not it's going to have these bubbles and go down, or if it's going to be like a viable way of actually using the Web3 technology in a way that is creating these new economic models. But my fear is that it's just like replicating a lot of the existing power dynamics of capitalism, where if you have a lot of capital, then you could have a lot of crypto. And it's like replacing the tech oligarchs with the new crypto oligarchs. that are there who got in early, who happened to get all these things that are now worth hundreds of thousands of millions of dollars, but that it's not really changing the overall dynamics of making the world more equitable. So that's sort of my rough take where I'm optimistic of the potential, but also extremely skeptical based upon the values that I see of this day trader, get rich quick multi-level marketing schemes and scams that are also infusing the culture of NFTs. that you have to kind of sift through that to kind of get to the heart of what is actually going to change and be viable for everybody in the future?
[00:30:19.747] Edward Saatchi: Yeah, well, I think that's well put. I mean, there's a lot to unpack there, but I think it's worth unpacking because I do want to bring folks from your and my community of XR into Web3. I mean, specifically into culture, because I think that's a good having a social token means there are lots of people in already who can say, hey, you know, here's how it works. Like, oh, you don't know how to set up a wallet. You don't know how to do this. That's how I got involved in Friends with Benefits, which is a social token created by Trevor, who created Brud, which created Little Michaela. So it's funny that the virtual beings folks seem to maybe have an affinity for social tokens. So tokens for a community. But a lot of people I know in the space of Pixar and the rest are very curious about how decentralization could affect their ability to have jobs, have more control over what they're doing, and actually contribute to a wider goal. So at The Culture, for example, the idea would be that thousands of people who are in our industry, who otherwise might be working at Pixar or maybe are working and they just moonlight at The Culture, would part-time be able to work on bounties for different projects that are useful to the virtual beings community in a wider way, whether that's building out an ACI model, so an artificial coherent intelligence model that members could use, whether that's experimenting on different avatars, whether that's building experiments around simulations, whether that is learning, so teaching other people skills that you have, whether that's around animation or the rest, whether that's a bounty around a creative product or small minigame or, you know, everybody being able to contribute their skills and work to a wider goal of getting to virtual beings. And it's the thing I've done that is the most connected to the Obama campaign, funnily enough, that decentralized, it's bottom up. There is an infrastructure so that you kind of know what you're doing, but there's a lot of trust and people can do what they've been trained for in a freer way. So Right now, a lot of people are in quite factory settings and creative technology studios, whether that's game studios or movie studios, where they have very little autonomy. They are part of a massive four-year pipeline, and it's slightly stressful, it's kind of crunchy, and it's maybe not challenging them enough. So I think there is a space for something that's a DAO that's decentralized for the whole community. And I do think Friends with Benefits is super interesting, and that's more in the music and metaverse space, and we're sort of more in the virtual beings and metaverse space. I can't sort of claim positive things about every DAO or every organization. I think there are bad actors all over the place. but I have seen negative things from what currently happens, whether that's Magic Leap getting blitzscaled way too early and sucking up a lot of the investment money that's out there, or companies being encouraged to silo information, startups in early-stage, high-innovation industries being encouraged to silo information so that then they're all reinventing the wheel and VCs can't quite figure out that everyone is basically building exactly the same stuff that is all completely commoditized by the time anyone releases it, and it's all wasted VC money. So having a more public good approach or an approach that is for the whole industry, where people actually in that industry are thinking through how to put millions of dollars, tens of millions, hundreds of millions in a treasury towards what's best for that industry, I think that could be really good. And I think that could be a lot better than panic blitzscaling of one or two companies in high innovation industries. You get a summer for six months and then a winter for three years. So we should compare everything to how it's done today rather than against like a perfect vision. I also have no interest in speculating on land in the metaverse. A lot of it is just not that cool