In this episode of #StoryFirst we're joined by Alessandro Botteon, CEO of Impssbl, which is using a combination of AI and NFTs to build a story engine designed to power the creation and spread of high-value, high-impact stories.
Impssbl has created three projects:
- Proof of Story, which features characters "with an AI-written story:
- UNTLD, a platform enabling users to "create or mint human or AI-written stories as NFTs
- Musess, an experiment combining AI- and human-developed art
During this conversation, Devin Sawyer, James Deveron and Alessandro talk:
- "Storyless" NFT projects
- How Impssbl is scaling decentralized storytelling and Web3
- Proof of Story
Speaker 1: I feel there's a huge, uh, uh, space, you know, like going forward, like the, the story based and FD and like what we are doing, I think it's gonna be, uh, extremely, extremely like important going forward, especially because right now, feeding in web trees that, uh, there's lots of like projects that are kind of like, you know, like story, less projects with, uh, with no background, no lower at all. So I think what we are doing is kind of like we are bringing these different projects to life in, in the first place stories exist for, for one, I think like, uh, evolution, evolutionally like kind of like reason, which is that people need to resonate with one which with each other, we, we don't wanna feel lonely. We wanna like feel connected somehow. And stories are bridges among people are bridges among cultures are bridges among, uh, you know, like tribes and, and countries. And, and in the end, that's what they do. They connect
Speaker 2: Welcome to story. First, a podcast where we shine a light on the web three Mavericks that are leveraging the power of narrative to create immersive NFT world. We investigate the art and science of story building through insightful interviews with creators, collectors, and investors. For those who seek a richer NFT experience. This is your portal to a vivid new realm of fiction. Story. First is a production of story. Prima doo, whose mission is to encourage the growth and success of story focused NFT projects through research, education, and project incubation story prima doo brings you the blockbusters of tomorrow.
Speaker 3: Hello, and welcome to the story. First podcast, a weekly show where we talk about web three NFT projects focused on storytelling. We believe story focused, NFT projects are built to last and will form the foundation of a new breed of hugely influential properties in gaming entertainment, art, and beyond the hashtag story. First podcast highlights the Mavericks who are leveraging the power of NFT to tell the blockbuster stories of tomorrow. My name is Devon Sawyer and I'm joined by my, his steam co-host. And co-founder James Devon say, hi James.
Speaker 4: Hey guys, my name's James I'm the art director in legends of cipher group and, uh, founding member of the story prima
Speaker 3: That's right. My name is Devon. I am also a co-founder in legends of cipher and story prima story prima is a decentralized autonomous organization with a mission to accelerate the growth and success of story focused NFT projects through media research, education and project incubation. But today on story first, we are most honored to welcome Alessandro from the impossible community. We're honored to host this podcast to highlight a great story first project, but one that is very interesting because impossible is an innovative platform being built to support web three and NFT projects with community and AI at the center. I think this podcast is going to be a look into the future of storytelling. So I'm really excited to have Alessandro on the show and we'll get started with an introduction, Alessandro, tell us who you are and how you got to web three.
Speaker 1: Thanks Le for the introduction and nice to meet, uh, everyone. Uh, so first of all, it's actually a pleasure to, to be here and like, uh, it's, uh, I feel there's a huge, uh, uh, space, you know, like going forward, like this story base and FPS and like what we are doing, I think it's gonna be, uh, extremely, extremely like important going forward, especially because right now, the feeding in web trees that, uh, there's lots of like projects that are kind of like, you know, like story, less projects with, with no background Nodo at all. So I think what we are doing is kind of like we're bringing these different projects to life in, in the first place. So just, just a few words about me. I, uh, actually come from tech. I used to work in Google for a few years between, uh, Ireland, uh, Dublin and, uh, Miano.
Speaker 1: I worked in startups all around the planet. I, uh, used to study like, uh, international management. I used to study Mo production. So kind of like a hybrid, uh, background there. Then I opened my first company in the 3d and AR space, uh, for fashion in, uh, middle and from then like last year in, uh, I think it was like June 15 last year, eye opening, impossible, which is a story tech company, uh, making a story based, uh, NFD projects. And that's where, uh, I'm actually like still kind of like currently focused on building right now. So what we do is like we build what we call, uh, decentralized autonomous stories, which means, uh, again, NFD projects back by stories that spin off the next generation of intellectual property rights. And, uh, we are focusing on like, uh, yeah, basically bringing alert Webre in a, and like kind of like bridging, not just like a human, uh, basis stories or human like written stories, but also AI written stories. We are kind of like experimenting as well on the, on the edge there of what stories and what narration actually look like today.
Speaker 3: Love it. Very exciting. Yeah. Um, you know, we're, we're, we at story premier are very, um, focused on the projects that are, um, you know, telling stories in the web three space and using NFTs as part of the technology for delivery. Um, you know, uh, agree that kind of the NFT space is story list. Um, and there's a huge opportunity. So talk a little bit Alessandra about what's really exciting you about the future of the NFT market now and how is impossible uniquely positioned to take advantage of the opportunity?
Speaker 1: Sure, sure. So like, first of all, I think, uh, I mean the DFT space like moves in waves, not like just, uh, bullish or Barrish waves, but like also what type of NFPS are actually like the, the trending one right now. Like we started like last year with all the, the PFP with, uh, like the cryptos, the board date, and so on, like in different like art styles, we moved on like to the 3d phase of things, like, for example, hide based and like a bunch of like other projects like that. We move on to the gift side of things as like, for example, invisible friends and like all the kind of like related projects. So it's kind of like you see like these different, like distinct moments in time of, uh, what is kind of like the current trending, uh, NFT, but like our focus actually, it's not like on the size or the visuals, it's, it's more on the story.
Speaker 1: So, like for example, one of the projects like inspired the most, uh, busy RT, uh, uh, uh, busy our team last last year was the dilute project. You know, when we, uh, saw the dilute project, when we saw like what they managed to build in a completely decentralized way where you like these, uh, bags of like, uh, invisible kind of like imaginary bags of like things containing words. And those words could be, artifacts, could be words, could be, you know, like, uh, uh, you like metal coins and stuff like that. And the idea was kind of like, uh, what's really important is what you build on top of it. What's the ground up kind of like narrative level that spins itself out of the, uh, project itself. That's kind of like what inspired us to, to make what we call the centralized autonomous story. So, uh, in, in short, everything we do can actually be divided in three main things.
