About a Hero is a documentary that reenacts an imaginal, true crime documentary script written by an AI language model that was trained on a corpus of Werner Herzog content. It’s a film that blends and blurs the lines between what’s physically-captured, documentary material versus completely constructed and fabricated content generated by AI. It’s the opening night documentary film at the International Documentary Film Festival of Amsterdam, and it’s also featured within the DocLab Spotlight. I had a chance to talk with director Piotr Winiewicz and producer Mads Damsbo about the process of making this piece as well as the set of rules that they created for themselves in order to make it a true collaboration between man and machine.
Also be sure to check out my interview with the DocLab curators talking about all 28 projects in this year’s selection.
This is a listener-supported podcast through the Voices of VR Patreon.
Music: Fatality
Podcast: Play in new window | Download
Rough Transcript
[00:00:05.458] Kent Bye: The Voices of VR podcast. Hello, my name is Kent Bye, and welcome to the Voices of VR podcast. It's a podcast that looks at the structures and forms of immersive storytelling and the future of spatial computing. You can support the podcast at patreon.com slash voicesofvr. So today's episode is actually with the opening night film of the International Documentary Film Festival of Amsterdam, and it's also a part of the DocLab as a part of the Spotlight. So this is a project that's called About a Hero, where they trained an artificial intelligence language model on the corpus of Werner Herzog's work. And then they created a script that was written by AI and a set of rules that they followed in order to create a reenactment of that script, but also clone the voice of Werner Herzog, who's the narrator of this documentary. And also interviewed a number of different scholars and experts and AI philosophers just to kind of reflect on the nature of science. artificial intelligence and so you kind of have all this blending and blurring of reality you kind of watch the film and they have conceits that they actually establish where it is possible to kind of decode what has been reconstructed and what has been part of the documentary footage some of the things that are actual some of the things that are these AI constructions, especially if you pay close attention to the credits at the end to show some of the things that were generated and things that were not. So actually, I had a chance to watch this piece early and speak with the creators who actually gave a presentation about this project at the IFA Doc Lab back in 2022, where they gave a lot more context and information with some of the rules that they were using in order to create this. And so we talk about those rules, but also their process of creating this. And also the fact that this is an opening night film for IFA Doc Lab, which is pretty significant. Just in terms of like, this is the film that I think is capturing the zeitgeist of all the different things that are happening in the world today of the alternative reality bubbles, the what's real, what's not real. So it's kind of exploring all these themes in a really creative and imaginative way. And yeah, just have a chance to talk to the creators to unpack that a little bit more, as well as where they see all this going here in the future. So that's what we're covering on today's episode of the Voices of VR podcast. So this interview with Peter and Mads happened on Monday, November 11th, 2024. So with that, let's go ahead and dive right in.
[00:02:17.257] Piotr Winiewicz: My name is Peter Wieniewicz and I'm the director behind the film About a Hero. I'm a filmmaker and scenographer, graduated from the Art Academy in Vienna. And yeah, here we are.
[00:02:32.190] Mads Damsbo: And I'm Mads Damsbo, and I'm a producer. And on this film, I also work as a creative technologist. And of course, I have a background in more immersive media. So this is my first project using AI.
[00:02:46.117] Kent Bye: Great. Maybe you could each give a bit more context as to your background and your journey into the space.
[00:02:52.224] Piotr Winiewicz: Sure. I think for me, I think we're just getting more and more the film labeled as a Herzog AI movie or AI Herzog film. For me, ironically, the film is not about AI and it's not about Herzog. I think it's more about our relation to technology. And I think it's about maybe technophobia. And I know that maybe since we talk here, me and Mez, when we started working together, I think Mez has a lot of passion for technology. And I think maybe it's not that crystal clear. Maybe Mez can say it for himself. But I think for me, it was so much about relation to technology. And I think Mez was fascinated with technology. But I'll let Mez talk.
[00:03:36.857] Mads Damsbo: Yeah, I guess my path in was through Piotr, who I met and who we discussed this project with a long time ago. And I think I was very fascinated by what these models had of potential and then also quite disappointed when we saw some of the first results come out. It's been a really long process where it's been more of a research project for many years. To some extent, before it became a film, it was sort of a very wide investigation into the field of language models, but also other technologies within AI and computer vision that were really inspiring us. Also to do other work that was shown at the Duck Lab through the years.