or interesting to me personally, but there's a lot of things that are interesting. And I think there's a reason why those Web3 folks have such high valuations. It's because they bet on something that seems like it's being proved correct, which is verifiable ownership of digital goods is important. And so good on them. Why shouldn't they reap the benefits of betting on the right thing? Really, I think we underestimate how mocked they were and how they had to believe in something, and I don't include myself in this, they had to believe in something when it was worth nothing and everyone was telling them, why are you doing this? and they were going to conferences that were like 20 people. So as much as we might now say, God almighty, what about Rec Room? Rec Room has like millions of people. Why are they not valued a hundred times the value of these Web3 people? I'm sure the valuations will stabilize and become more normal and more logical. But I do admire, and I think we should all admire, people who really believed in something, which is what I feel when I talk to the CryptoVoxels team or the Sandbox team, really believed in something when it wasn't cool, worked on it for years, and now it's become hot and good on them to try to take advantage of that. But I think there should be some admiration for their ability to see something that we now can all see, which is that digital ownership is important. And I, myself, like you, have to edit out a huge amount of froth in order to remind myself that the core big idea of digital ownership that transcends just buying a skin on one video game that one game company controls, that's a really big idea. It's kind of hard to see how the metaverse works without it. So those people who saw it first should be rewarded. Yeah, I guess your point about the capitalism stuff is a terrible example. We tried to talk through at the summit with Jacob and with Jiho was what about the terrible exploitation of the first world and the third world? Like, are we creating like a terrible situation here with play to earn, you know, where a lot of people are leaving their jobs to go and work for games? You know, I'm very worried about that. I'm less concerned about the scams and I'm more concerned about a financial crisis based around crypto, where you have a lot of people who created no social security for the people who now have jobs because they all got excited about pay to earn. And then government has to foot the bill when those games companies go down or miscalculate something and there's a crisis. And now people are actually losing their houses they left their jobs. Those are the kinds of worries I have. But whether I mind as much about people becoming extremely wealthy, some of them already were, or some of them were very poor and now have a lot of money, that just seems like the normal churn. I'm not sure that it is going to refactor capitalism. but I do worry about a financial crisis that would affect normal people and people lose their houses and people lose their jobs and then government has to step in. I worry about that on a daily basis because I think there's such a contempt of government and so much of crypto, but I think Who's going to pick up the pieces when there's a financial crisis like 2007 and 2008 created by all of the money that's gone into crypto? Who's going to help those people? It's going to be government. So I'm very, very, very skeptical of that side of it.
[00:37:49.770] Kent Bye: Yeah, just to elaborate on that play to earn a little bit, you're talking about taxi infinity, which is the game where, and you could earn up to 15 to $60 a day, but in some countries where the average amount of money, they may make maybe three or $5 a day. So you're talking about three to 10 or 20 times more than what they make per day. And then what does that do to entire cultures that are playing these crypto games? What does that do at a sociological level for entire countries? On one hand, getting benefit for some of the people, but it not really being in right relationship to the world around them. I think that's the escapist games that are helping profit these companies like Axie Infinity by having a cut of the money that's being made, but also are maybe disrupting larger things that are making a handful of people really wealthy, but also potentially disrupting these larger cultures dynamics. And so, you know, I think those are larger questions as we continue to play out. I mean, this is part of the globalization at large, just another vector for that. But yeah, I guess I guess for me, I just think about things being in right relationship and maybe part of the ways in which that the data has been aggregated and trained has not really been in right relationship for people. So that if you're able to actually compensate people through pray to earn, where you're actually training an AI model, then there's more of a direct incentive structure for rather than using AI, that's going to displace jobs. You're actually using AI to both create something that's viable, but also compensate the people who are making it valuable.