Speaker 1: So, uh, thing, number one is what we call our original series, where we decide everything, we decide the visuals and we decide the stories. And, uh, it's, uh, um, for example, one of the, the examples that are, uh, live already is that one is pro story, which is our first project, like launch between, uh, December and January, uh, December last year in January this year. And it's a project where we, uh, basically like paired a traditional TFT, uh, with an int inte an artificial in intelligence story, uh, written specifically for that character. And that story could be lots of different things, could be, you know, like, uh, love story, uh, drama sci-fi it could be like lots of different genres. And at the same time could also be different formats. It could be a plot, could be a dialogue, could be a description, could be a scene.
Speaker 1: It could be lots of like different things. And the idea was that we wanted to make these characters more alive. We just didn't want like them to just be, you know, like visuals, uh, like profile pictures as well. And, uh, uh, another original series, which is actually launching about like a couple of weeks, uh, kind of like shares the same, uh, mentality, but completely different work of art. So it's, uh, the second project is called music and it's a collection of 999, uh, portraits of, uh, women that celebrate a borderless inclusive and veil idea of like female beauty and the, this project actually, it's actually a collaboration between, uh, uh, impossible and an Armenian artist, Armenian Pinterest that made the, um, portraits in the first place. So it started as an experimentation because it's, uh, we started like from 100 portraits, black and white, super elegant, super beautiful made by her.
Speaker 1: And the idea was kind of like, we wanted to amplify them with stories, but going forward with this, like we, uh, amplified them. We expanded on the role of artificial intelligence because we not only ended up like giving every single MUEs a story, but we also like ended up using artificial intelligence to expand the number of MUEs as well. So we are like the original MUEs, which come from a story written, not by AI, but Britain by a talent network of like, uh, roughly 20, 25 female writers in web stream. And, uh, and we have, like, for example, you know, screenwriter from Netflix, we have like nonbinary writers. We have like a huge amount of representation in the, in the, among the writers. And then like the AI painted news is, which were trained. Basically we trained a model like specific styles of like painter painters that we love for the project, for example, Mo uh, clay, uh, if line and, and so on.
Speaker 1: So we generated other kind of like 899 sisters, AI sisters to the original newses. And then we gave every single, uh, AI painted newses an AI Bri story. But in this case, we kind of like, uh, um, you know, like, uh, uh, did a much, much, uh, more precise you like learning in terms of like, uh, how we train the model, you know, like how we train the entire, like literary model for the, for the project. And therefore, so, uh, uh, the, the, the story components in both projects, uh, is actually, is such that, uh, it's not only about the character. So every character has a story, but there's also a super story to the project, meaning that every single project has a master story, but that master story is not decided by us centrally as a studio, but it's decided by everyone that holds one of the characters in a way holding a character is kind of like an executive producer pass, decide where the story goes in the first place.
Speaker 1: Right? So the idea is that, for example, in mu, at the end of the project, we have like these 999 newses with 999 stories, and we run this sort, like hunger games competition, where we ask people, you know, what, uh, let's vote, what are the five coolest stories that resonate the most with you out of the entire drop and those five MUEs with their five stories, they become the protagonist of a graphic, novel, and FD that we as a studio make. And then we give to the entire community so that we basically like you as a holder, no longer only hold an NFT, like a cart, uh, with, with also like an attached story. But you also hold a piece of media built on top of the project. So that's, that's, you know, like the beauty of it it's, uh, it's, uh, uh, think about that.
Speaker 1: Like, if you like of the blue sheet projects, like board date, or world of women and so on, they are all becoming other things. They're becoming, uh, Netflix series. They are becoming documentaries, they're becoming brands. And the beauty about like, you know, like working with tennis pieces that I don't like board date 2100 can be a t-shirt brand, but board date 79 and Android can be the next Ironman. It's up to the story to decide that, and that's the beauty of it. So this kind of like sums up our
Speaker 1: Board date and so on. So what we realize is that, uh, uh, we already own the IP at the indeed level of the character. So what if, uh, for example, one of our next original series who can be, uh, is gonna be a comic book, an NB comic book that tries to answer the question, what happens if a board date meets a crypto bank that meets a word of women in a, this topic, New York city in 2300, right. And the idea is that it's gonna be kind of like this noir thriller story. And, uh, but the beauty is that, uh, we own the APS, uh, of the digital characters. And we are just building a super story, a narrative bridge to connect all of them. And that's what I mean by non-original story. So the visuals are not our, but this story is, you know, like it's, uh, it's made together, uh, together by us and by our network of, uh, creators as well.
Speaker 1: And lastly, our, our tour bucket, it's what we call our meta projects, which are not stories. They are new ways of telling stories. For example, one of the upcoming ones, which is gonna be live, uh, between may and June, it's a platform that's gonna be called, uh, untold. And this platform, what it does, it, uh, allows you to create plots, to create stories using AI and meet them as an Ft. So what we are doing is like, uh, imagine you on a website and it asks you, okay, like pick a genre, like love story, pick a format, like, uh, video game or film, or, you know, like novel and so on. And then pick a couple of random words, for example, table and light. And you can, uh, it keeps generating stories and stories and stories for you because we are training the model on extremely high quality narrative content.
Speaker 1: And, uh, the idea is that, like these stories that you like the most, the ones that resonate the most with you, you can actually meet them as an FTS. And we are partnering with, uh, three different law firms at an international level to make sure that, uh, uh, whatever gets, uh, produced by the AI engine is specific enough to be, uh, protected from an intellectual property rights perspective at an international level. So in a way it's kind of like building a marketplace for ideas and this, this platform, as you see it, it's not a story. So it's a new ways of building stories and it works more like, but like for people, uh, and for projects, cause we're gonna be giving access to the, to these platforms, like to projects that, uh, kind of like wanna build their own AI stories. Cause we will receive lots and lots of requests like by different like NFT projects that they simply stated could be, can you do what you're doing for your projects for us?
Speaker 1: Can you generate stories for us? And, um, and uh, either like written by our network of artists or written by our, you know, like AI engine as well, and to conclude the idea is that, uh, I think like impossible is extremely well positioned to, to do that, like to, to address like these, uh, kind of like need like for stories in the Webre market. And, uh, basically everything we are, we are doing right now is we are doing it as a studio, you know, like, so it's the studio deciding the roadmap of the next project, like going live, but going for like in the next maybe four or five months, we're gonna be operating as a, as a Dow as well. So it's, and the main difference that like, uh, uh, the Dow is gonna be kind of like a double sided platform where on the one hand you have the creators.