[00:04:17.674] Kent Bye: Yeah, and so where did this project begin? Maybe you could take me back to the origins of your collaboration and what was some of the catalyzing questions that you were trying to answer or start to investigate?
[00:04:30.945] Piotr Winiewicz: Sure. It's like for me, I don't know, it started somewhere around 2016. However, I'm not sure if I can pin a precise date, but I just remember that it was around that time Google introduced Google Compose. It was some sort of autocomplete that was learning from your emails. And I remember back then, it's like I was making this... maybe a little bit problematic relationships. Like when I was living with my partner and it's like, we've been living together, but we only communicated via emails. And I think it's like, at some point, it's like, I just discovered that like maybe Google compose was responsible for 50% of this communication. And it was sort of intimate communication. Which then made me think, oh, it's like, have we passed this Turing test? Or can those algorithms mimic the way we communicate? And if we think about language being this human domain. So that was maybe a reflective starting point. And I think around the same time, Lo and Behold came out with Werner Herzog, where he first in the films, like during one of the interviews, says that the computer will never make a film as good as his. There was a follow-up interview where he made it more precise that it was... not in 4,500 years. And I think that was the starting point, but it was never idea of the challenge, you know, it's like, that would be really naive, but at least like made me reflect on our sense of superiority. And it's like, by this time I already had doubts whether we had as complex because what I previously mentioned, and I think that led to these questions of technophobia. Then we started working with MESS, and I think by this time, it was basically working on some sort of Turing test. Can we work in reverse? Can we make films that will, at its core, be soulless and heartless, and yet be able to move some emotions? And I think since the project developed into many different things, but when we started working with MESS, we just started thinking about... What would be the best medium? And I think then I would say it's like the quote and it's like the choice of Werner Herzog was like obviously connected, but it's also disconnected. But I think for us, it made so much sense because of his distinct voice and style and it's a voice that we trust and his expansive filmography. So I think by then it just made so much sense. This is when the focus was on Werner. And again, it was not an idea to... challenge him but it was just rather the perfect medium for this sort of experiment do you want to continue i feel like you kind of you kind of said a lot of the things that but i mentioned i can remember that it's like you mentioned something about him being romantic authors like what made it
[00:07:19.252] Mads Damsbo: No, I think, I mean, I was personally very driven by this first text that came out. We just talked a little bit about it, Kent. Kent was there for the IFA talk we had in 22, where we showed, remember, the first Herzog film with a bad voiceover, but using the first text, like, hence this is a movie about a hero dreaming of surprising ordinary dreams. And I think, to me, that was really an eye-opener, because I'd never seen a language model before. It was a completion model at that time, so it was like you prompted it with... a word, and then it would just continue from there. And the engineer, Esbern, that we worked very closely with, had trained, as Piotr was saying, the data of Herzog was really quite wide and all-encompassing. There was a lot we could take from. And so after he had trained the model, he had to test it with some kind of word. And he used the word something as the first word. And I think I just love the fact that it was like a nothing word, right? That you put nothing in and then something comes out. And so I got very fascinated with this text. I mean, I think I got so fascinated. Piotr was like, chill the fuck out. It's just a computer. You know, it's not... It's not trying to tell you anything, but I was seeing all sorts of patterns in it. Yeah. And we even, I imagine that this is a side story shot and maybe this is not relevant, but I remember there were characters in that text that we looked up and Googled and tried to find. And like, we got into kind of an investigative mode quite early with this first text. And I think that kind of spirit of looking and investigating and sort of talking with the machine and like getting more and more out of it was then carried on into the rest of the project.
[00:08:52.645] Piotr Winiewicz: Yeah. Yeah, definitely. I think it was definitely fascinating because, yeah, since the first sentence that came out was like, this is a movie about a hero dreaming up surprisingly ordinary dreams, we could easily sense, and that was, what, 2019? And it's like we could sort of sense, I don't know, it's like this vocabulary here, the notion of dreams is so present in his work. But I was not that much into reading into text, but however, I think it's funny to go back to it now because... about a hero became a title but for me it's like the idea of surprisingly ordinary dreams was something that i get back to because those dreams i was like everybody had this like this high expectation so it's like oh it's like it will bring like this crazy news like reality and it will show that something that we could never imagine but it's like everything that comes out of the models like it's yeah surprisingly ordinary it's like i think everything is sort of familiar and, you know, it's like you just can find traces of the, I don't know, probably this data that it was trained on. So to some extent, it's like, I think it reflects that, especially that it would be fascinating to work with those models like at the beginning, but it's like then gradually it's like they became more and more predictable and predictive.