[00:39:14.998] Edward Saatchi: I've been thinking a lot about a company that maybe is fable at the intersection of Axie Infinity and Scale AI, if you've heard of Scale AI. So Scale AI is about labeling data. And it is surprisingly similar in the sense of first world, third world to what you see in Axie, where people in poorer countries are labeling our data, especially in edge cases, where a car can't interpret certain stop signs and they get sent to scale AI and scale AI can help label that data. Obviously, that's quite grim as an example, but it's also real having tens of thousands of people labeling all of that data. If you could bring those things together where it's a game and it's fun and it's not just labeling data, it's actually having a relationship and you do it both because of the financial incentive and because it's enjoyable, then maybe you've got something that is so turbocharged that it actually creates a liftoff potential for how you get to AGI characters. And that's my personal focus, is how do you actually get to a true virtual being that's really as intelligent as us? Nobody is on a path to that today, if we're all being honest. And so that's why I think radical thinking is so necessary and why looking at what play-to-earn and train-to-earn could unlock, thinking about what an intersection of scale AI and Axie Infinity could bring to the world. I think that's super interesting. And I really admire Geho's, the person who talked at Axie. I liked that he talked about dreaming of being a central banker when he was a kid. And I think there's an earnestness there. And I feel the same about Yield Guild Games, which is trying to build it, but I don't see earnestness everywhere. And I think the risks are definitely underplayed and the opportunities of lifting people out of poverty. I don't think they're overplayed, but they're maybe not being taken into account with the risk that someone might leave a real job, not a real job, leave a job that then they can't get back to if something goes wrong with the financial calculations of each company. But my experience has been that there is an earnestness in the people who are putting those other people most at risk and a kind of a willingness to be honest about the risks that I think is good and the minimum that is needed. And I think, you know, I would hope that if Fable explores similar areas that we would approach it in a similarly earnest way with, I hope, an inspiring mission, which is not just how do we build a big business, but also how do we actually get to an AGI character and I think incentive design. is an area that has been really untapped. And part of it is that we all just have faith that if we just keep waiting for the internet to get bigger and bigger and bigger and just keep scraping more and more and more and build bigger and bigger processing farms, then good things are going to happen. And I just don't think that's going to be the right path to AGI characters. We have to take radical steps and try unexpected things.
[00:42:23.232] Kent Bye: So there's a moment where you're going around into different virtual worlds. Somebody was in Decentraland and they had like an NFT or some sort of accessory and that gave them access to like a secret area.
[00:42:33.980] Edward Saatchi: That was a guy called Tox. That was so cool. I was really, really impressed with that.
[00:42:39.517] Kent Bye: Yeah. So the crypto avatars creator. So there's the idea that you could log into a wallet without having your username or phone number, just kind of like this level of anonymity, but also identity control and your wallet being your trace of your things that you own and ways that you've exchanged value. And, you know, for me, I have been resistant to like NFTs generally, I'm a, I'm a fan of like eco NFTs and the proof of stake versus proof of work. And so there's ways in which that I watched the discussion around things like CryptoPunks, which you were able to acquire one of the CryptoPunks and talk about it with a lot of different people. And then first my reaction was like, ah, like I'm not really into like the reputational aspect. I guess I'm more into the experiential aspect of what kind of experiences is this going to unlock for me if I have one of these things. So I ended up After that, watching the Virtual Being Summit, I actually bought my first NFT. It was by Sutu and he's got neons on the Tezos. So a clean NFT, but also he wants to do these AR and VR integrations with these, which I'm really excited about having an artist that I like and then want to support, but also be a part of these experiences that I may not have access to. But I'm just curious to hear your own take of why you chose to invest in CryptoPunk as part of your avatar and identity representation.
[00:43:54.630] Edward Saatchi: Yeah, well, I think they are oddly connected. The idea of you're going to wear a jacket that will give you access to secret parts of the metaverse is oddly connected to you're going to buy an NFT that's going to give you access to a community that is ambitious and that's actually, again, financially incentivized to push things forward, whether that's creating VTubing as CryptoPunks or creating AI CryptoPunks, which we've seen with intelligent NFTs. there's a mutual financial benefit because everybody has a CryptoPump. And if you do something that makes CryptoPumps more valuable or cooler, then everyone wins. Everyone who has one. It's very weirdly motivating to unlock people's creativity. So I think that is an odd connection. There's also a connection with culture where you need to own 25 culture in order to like see into that separate part of the metaverse. There's something about guilds, which is what I've gotten very interested in, of kind of secret knowledge, that a mutually incentivized, trusted group can unlock when they have those mutual incentives and trust. So, I mean, short answer is complete FOMO. And a mix of FOMO and greed and panic was one answer to the question, which I'm sure would be the most honest one of getting a crypto fund. But then the more intellectual one would be that, you know, I like this idea of being in like that movie John Wick where there's the Continental and there's kind of that hidden world of people who are in your kind of sub-community and your guild. That's so cool. And that's like very odd and weird and I guess must be how the early internet must have felt. I definitely don't feel that on the internet generally. And so I kind of like that weird, guilty element where you get to see some kind of secret other world with peers and people who you trust and are in the same kind of, whether it's industry with, or the same aesthetic with, that's kind of cool. And I like it. And that's part of why I like the jacket opens up these secret things. It's very John Wick and the Continental. So I like kind of seeing that idea come to fruition.