Speaker 1: So like our storytellers, our, uh, artists, and, and so on that, uh, are really like a few hundreds come of, like from all around the planets. And, uh, on the other hand, you have fans, the people like reading, watching, listening to the stories and, uh, the main difference that right now we are deciding the roadmap as a studio, but going forward, it's gonna be the downs. Like everyone is gonna be deciding what are gonna be the next projects. And additionally, if, uh, you, uh, if, if you're part of the Dow, you're gonna be one having like a presale access, like for the future impossible project. So it's gonna be like a new fleet story ecosystem in a, in a way two, you're gonna be accessing the meta projects like I'm told. So for example, I'm told, is a project that is public to access. So you can actually like, uh, go and generate as many stories as you want.
Speaker 1: But if you wanna min them as an Ft, you need to be part of the Dow. You need to have like a Dow access, uh, in order to, to use it like to min uh, plots. And, uh, and also the third part thing, which I think is gonna be like the most relevant going forward is that, uh, uh, imagine that the room is gonna like be made by 10 different projects and you fall in love with project number. I don't like three, which is gonna be a movie at the centralized movie. What you can do is actually on the specific project. Uh, you can mean what we call like an executive producer pass them, uh, which makes you busy. Like it has like two main consequences consequence. Number one is that, uh, um, uh, the, uh, you become, uh, decision maker in the plot you become, uh, you decide, do we go a, or do we go B, do we go C?
Speaker 1: Or do we go B and, and so on. So you kind of like become a, basically a decision maker, a producer total effect. And on the second hand, you become a producer because you, because you become kind of like an equity shareholder into the specific media, into the film itself. So you help fund the movie, but you also like get parts of like the revenue that the movie's gonna be doing going forward. So in a way it's a new way also, not just like to produce art, but also to like, you know, like prefund it in the first place. Uh, so yeah, that, that's why I think like one, there's a huge space, like for, you know, like story based, uh, architectures, ecosystems, organizations, dos going forward, and sure, it's, it's a blue right now. Like this space, uh, the D I P market, the licensing market is like, uh, uh, 10 different tenfold, bigger than DFP ones. And right now there's a huge space bridging the shoe, bridging the NFP market and bridging DP market. And that's what we are trying to do.
Speaker 3: Wow. Yeah. I mean that, that's the, that's a great overview. I've got a few follow up questions. Um, but, but James, I'd love kind of an artist perspective here based on what you heard a, any follow on questions or thoughts for Alessandro before I start to pick apart this really cool brand.
Speaker 4: Yeah. Aandra I do have a question. Um, I looked at the, um, I looked at the project, uh, mu users, and it's obvious that the AI, um, generator is doing really well at creating these, um, these beautiful portraits in a style that's somewhere between impressionism and, and, uh, an abstract. Um, mm-hmm
Speaker 1: Yes, no, absolutely. So, first of all, um, uh, the AI role depends on the project. So it uses there there's two, actually there's three different AI engine. So one is for the, uh, image generation. One is for the image enhancement and like for the up scaling side. And one is like for the tech side. So like, uh, the visual the, the way. So first of all, to answer your question, there's a huge degree of like human creation in, uh, especially like in the individual equation and in the store creation. There's a lot of like human creation involved. Uh, second of all, the, um, so in the individual side, what we did is like, uh, we started like from the original news is the black and white ones that kind of like work as a negatives in a film in a, in a way we redrew them, uh, as like simple sketches in order, like to make them like understandable or better, more understandable by, by the engine, by the Dcon engine.
Speaker 1: And what we did is like, we, uh, produce different several different learnings based on specific different like styles. And it was a, uh, it was not, not exactly like my call. It was the call of like the main artist of the project since this project is a collaboration with, with which is the main artist to decide which, uh, particular painters to do the, the study, the learning on. And I was like, uh, as I mentioned, like moly, it decline where like, some of the references, uh, used there, but again, like out of, uh, 899, uh, AI painted uses, I guess we generate like a few thousands, you know, and then the rest was kind of like a process, like discarding the ones that were not really legible from, uh, shapes perspective and like, from a color perspective, like from, uh, you know, like, uh, all the different, uh, uh, the, the different kind of like, uh, um, outlines that we give to the engine. OK. I see.
Speaker 4: So you, so, so you and your team will, um, you and your team will go through the, uh, the, um, the, the portraits that the engine creates and you'll exclude the ones that aren't palatable. Is that correct?
Speaker 1: Yes. Yes. It's, there's a level of human creation there. Uh, and roughly, yeah, I guess like probably would generate like a few, maybe two, 3003, 4,000 in order to, to generate like 899, uh, final AI painted music. So, yeah, that was like more of a, um, like the, the, the judge, there were actually two layers of like cleaning. So layer number one is that, uh, how understandable, uh, is the original shape, the original MUEs in the AI painting redrawn in a particular style. So it was like two, let's say not understandable. It was like discarded. And then the second one was like a more kind of like artistic judgment, like from the main artist, you know, like, uh, which ones to pick, uh, given that they were like already, uh, you know, like, uh, uh, understandable, like from, from a shape and kind of like lines perspective.
Speaker 1: And, uh, on the, um, the AI announcement side, it was like pretty much straightforward. It was like just an AI, like to announce the quality of like the overall paintings. And, uh, the, I think like lots of like, the work was like done on the story side, because the, I did it like in order to generate 899 stories, uh, since it was like something I, I hand hand that created myself as well, I had to generate like roughly between six and 7,000. So it was like roughly one in eight was, was saved up. And the way it worked is that there were like two different, uh, learnings there. So learning number one is that we trained the model, not on specific styles, not on specific, not on ES or Hermans on specific writers. We trained the model on the building blocks of stories. We kind of like went back to what are the most important theories in story creation in morph of stories in the world.