[00:46:10.893] Kent Bye: Yeah, just from my own brief experience, I've started experiencing some of those things as well. So I can definitely see the appeal of being a part of that community that's mutually incentivized to make that community more valuable by unlocking my own creativity potentially or whatever else to come for that. But yeah, just to wrap things up here, I'm just curious what you think the ultimate potential of all these immersive technologies and virtual beings might be and what they might be able to enable.
[00:46:35.412] Edward Saatchi: Yeah, AGI characters. I mean, I think that's why I do what I'm doing is to build out those virtual beings. There aren't any virtual beings that exist today. I think it's an aspiration. And every technology that I'm trying to have us look at as a community and ingest the ideas of, is all about, could this be one of the pieces of the puzzle? Could this be part of the grammar of this new art form, this new technology of virtual beings? And so that's the end point. And I think we all gut level know what it looks like because we've all watched sci-fi movies or read books about it or whatever. We all kind of know what it's supposed to look like. And I think ideas like the metaverse, ideas like train to earn, ideas like machine learning and multi-modal learning. These are all tactics for how we get to that ultimate space of how our species could potentially create another species, which is artificially intelligent virtual beings.
[00:47:38.852] Kent Bye: Awesome. Yeah. And I know in the virtual being summit, there's lots of discussion. And one of the things I heard was these virtual beings as a communication medium, just like XR virtual and augmented reality as a medium, but these virtual beings within themselves are kind of like a medium of expression, which I think is also very intriguing to think about these virtual beings as a medium of expression within the context of these stories. So yeah, just to, you know, final thought, is there any other things that's left unsaid that you'd like to say to the broader immersive community?
[00:48:04.887] Edward Saatchi: Definitely. I'd like to say that I must sound like I'm so certain about things, and I'm definitely not. I'm just trying to muddle through just like we all are. And, you know, I would hope not to discourage anyone on the path that they're on, because we're all trying to do difficult things. And my own hope is that we can do it together. And so I wouldn't want to give the impression that I've figured out this path is wrong, this path is right. You know, I'm just muddling through myself.
[00:48:34.826] Kent Bye: Awesome. And for more information, people can check out the culture Dow, and you also have fable studio and Lucy. Is there anything else that you'd like to point people towards?
[00:48:41.637] Edward Saatchi: No, I think take a look at the app, the culture Dow on Twitter, take a look at fabledashstudio.com and come and get involved.
[00:48:50.622] Kent Bye: Awesome. Well, Edward, I'm really excited to see where this continues to go. You brought together a great cross-section of the community through your Virtual Bing Summit, and I really learned a lot over the weekend just hearing all the different perspectives. So thanks for doing that curation and exploration through different corners of the metaverse that I haven't spent as much time in. So thanks for doing all that curation and advocacy for this community that you've really been cultivating. And yeah, it was great to be able to unpack some of it here with you today. So thanks for joining me.
[00:49:15.062] Edward Saatchi: Thanks, Cam. Speak soon.
[00:49:17.398] Kent Bye: So that was Edward Sachi. He's the co-founder of Fable Studio and the instigator of the Culture DAO. So I have a number of different takeaways about this interview is that, first of all, well, I really had a good time at the Virtual Being Summit and learned quite a lot because there's a lot of things that I haven't necessarily been tracking all the nuances of what's happening with the DAOs and different crypto worlds and this confluence of the Web3 technologies and the decentralized tech and decentralized finance and NFTs. So it was interesting to see where he's seeing a lot of these touch points and confluence into the metaverse and these 3D virtual worlds. So a big takeaway here is incentive design, and how much incentives are going to be a big part of what is probably going to be a key factor for generating an artificial general intelligence virtual being. So the existing model right now has a very subtler colonial mindset where you just go out and seize all this data, you scrape all this data, and you use that to train these big models. But it's fairly indifferentiated and absent from a lot of the deeper context. You know, they're just basically taking a sampling of the cultural biases and what is happening at a collective level, and then distilling all those biases down into an AI language model, which is not necessarily going to get to that point where Edward is saying that you're not going to have a rich character that has a life that knows what it was doing an hour ago, day ago, week ago, a month ago, or a year ago, and can talk to you about that and knows what they want to do in a month or a week or a year from now. And that could change over time, and that GPT-3 are not going to get us there. So that existing language model, the way that it's working is not really getting to the nuances of an individual character. And it's also scraping things in a way that takes into consideration things like feminist theory around situated knowledges, that would say that people have a certain orientation of a power and privilege within the context of a larger culture, and then that gives them specific epistemic privilege and knowledge about that culture that would be unique to that individual's perspective. A lot of these AI models are not really taking into those larger contextual dimensions and modeling that and trying to, at the end of the day, create a rich character that has an individual personality with its own motivations and incentives and everything else. That, I think, within itself is something that may need to still be programmed and coded, because you're talking about, in essence, a sense of free will and to what degree are you going to be able to create an organic consciousness that's able to really have its own desires and motivations. independent of any underlying phenomenological or experiential dimension other than ones and zeros. This gets to the nature of consciousness and whether or not we even, as humans, have any of that, as well. I think we do. I think we have intentional actions and that we're driven to make and create things. But we also have behavioral patterns that are fixed action patterns that are completely unconscious and driven by a lot of our own ways of coding that we've had throughout our lives through the various experiences and trauma that we have in our lives. But getting back to the artificial general intelligence, it's very interesting to hear Edward say that this is not a viable path to use GPT-3 or any of the other existing AI technologies that are out there to really get this rich sense of a character and moving away from this, thinking about these chatbots and one-on-one interactions and creating a whole ecosystem of other characters that are learning from each other. So Edward sees this as a combination of something like the SENS plus GPT-3, which is a 3D simulation with complex lives, routines, and rhythms that gradually becomes emergent as those simulation loops conflict with each other and individuals are overseen and are compensated for ways in which they're helping to train these AI models through something that is called the train-to-earn There's free-to-play games. That was a big transition into the gaming market, where you just give away games and there's advertising that's supporting them. Now is the train-to-earn, which is having some level of crypto-based exchange that's happening within the context of that game. As you earn that cryptocurrency, then the game is basically taking a cut of that, but you as a player are able to earn as well. This is something that Steam has looked at and said, nope, we're not going to do that. They basically banned all of the crypto-based games. Epic Games has actually been a little bit more open to including some of these different cryptocurrency games. Certainly, what Edward is saying here is that there are a broad range of different incentives that are happening within these different games. Whether or not there are different levels of exploitation that are happening, I'm sure that Valve saw enough concerning behavior within the broader cryptocurrency community to see all these various different scams and schemes and whatnot, where they just were like, nope, we don't want to have to deal with that. I was on the Bored Apes Discord for just a little bit to lurk and see what was happening. And every day, I would get two to three different crypto scams. People direct messaged me saying, hey, you need to join this or that. And it was just so insufferable that I just had to leave that Discord because it was just attracting all of these different crypto scams. So anybody who is in this community has had to probably block their DMs. And there's just a lot of grifter scams that are happening within that larger community. So I can understand why Steam would want to just, like, block it altogether. But in talking to Edward, I could see that there are certainly other entities that are out there that maybe have a little bit more earnestness or a little bit more integrity. But I do think that if you look at the future of how artificial intelligence is trained, that there is an existing settler-colonial mindset that is just seizing all this data and using it in this aggregate way. But what if you do move it into these other incentive design structures that when you contribute the training or your emotional labor or your knowledge into training in AI, if that is becoming more valuable, then are there ways for you to get some sort of crypto token that then would increase in value as other people do the same thing? This idea that as more and