Speaker 1: So we started like, from, I don't like Vladimir props, morphology of detail. Uh, we started like from Chris Anderson, the journey of the hero and, and, and so on. So kind of like the, the structures that, uh, uh, narrate how a story is built, like the three act of like a classical Hollywood movie or, uh, the 31 functions of the story, but by prop and so on. So the idea is that, um, we started like, from those, those elements, and then we like narrowed them down in like the core elements that, um, in those like five, 10 lines of like a story that the, the engine needed to generate needed to be there. So for example, a story needs to have a character or needs to have like an intention. So you want to achieve something and a story needs to have an obstacle. So someone or something is preventing you from achieving what you want.
Speaker 1: And that, that kind of like, uh, uh, you know, like, um, conflict between like, uh, uh, intentional and obstacle creates the, is the engine that moves the story forward. Right. And this was like, used, like to, um, create the back story, like the skeleton of like, lots of like the stories that were created. But again, like most, uh, many of them like were not simply plots. Like this was like a, um, um, a type of like reasoning and type of like learning that was used once the, uh, format was like the, one of a plot, but like, in other cases where like, uh, letters where like, uh, poems where like, uh, uh, different things, because the idea was like, this story element of ause in this case was not supposed to be like only narrative, but was supposed to be a portray in space and time of the MUEs.
Speaker 1: So it could be a plot with all the different, like, elements, like, uh, it could be like a sci-fi intercellular saga where she's the protagonist, or it could be just a moment in space and time, like a four lines description of, uh, one of her afternoons, uh, in, uh, in the house. And that will be enough because you, you know, like you already like, see an element that makes her alive in the first place. So again, the, the, that was like the first step of learning. The second one was kind of like, uh, uh, based on like, context. So for example, every uses as a name. So I don't like ELs Ali and so on. And we use the name as one of the inputs to, um, like in the engine. Another was like the format, another was like degenerate and a few others were kind of like the, uh, kind of like randomized the S that we were using.
Speaker 1: So, um, in, in a way it's kind of like, uh, the pre, uh, when I was talking about the untold platform, it's kind of like the manual version in a way of the untold platform, because we were like, for example, giving us a prompt into the engine, a bright, uh, short story in decipher in general about Ellie, uh, with a woman, uh, with, uh, X, uh, um, the characteristic characteristic or, or features, and, uh, about these and try to integrate these two particular keywords, like table and light, you know, and then we're like generating and generating and generating like lots of like different, uh, outputs until we got like one that was, uh, uh, specific enough. And, uh, that we liked as well, like in order to, uh, associate what MUEs as well. So it was kind of like, again, it was, uh, lots of like technology involved and lots of like human creation involved, but what we are trying to do right now after also the, the new project is that, uh, we are building a much more automated process, whereas like these, uh, kind of like learning has been, uh, we're basically like re-putting the stories produced for the new project as part of the learning, so that, uh, the next wave of generated results are even like more precise or even like more narrative as well.
Speaker 1: So every single project, like act as a, sort of like a reinforcement learning for the engine that we're working on. So yeah. That's wow. That's what we're doing.
Speaker 3: Amazing. So if I kind of like, I mean, you mentioned a couple key things, like you started out, you're a story tech company, um, mm-hmm
Speaker 1: Yeah. Yes. And so we use the, the, let's say the untold platform is not public yet because it's, uh, I think we just finished like the front end side and we're connecting like the, the, the back end with the AI engine that we, we trained so far, but yeah, ideally it's, uh, the, the, the mechanisms was the same. So like all the learnings that we did, like behind the scenes in this months where already implemented the project by project, you can actually really like, see the difference. Like, what's the quality of like plot generations between one project and the other. And, uh, the next way is getting even like more precise. We, we use the, let's say the, the not, uh, the pre untold platform, uh, behind the scenes, uh, um, built on kind of like using like the digital three data engine, uh, open AI platforms to start from.
Speaker 1: And, uh, and from there, like on going forward, like in the next couple of months, we want to make sure that that platform gets, uh, uh, can actually be accessed by, you know, other projects and other people as well. So it's not just gonna be like for, for us as a studio tool to make stories. Cause the idea is that, uh, one of our key, uh, pillars is that, uh, uh, next wave of like stories is gonna be decentralized also in the way it gets generated. So the good thing about, uh, I'm told is that, uh, if you create one of the coolest stories of the word, but you're not a pretty person, you're not like a brighter, you're not a screenwriter. You can still monetize that story by giving it to someone which is more expert than you are at making that story, uh, you know, like into a more complex, medium.
Speaker 1: So the, is that like, uh, we are working on like, uh, uh, also like building different types of licensing, whether it's like, uh, from a CCC zero perspective, like more complex licenses in order to make sure that, um, everyone gets recognized along the way. So that's, you know, like the, the, of it. And, uh, cause right now, like we, we live word in which like, uh, we, we, we were born with the assumption of like stories and ideas are not protectable in a, in a way, like, because you can just like change a comma and, uh, it, it makes it like completely different. So in, in a way we were trying like to figure out a way where, you know, like people are not like just, uh, E egoistically protecting their stories. They will never write
Speaker 1: For example, uh, personally, like, I, I love stories since I'm a child, right. I used to like write novels when I was like 80 years old on my painting to computer during the summers with my brother and my cousin. So the, the, the, I love stories since I'm a child, I, I wrote like a few books as well. And, uh, the, uh, I have like a folder which contains hundreds of like story plot, hundreds of like story ideas. And first of all, I, I can tell you, like, I recently upgraded that folder to contain a lot of AI written stories that I was like exploring like for completely different from the project, like, uh, that I was exploring for myself. Um, uh, while I was training the model, because some of them are, they could be the plot of the next Christopher, no movie. Some of them are incredible are really incredible.
Speaker 1: And the second thing is that, uh, again, I will never be able to write hundreds of like, no, like 1000, 2000 novice by myself for like 2000 scripts by myself. So, uh, in a way there's a huge amount of like story plots or story ideas that could have a life, you know, so I'm, I'm trying like to, with, I'm told with like the, the following platforms that we are, we are building, we are trying to understand how to make sure that, uh, there's always an incentive for people to be creative and, uh, like still see some, some revenue like from, from there. Cause idea is that there's a radio market for this. I mean, think about like the, the entire fun fiction phenomenon in, in the first place. And also the, these spins from the fun fiction phenomenon, like, uh, for, and there there's cases, like from, I don't like some super famous movies that came out from fun fiction.