more people participate in these communities, then they become more and more valuable, and that you have a little bit more of a cooperative working situation where everybody's co-earning together by building something that's valuable. that's the idea of something like a DAO that is like a guild of using social tokens to be able to exchange and to pool resources in a way that you know what Edward was saying is that there's a lot of people who are in these like say Pixar or these game studios where they're in a four-year pipeline where they're working on different projects and maybe their input isn't really taken that much but it's something that's sometimes high stress, but also something that doesn't really tap in their full potential of their creativity. So having something like a DAO where people could join in and have more of an open collaborative, pure production where you're able to have these bounties, you know, in open source world, you have bounties where you're able to solve something, maybe get some money. In this case, you would get some tokens by helping to do one of the different things and the initial Manifesto that edward posted on october 6 there's things like different committees for the virtual being summit that was held on the 29th and 30th There's the vault committee which is giving members usage rights over a number of different types of virtual identities crypto punks and mutinate me hot clubs and the me bits and So pulling together people who are members' resources to be able to have a collective expression for how people might want to use some of those identities. There's a single character committee, so building a single character as a DAO. Avatar committee, building avatar tooling in different sets of culture. And then the hackathon community, discounts community, and learning community. So a number of different committees that are really trying to create this guild of folks working on AI, artificial intelligence. And I think generally, if you look at artificial intelligence when it comes to academic publishing, It's different than a lot of the other types of intellectual property just because so much about AI is about the data and the algorithms that are driving the data is something that most of the major companies now have been a part of Publishing tons of different papers I know the company formerly known as Facebook now known as meta has like 79 different AI papers that are going to be at the NIPS conference that's coming up here which is the big annual machine learning conference and A lot of open-source sharing of knowledge when it comes to AI. There's just been a huge amount of progress over the last five or ten years in the AI community, in large part because you have a lot of this shared knowledge. But at the end of the day, you still only have that data to really make the most use out of a lot of these AI algorithms and neural network architectures. At the end of the day, you still need a lot of compute power, as well as the data to actually make use of it. So, if you're able to use something that's more open-source, peer-driven, where the people are contributing their own knowledge into cultivating these virtual beings, then you have completely new incentive models that are being created there. Right now, there's no real good way to do that, but it could be that some of these Web3 technologies may actually be the most viable path towards doing that. Aletheia AI, which is these intelligent NFTs, is one example of that, which is trying to create these virtual personas that people then take ownership over and then perhaps do certain things to increase their value and help train them. Yeah, so there was a whole session that was with Aletheia AI that was happening at the Virtual Being Summit. Last I talked to Edward, he wasn't quite sure whether or not a lot of those sessions from the Virtual Bing Summit were going to be made publicly available, or if that's something that he's going to gate behind the Culture DAO to be a part of your member, then you'd have access to a lot of that learning repository. So definitely check out if you want to get more involved into this decentralized autonomous organization, the Culture DAO. I imagine that we're going to start to see a lot more of these DAOs. I mean, this is the first DAO that I've had a chance to talk with, but there's plenty of other DAOs that have been around for a while. Actually, that's not quite true. I think the Decentraland may have actually been a Dow. I talked to them a number of years ago. One that just recently launched here officially on October 29th. Definitely go check out the manifesto, which I'll link here down in the show notes. Yeah, I feel like that this conversation was just so rich and dense, with so many different insights. I really tried to distill down lots of different stuff here, in that there is this path of wanting to create these AGI virtual beings. I just happened to watch last night Free Guy with Ryan Reynolds, which is the story of a non-player character within the context of a game world that is sort of like a Grand Theft Auto type of game, but similarly a little bit more of The Sims, where each of the characters have their own rich lives, and they're in their loops and routines, and through these different disruptions, then he finds more and more agency and autonomy to break out of some of his existing game loops. So, it's sort of like this mix between the Grand Theft Auto, Sims, and then aspects of The Matrix and The Truman Show and Ready Player One, but with more augmented reality and more of a romantic comedy element. It's actually quite good, and I recommend checking it out. I think it talks a lot about these things, especially in the context of creating these AI virtual beings, because in that film, you start to depict that. I think Westworld's another good science fiction reference point to think about. going into these ecosystems and these worlds and to be able to interact with these different characters and these loops where, you know, they have their own intention and motivation for what they're doing, but they're interacting with these other stories and these other humans to have these different types of emergent behavior. So that was, I guess, the other big thing that I got out of both the Virtual Being Summit and seeing some of the latest presentations for what Fable Studios is working on, but also listening to what Edward is saying is that really getting away from this one-on-one mode of training AI. and getting more into this communal, systems-driven, having lots of AIs with their own loops, and that there's