Speaker 1: Right? Yeah. And it is that, uh, uh, the thing is that, uh, I think one of the problems with fiction is that, uh, you need to create the quality cause you can find like, uh, gems, but you can find also like, uh, um, projects that have like know literally value at all. So, uh, that's also one of the ways in which we, uh, the, the way which we're structuring it out is that we are posing a lot of like attention to the, uh, creative side level. So what's the level of the artist we want, like to partner with the tier a artists, we don't care how famous they are. We, we only look at how good they, they, they could be. So also one of the missions for, for, for impossible platform is that we, we partner with underrepresented artists, we partner with, uh, uh, under, uh, represented voices that are still unheard in the space that they have a huge amount of talent to, to be on like, uh, uh, to do something in, in Webre.
Speaker 1: And Webre is like a way, you know, like better space to, to, to, to start, like, in terms of like speed and like a community, uh, than like nutrition, like media media, you know, like
Speaker 1: And, uh, if your work spoke for, for itself, it had a house already at a community. And, um, so yeah, that, that's what we're trying to do, you know, like, so we are putting a lot of attention to, who's gonna be the artist. Who's gonna be the centralized network of artists, like building a different projects, but, uh, that's why we didn't wanna start, like with the do right away, we build the Dow behind the sensor right now we are a manual do because we have hundreds of creators all around the planet, but we spend that year working with them, selecting them, uh, testing each other, liking each other and, and so on. So that's, uh, yeah, that, that's what we did
Speaker 3: Amazing. Um, I love that you you've described what, what I would play back as a value chain in intellectual property that starts with anyone really interested in storytelling. And so, from what I understand through the untold platform, which is now powering your current projects, um, you know, community member could go in generates some ideas. They really like them work with the AI to maybe spin them up into more story narrative, leveraging some of the best, um, theories and storytelling find a professional writer, artist in your community, um, possibly publish that in your community as an NFT on your platforms and everyone maintaining their piece of the intellectual property on the value chain. Is that sort of way to describe what your experience might be?
Speaker 1: Yes, yes, no, exactly. So we are, we are still, um, trying to understand, like from a blockchain perspective, whether the digital, like blocks of stories are gonna be treated either as an FDS or as a, a year 1155. So like as the, the, or like the different kinda like side of FTS, because the, the main point there is that, um, it's, uh, a similar part. I think it's like gonna be fashion, you know, like, uh, you can have like, uh, that there are people like building digital, uh, clothing, digital, like cotton digital, um, material that are later gonna be used in, uh, garments later are gonna be used in, uh, uh, more complex creations and more complex, uh, kind of like works of, uh, um, works of fashion. And the same thing is gonna be applied to stories. So if you really love, like my story and you are better at me at writing, you can take that story and amplify and make like something even more, uh, you know, like complicated and more like big as a media as well.
Speaker 1: So the idea is that it's gonna be either an NFT part of a bigger NFD, or it's gonna be, uh, near 1155 part of a final NFD, uh, movie or film or series, whatever DB format is gonna be is gonna be like, you know, so that's, um, yeah, so, but the, the, the chain of value is gonna be, you know, respected. So if you were the one that the were kind of like the first one, kind of like to generate that story using AI or connecting those very, very far away words, uh, with your inputs, then definitely like you should be like the one being recognized for that in the first place. So, yeah.
Speaker 3: Wow. So almost, you know, uh, your platform could be, um, a way to just simply at the very basic protect the intellectual property of my ideas.
Speaker 1: Yes. So that's why we are, so the, the, the complicated side is that, uh, uh, in the tradition world, uh, there are things that are protectable and things that are not. And, uh, there's for example, when you, when you try to protect like a plot, it's a similar reasoning is like, uh, I think about like television formats, like who wants to be a millionaire, or like, you know, like similar formats, they are protectable. And they are like one page of like maybe five, 10 nights describing a format describing like a quiz describing like the TV show itself. And that's protectable. And once it's protectable, you can license it, you can syndicate it. And once it's accessible in one country, you can like sell it in other 190 countries, you know? So it's, it's a way like television made money, like through the nineties and the, and the first and like two thousands.
Speaker 1: But it is like with plots, it's harder because it can just like, take a plot and like modify just a little bit. And, uh, it still, like, you know, like gets recognized like something new and there's no like, uh, uh, and there's kind of like, no, uh, no, no appeal there, because there there's no case. So what we are trying to do is like, we are applying like blockchain technologies where, whereas you have a timestamp, uh, of, uh, creation of that like particular story, and we are trying to make them specific enough to be protectable. So the, the idea is that, uh, of course you are never gonna be solving the problem completely because if you take the core idea, the core engine of an, of a story, and you rebuild it and reapply in different ways that then it becomes something else. But again, it's, um, it's, it's a decentralized mechanism.
Speaker 1: So it's in a way it's similar to look, we wanna like, uh, use those stories, those like plots of elevator pitches in a way like of, of a movie of a tray of film and on to build an entire movement of the centralized storytelling as well. And the idea is that, uh, um, so for example, uh, the, the do structure, which is like, where we're going towards is only fundamental because, uh, like via process, like century decided by a studio, otherwise we become, you know, like Disney or Dreamworks or Pixar where the movies of the next 10 years get decided by 10 people at last floor of, uh, skyscraper doesn't make sense anymore. But at the same time, you cannot like leave it entirely decentralized because, uh, it's, uh, it's gonna be chaos as well. So what we're trying to do, our, our Dow is gonna, we, we just like spent, um, eight weeks in the, uh, sit cloud accelerator, like, which is one of the biggest accelerators in the world for, for Dows and, uh, talking like almost every day, like about the governments do structures.
Speaker 1: And one of the key takeaways is that like Dows are dynamic governments and, um, they should change depending on the type of decision that you actually calling in into, uh, into objects. So, for example, if I, if I need to vote, uh, what's gonna be the 3d artist for a specific project. Doesn't make sense that I open this decision to everyone, because I'm not an expert in TRID end. You're not an expert in TRID end, doesn't make sense for us to, to decide that. So it's gonna be like a, more of a technical committee team decision. And at the same time, like one, one, um, for bigger things, like we, we need to decide the roadmap. That can be like something that can be like, uh, decided by, by everyone. So kind of like more so we're, we are isolating between like monarchy in a way, an empire and, uh, uh, kind of like an Enlighted form of like communism in a, in a way.
Speaker 1: And, uh, and, uh, again, the type of like government keeps changing depending on like what's type of decision and what's life cycle of the decision as well. And, um, yeah, again, like for, for example, like, uh, uh, one of the best systems that we're probably gonna be abusing is like, kind of like, uh, we're adopting the short runner system of like traditional television, uh, production companies right now where the sh runner is usually the head writer, or usually the person coming up with the idea and he, or she's gonna be overseeing a team of brighter that like, uh, time by time is gonna like be producing whether it's chapters, whether it's episodes, whether it's, you know, like, uh, modular works of the bigger, like work of art. And at the same time there's gonna be, uh, you know, like other quad in the process that decide other things in the centralized way. So for example, who's gonna be, uh, the core team, uh, deciding. So for this type of project, we need like a storyboard artist, uh, visual designer, um, uh, screenwriter, we, we need like, you know, like, um, comic book artist, and the end format is gonna be that of a comic book, for example. So the idea is that, uh, we, we are trying to find like the optimal point between like decent centralization and centralization in creating work so far, because none of the two extremes, uh, works, uh, or works efficient things. So, yeah.
Speaker 3: What was the name of the Dow accelerator? Just for the audience you said? Uh,
Speaker 1: Uh, sit club. Sit club.
Speaker 3: Oh yeah. Okay. Yes. I've uh, yeah, I, uh, I did check out one of their, um, oh yeah. When you guys were, when yeah. When, when impossible, uh, I think you were featuring proof of story or impossible on, on there. Yeah, absolutely.
Speaker 1: We were, yes. We, we possible impossible. Okay,
Speaker 3: Great.
Speaker 4: For, yeah, go ahead. More Aandra for, um, capable and traditional writers. Can they use your AI engine for any other utility? Can they let's say if they get writer's block, can they throw in a couple of chapters and,
Speaker 1: Yeah, so that that's actually a separate platform from, so I'm told, is a platform that does one thing. And, uh, and that's like generating stories in AI that that's all what it does. It's specialize in that we are having like, uh, so I was actually like summarizing all the project that we're doing right now. They're like 30 between platforms and original non tourism meta project that we're like tackling between and next year. And one of them is specific for that. It's, it's an AI, I'd say like an AI research slash brighter assistant, you know? So in a way it's, it's an, an assistant that like, once one hits you, like with, for example, brighter box, or not just like brighter, but like how to continue in the story. Right. You get like stuck in a specific dialogue, you gets stuck in a specific, um, type of like interaction.
Speaker 1: And it has to like, get out of like that, uh, that mess. But secondly, it kind of like helps, like for example, creating like backgrounds, uh, creating like descriptions, creating like character back stories as well, in a way, and like, make sure that everything adds up. You know, like, for example, when I write, I I'm like, uh, my style of like writings, usually, like I start with an idea, I met entire the entire kind of like roadmap of the novel. So from chapter one to chapter 30, and then by chapter 10, I'm writing a completely different novel, right. So there there's like a, an emergent kind of like storytelling happening, happening there. Cause, um, I really believe, I think it was by, by foul that the sentences that, uh, that, that quote that, uh, it is only when characters and events begin to disobey you, they truly begin to live.
Speaker 1: And I really believe that it's, it's, uh, you need to let them disobey. Otherwise they becomes like an artificial plot, you know, in a, in a, in a way. And, uh, and readers are extremely smart. Readers can pick it up like right, right away. So in, in a way we are, yeah, this tool, it doesn't have a name yet. It's gonna be a separate platform, but it's gonna be using AI to kind of like, be a sort of like writer's best friend, you know, like a writer's best companion. And, you know, like how big writers have like entire research teams, so like Wils meet or Steven King or whatever you have, like, um, lots of people like researching like historical facts and like, uh, specific locations. And imagine you, you wanna like set the, um, novel in a place, I dunno, in South Africa, in Cape town that you never visited and you like to have like specific ideas of like, uh, uh, you know, for example, a route through the city or like specific like, uh, um, bars and restaurant that you can actually name in order, like to make, uh, your narrative lie more, you know, like subtle in a way, because it looks like, uh, as if you, you have been there in the first place, so it's, yeah, it's, it's a writer's best companion.
Speaker 1: We are. Uh, it's gonna it's we are, um, gonna be working on that probably like between, uh, June and July. So it's still like, gonna be pretty, pretty short, uh, uh, we we're working at full speed, so it's, uh, uh, it's lot of like projects all at the same time, but yeah, there's gonna be like lots of like things for, for writers as well.
Speaker 4: Do you think it could go as far as saying, like your example of South Africa, do you think it could go as far as saying what the air smell like in a particular month, in a particular part of the world? Could it be very specific?
Speaker 1: Yes. It depends on the training that you're gonna be applying to that, but if you wanna, that's, that's a beauty of it. You can train specific functions. So like, if I wanna train, uh, the, uh, model on, uh, for example, what you just said, like, for example, I wanna, uh, summarize a city by it, stars or summarize a specific feeling or a specific climate, or a specific light that you get in a specific city. You can train that specific function and like, uh, make sure that, uh, it gets like, uh, extremely realistic results as well. And, uh, so yeah, it's, uh, usually like what, what we do is like, uh, it's really hard to train, like, uh, thousands of functions. So at the same time, so we start like, with a package of like the most important things, like, uh, uh, the anti block function, the character backstory function, the, uh, description enhancement function, because you're not really good on descriptions.
Speaker 1: So you kind of like amplifies them for you. So it's in a way it's, it's a, it's a Zas plus, uh, uh, best friend plus like a research assistant plus, uh, um, yeah, like, uh, an expert, like in different like realms, like whether it's a, a climate guy, whether it's like the, the, you know, like the, the Soma guy and, and so on, like all in, in one, you know, so it's, uh, wow. Yes. So that's, uh, I mean, I was, I was like building already this because I, I need it as well. Cause I know, like when I write, it's usually like either I do something very sequential, like where I spend months researching something, then I gather all of that research and I try not to use it, but very, very, very subtly in the novel or whatever I write. But, uh, it's also very inefficient because you spend months and then you spend like months writing and you go back and then you will lost something in research. And, uh, it's like a constant like back and forth, uh, process. But what if you can do it like, uh, there's not pre-production production, post production, there's just production, you know, that's, that's the beauty of it. And, um, yeah, so it's actually already in, in the pipeline.