growth that happens through the collisions and interactions of these different incentives and game loops that they're able to potentially learn from each other, but not only learn from the other AI, but also learn from the humans that are in the context of that, as well. What's it going to look like to create this learning context for AI for humans that make it fun to interact with? I think that's the other part that I got out of this is that to really turn it into some sort of game mechanic loops where it's actually fun and enjoyable to interact with these virtual beings, but in the same time that the virtual beings are able to learn from their interactions. And so there is a bit of this mentorship where you're able to train the different AIs. And I just can't help but to think of the Microsoft Tay example where Microsoft released the AI out into Twitter, and then the culture at large immediately taught it how to enact all these different racist behaviors. So just the same as you have these culture dowels, then how are you going to prevent people from coming in and trying to disrupt the learning by passing along some more discriminatory hate speech, racism, whatever it ends up being, just things that are not going to be helpful for the cultivation of these AIs. So how do you start to embed a certain layer of ethics, not only in the training process, and is there social tokens, or are there revisions so you can actually revert different trainings if you want to remove different aspects of things that these AI beings may have learned? You know, what is the process of that? Can you actually memory hole different aspects of things being learned? So, yeah, I think these are all these different types of cognitive architectures that are still yet to be fully fleshed out. So what are the all the various tradeoffs for being able to do things like that? I guess the final thought is just this verifiable ownership of digital goods is something that is important and something that is being recognized larger by the culture. I'd say, in some ways, any type of currency or any type of social system, there's a lot of normative standards that the community has collectively agreed upon as being valuable. There's a lot of that trust that is put into whatever system it may be. A lot of our existing systems are based upon that trust. By creating digital replications of those, then you have new opportunities but also new challenges as you start to navigate the ways in which those incentive structures may not be in right relationship with the world around us, which I think is probably one of my biggest critiques of the larger crypto community, which is the ways that the proof-of-work is not ecologically sustainable and that it is not only creating an ecological impact, but also larger supply and demand pressures. The more that the crypto community has been using some of those different GPUs or computing resources, then it's less opportunity for other people to use some of those, as well. It's a disruption. It's a part of the larger economy, the supply and demand that I think is just natural, but there's also open questions in terms of how to make it really sustainable in a way that it's not only in right relationship with the world around us, but all the other people that are involved with those worlds. I was a fan of what SUTU did with the SUTUverse because there's ways of not only using eco NFTs with Tezos, but also, you know, when he did this drop of 10,000 avatars on neons, the traits weren't revealed until like a day later. What that means is that you don't have people that are swooping in and trying to buy up and acquire the most rare traits. It's a little bit more of a lottery so that you really don't know what you're getting. From there, it creates more equal opportunity for people to have access to something that may be way more valuable than what they've would have otherwise been able to afford. So things like that were how do you make the architectures of these systems a little bit more equitable. I was a big fan of that. And I just wanted to not only support sutu as an artist, but also some of those deeper practices. But also I wanted to have my own direct embodied experience of what it felt like to buy and own an NFT. And I was surprised about how much it started to, in my own mind, bring up these different aspects of my own imagination of, okay, now that I have this virtual avatar that's a part of this Sutu-verse, then do I want to try to make this into a 3D model? Or I'm just excited about where this goes in terms of the different integrations and what that's going to unlock from an experiential perspective, either through Sutu's Hijack application or through some of these different virtual augmented reality experiences. Now, as I say all that, I just want to take a step back. As I was listening to Edward talk about a crypto-punk, I had this visceral reaction of, man, it just feels like I'm being preached to to try to join into this club of the crypto-punks. Those crypto-punks are costing anywhere from $100,000 to $200,000, so they're not cheap. One of those things where people who got really early into the crypto community, they were able to acquire before it was worth that much and before anybody that was really thinking about it in any way. But it's now a status symbol to be a part of that early adopter community who are really innovating. But also there's different ways in which that Larval Labs has a whole agreement with UTA to be able to do different licensing agreements and whatnot. So there's certain ways in which that these digital identities are being validated and potentially have different intellectual property that's created out of that. So even though I was having a lot of visceral reactions of hearing Edward talk about what it meant for him to be a part of this crypto-punks community, I started to understand that a little bit more by being a part of the Sutuverse in my own small way, but also being cautious to try to avoid feeling like I was trying to recruit lots of people into this or that particular type of NFT or universe. Because, like I said, there's plenty of reasons why people may not want to dive into the Web3 or decentralized finance just yet. But I do think it's going to be a part of this incentive design as we move forward. So that's all 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 list of supported podcasts, and I do rely upon donations from people like yourself in order to continue bringing this coverage. So you can become a member and donate today at patreon.com slash Voices of VR. Thanks for listening.