Speaker 3: So I guess, you know, cause as, as I kind of wrap my head around your story, you know, one of the things I was thinking you you've got a lot, a lot of projects, you said you've got 30, sounds like a combination between building some of the story tech, uh, and the AI functions. Um, but also testing and learning a lot of, um, different stories and reteaching your models and improving your models. Cuz to me, obviously, you know, the race for AI is very much gonna be well, how quickly can you build the smartest best model? Uh, if you wanna be the, the, the AI of storytelling to go to, you've kind of gotta have a, a volume of stories built and told to teach the models. So mm-hmm
Speaker 1: Yes. In a way you can see all the original and non-original projects that we do as a, uh, learning playgrounds for the AI in a way. So you, you can, I mean, they are already monetizable because they are projects. They are like full on projects, like sold as and Ft with their own roadmaps and, and so on. But at the same time, the main effect is that they advance our AI as well. So it's, uh, all the hundreds, sorry, all the thousands of stories that we, we produced between pro story and, and the music that they made our, uh, faces in, in AI, like a lot, like, uh, better than they were like, uh, just like four or five months ago. And, uh, at the same time they are project, they monetize like, and they attract a community. So again, the, the, the thing is that like in web 3d hardest, and I think aspect right now, it's not like the, uh, tech or creative, how your
Speaker 1: Cause right now there's, uh, I dunno, thousands senses of thousands of projects that the market is extremely saturated. It uh it's uh, and it's also like, uh, uh, in some cases can look like a very unhealthy environment in a way, because it's, uh, it lives like on, on hype and it can attract like very bad actors at the same time. Like, uh, yeah, for example, what we're trying to do is like, we are trying to be like, uh, uh, not crazy big communities, but like communities of people that could be passionate about, uh, the mission of every project. And I bring you project is kind of like a separate standalone project, you know, like they have their own websites, they have their own search, they have their own community, but the, the idea is that, uh, uh, we don't wanna attract like a community of like flippers and, you know, like, uh, and sheers and so on, just because it's, uh, yeah, maybe in the short round you win, because like you get like lots of money, like how to like drops and stuff like that.
Speaker 1: Like in the longer, longer term, you're not building something sustainable. So even the entire philosophy, like working towards a drop, I think like from, uh, after the music drop, we're gonna be, uh, um, kind of like abandoning the, you build towards a drop structure for a more ever green kind of like, uh, um, look, this is a product, this is a project. And it's gonna be there like for the next upcoming months, which is way more health, you know, like don't just like have to be like for months and months and must and months. And then it becomes kind of like an emotional, physical rollercoaster all the time, which I don't think it's sustainable in a three long term. And especially cause if, uh, it's a model that's born to generate and sustain hype and usually the type mask, the fact that, uh, your project, uh, uh, doesn't have like a lot to stand on in, in its lives, you know?
Speaker 1: So like the only way to market it is like through time and like formal and like, uh, you know, so it, it's not sustainable long term. And that's why we are, uh, again, like NFD are gonna be one piece of, uh, what we are doing. Like the rest is gonna be like, uh, uh, between like, um, traditional licensing, like, uh, how, the stories that we build become other things like how, the stories that we build spill over, how do they become, uh, movies and books and, you know, like other things and the NFD is gonna become just a little part of that. And at the same time it's gonna be, uh, we don't just wanna write the stories. We're not like build the tools to invent the machination of stories together. And, uh, uh, I come like from, from, I mean, as I mentioned, I love stories in Ima child.
Speaker 1: The stories are like the things that, uh, I'm passionate, I'm most passionate about. And, uh, uh, the, I think there's a huge space right now, uh, to tell stories in, uh, a different way where like, it's not like an AI versus human kind of like, uh, um, conflict. It's an AI plus human where the AI component kind of like helps you boost, whatever work works you already doing, you know, and in a way I think like the cool part again, I was extremely skeptic at the beginning was when I was like, yeah, okay. Let's see this AI and let's see what it can do. And after like a few months, I was like, wow. Like in, in some cases it can really, really, really like, uh, uh, basically connect. So such like far away words that you will never like get there in, in the first place.
Speaker 1: And that's like one thing. So like the result horizontal, um, kind of like, just like, uh, exposition of like different like concepts, uh, that an AI can actually like do working like better than, than a person, unless you like spend all of your days like doing that. But the second thing is just like the way of like, uh, thinking and the way of writing sometimes is like different. Like we, we reason, like in a, kind of like a horizontal diagram and vertical type of like language where, uh, our horizontal language is like, uh, what, when, who it's written by questions like our diagonal one is how and the vertical one is why. So the, our type, our language also includes our, the way we think, like the, the most kind of like vertical question web is like, is a why, but like sometimes what if you, you train the, uh, an AI language, sorry, an AI, um, model, not just on stories, but like on more complex particles, for example, what if we push the boundaries of language to AI by pushing the boundaries of language, we push the boundaries of thinking. So for example, what if we build, uh, complex particle, like the why of how of a when? And we see that in action in a text generated by AI in order to, uh, produce more complex cows and consequence connections in the way, a character things in another, right. It's super complicated, but it's, it's also beautiful because it's, uh, it's, uh, it can train ways of writing that do not exist in right now. That's, that's the beauty of that
Speaker 3: Based on the complex minds of a human.
Speaker 1: Yes, exactly. But you build from there and then, uh, you, so for, for example, like, uh, uh, one, one thing I know, I remember like this story specif there, there was these, uh, uh, I was stringed like the, the, the AI, like on a, on a dialogue, I said, like just was very, very, um, wide kind of like input. It was like, just write a dialogue between Skar, something like very, very broad. And I remember like, it started like generating this dialogue between two very, one, very angry and one very sad charact, like one girl and one guy. And, uh, it's like back to back, like they keep like, uh, screaming at each other. They keep like, uh, uh, hating each other and loving each other. And only in the very last line, you just read like this maybe 20, 25 lines of like dialogue, and only in the last line, you realize that they are father and daughter. And so whoa, that changes like everything like that changes like my entire of the dialogue. And that's a, a way of writing that I never witnessed before. Like this kind of like, uh, uh, did the, you can see like the turning point in the thinking from the AI in a single word or in a single sentence. And you have like, these kind of like turning points, which, which more like in a, in a non-human way. And that's also like super be it's a little bit scary, but also super beautiful.
Speaker 4: It is a little bit scary. Um, but it's also really, it's also, uh, really thought provoking. Do you think that the AI could, um, could assist us thinking in different philosophical ways as well as generating stories?
Speaker 1: Uh, yes. I mean, philosophy is, is about asking questions, right. And, uh, uh, if you manage to, to ask those questions in, in, um, in a novel way, uh, then yes, I, I can see like, no, no reasons why, why not? So I think we're a little bit like still far from, from there, but, uh, uh, I can really see like in just how
Speaker 4: You could have suggestions for resolutions to global problems or political tensions.
Speaker 1: Yes. I mean, the, the, again, it's all about what data you fit into a model. So like the, the models we're using are like models that, uh, uh, in their general shape, they were trained on the entirety of the internet in, in a way so on, uh, gigabytes and gigabytes and gigabytes of like, text that coming from all over the planets. You know, it's kind of like a mirror of what what's, uh, happening right now in the world, but, uh, yeah, again, depends on like, what type of training you wanna, you wanna do it's uh, um, the, the trainings that we are doing are specifically on either authors or like on, uh, story structures and story morphologies in, in the first place, uh, if, uh, the type of like, uh, training you wanna do is like on the philosophy of like platform niche, and, uh, you know, like all the, the big, the big pH through, through history.
Speaker 1: Uh, and for example, what, what if you take the entire, uh, I don't like pH like works of like, uh, haggle or can, and like, uh, the biggest ones and you ask an AI to either complete Orite and let's see what, what happens, you know, like it's, it's, uh, I dunno, probably like probably a few people have already done that. It's, it's not our Fouse, but, uh, uh, yeah, again, it's, it's a tool. It's the, the, I think the bottom line is that, um, AI is, is what in late in, they call a box media. It's not something bad, it's not something good. It's something that, uh, has power. And it's where you bring that power. You direct that power that, uh, uh, ultimately it's gonna be decid, whether it's like a good thing or a bad thing in the first place. And, uh, so yeah, it's in a way it's up to us as well,
Speaker 3: And I'm really excited, um, Alessandro to see impossible, um, tapping that power and directing that power to, to storytelling of the future. So, uh, congratulations, um, you know, your vision is brilliant, exciting. Um, you're executing, um, on that project, uh, a number of projects to bring this to life. So, um, you know, certainly I'm quite excited. I think it's safe to assume our audience is gonna be excited. So, you know, how, how in a nutshell, could, could folks get involved and, um, participate in your communities and projects, ultimately, you know, um, untold as a platform being available to the public is gonna make what is driving and empowering and possible projects today available to everyone. So tell us the story of how, how, how folks can get involved.
Speaker 1: Mm-hmm
Speaker 1: But again, like news it's, it's a project. Like it's a standalone project that like, has a lots of like, you know, like beauty and lots of, kind of like value in, um, uh, just aside, like from the AI component, you know, it's, uh, it has like, it's a mission it's it's about like women empowering in, in arts. So it's, um, I think it has like lots of, um, uh, reasons to, to, to, to be there in, in the first place. And, uh, yeah. And then the second one is, uh, I mean, if, uh, if you check like, uh, the, uh, impossible either Twitter, like, or, um, or, uh, websites, it's where we are gonna be posting the, all the different outlines regarding when the meta projects, not just like original projects are gonna be lives. For example, I'm told, and I'm told as, as I mentioned, it's gonna be like, probably like maybe still one month behind the scenes before it's kinda like ready to be deployed, uh, to the public. Uh, but yeah, every, every single information is gonna go through either the, uh, website and attach newsletter or through the impossible Twitter.
Speaker 3: Awesome. Great. Well, we'll definitely, uh, encourage our audience and our community to check it out. Um, yeah. So, um, any, any final closing thoughts? I've got a couple, two, two more, uh, fun questions to close the conversation, but Alessandra wanted to just see if you had any closing thoughts, um, as we wrap up.
Speaker 1: No, I mean, the, the closing thought that probably is kind of like the, the central, you know, engine of everything that we are doing is what we call the, uh, Mickey mouse serum. And in a way, uh, in, in a sentence can actually be, uh, said like this, a story spill over if something is born as a comic book in 1928, like Mickey mouse in 1970, like Ironman, or in 19 96 97, like Harry Potter in a, as a, as a book, it doesn't mean it's gonna be dying as a comic book or as a novel, it's gonna be evolving. It's gonna become other things. And that's because one, the stories were great and should the communities around them were equally, equally strong and they managed to navigate and bring, uh, and BA surf with the story for decades to make it evolve and make it like, become an emperor.
Speaker 1: So in a way, the same reason it applies to NFTs. Something is born as an FD doesn't necessarily mean it's gonna die as an NFD can become other things can become movies and series and, and meta versus, and, you know, like it's up to us and it's up to the story. And that's why the story component it's, it's kind of like the most, uh, I think important one, because at the end, like we are stories exist for, for one, uh, I think like, uh, evolution, evolutionally like kind of like reason, which is, uh, people need to resonate with one which with each other, we, we don't wanna feel lonely. We wanna like, feel connected somehow. And stories are bridges among people are bridges among cultures are bridges among, uh, you know, like tribes and, and countries. And, and in the end, that's what they do. They connect things. And, uh, and, uh, I guess, like, there's a huge space going forward to connect. Like, not just like, uh, Webre projects, like, but, but it was like Webre with the rest of the world, because we have a lot to say so on. So yeah. Yeah.
Speaker 3: I love it because you answered the first question of the two final surprise questions, which was NFTs and web three fatter enduring. And I think everything you said in your closing remarks paints a picture of, um, a very enduring, um, well, a legacy, if you will. Um, so the final fun question is, um, given your launching a Dow in the not too distant future, how long do you think it will be before a Dow accepts a mainstream award? Like an academy award or an Emmy
Speaker 1: Oh, wow. 20 years. So
Speaker 1: I'm pretty sure it's months away. Uh, not, not years away
Speaker 3: Love it. That's it. Um, Alexandro, congratulations again. And, uh, a huge, thanks on behalf of story prima, um, for joining us on the story first podcast, um, been an absolute pleasure to have you on the show. So I wanted to say thank you and, uh, to, to you and the audience, um, have a great day, and we'll close that there.
Speaker 1: Thank you very much. Thanks for, for having me and have a super good day as well.
Speaker 3: Cheers.