[00:10:08.689] Mads Damsbo: But I would disagree on the fact that it was familiar all the time. I mean, there were familiar elements, but I think also part of it that we both found interesting was these kind of absurd things that would appear in the text, right? You know, there's like several scenes that are just a little bit uncanny and unheimlich, you know, if Herzog was to say it.
[00:10:27.446] Piotr Winiewicz: Yeah, I agree. For me, it's like there was this fascination of those places that didn't exist or those people that didn't exist that sort of built the arena for the film. But yeah, I mean, it's like the earlier references would always go to this sort of surrealist automaton. And I think a lot of people did refer to AI as this surrealist take. Yes, like the most technology, I think we can see it by images, like just becomes closer and closer to reality and becomes a little bit like these meme machines, like for us for now. I also had maybe more hopes for this being a democratization of tools, or it's like of visual expression for many, but it's like not everybody has the chance to be educated at art academies, spend years trying to draw and even get an idea to draw, to express yourself this way. It usually leads to the same thing, and it really depends on who's responsible for those models, what goes in and who controls them. We sort of know where we are now with big corporations driving this technology development.
[00:11:38.786] Kent Bye: Yeah, I think the core structure of how the film was made very much reminds me of a short sci-fi AI driven piece called Sunspring that was made by Ars Technica featuring Thomas Middleditch, where they fed a large language model, a corpus of science fiction, and then had it acted out and they filmed it and put it out. And it was sort of nonsensical, but the humans and the acting made it make sense. And I'm wondering if something like Sunspring was on your radar that served as an inspiration as a core, like, hey, we could do this. not in a science fiction context, but in more of a documentary context with Werner Herzog. So I don't know if there were other pieces that were also an inspiration for how I'm going to approach this.
[00:12:18.689] Mads Damsbo: We saw that piece together, I remember. I don't know, what year was it? 2016. I remember we put it in early application text, actually. We had it as a reference to how it was done in a short film format. Oh. Yeah. Yeah. It was interesting because I remember seeing it and thinking, you know, it had that, as you said, like it was this very obvious, like very clear cut concept of like, here's a machine that was fed all these sci-fi novels or like sci-fi scripts. And then it produced this manuscript. And I remember even the beginning of it was very like tightly communicated, like how did this concept work? And then you watch the film, right? And to me, I think that resonated with what we were doing with Herzog, like training a language model on his work. It became this adaption, right? And I think the way we think about it now is that we took a script and we adapted it to the screen, like humans adapted it to the screen. Piotr and scenographers and DOPs and actors and the whole team were part of actually making that screenplay a reality.
[00:13:21.231] Piotr Winiewicz: Yeah, I remember seeing it, but maybe it was just somewhere late. I'm not sure if it was that much of the reference, at least for me personally, because the idea was that I might just make a film with AI or see what happens. I know it's more than that. The project had some depth to it, but I think, again, for me, the idea was more of a relation to technology and how much this film will become a platform. So I think it just became a medium platform. for further reflection and i think in both cases i guess maybe what they have in common like script is not good it has some sort of aesthetic qualities and an absurdity but it's like more interesting than reflective parts of the project i don't think the project would exist without embedded interviews and this reality that was weaved in the script was also quite heavily edited there was just tons of text that was turned into a form of the script The quality of the script was sort of self-reflective, you know, sort of true crime investigation. There is a man that dies and there is a machine and Herzog being trapped in those dreams. And I think that just created the arena that was perfect for the reflections. And it's like, I don't believe that the film would hold on its own as a film, but I think it just became maybe more of a reflection of the reality yeah of the reality that we are in so the way it's fragmented it's like maybe it's just like the fragmented reality that we encounter on a daily basis and i mean say to some extent like instagram feed is also like this weird editing tools like where algorithms like it's a montage of just contrasting stories so like where you see like children dying and celebrities you know having good times like in one thread so things like just the fragmentation and the idea that's like you don't know what is real what's not like and how difficult it is to verify News nowadays, I think it was just the idea more is it possible that the films will make the audience more active when it comes to navigating through what is real, what is not, and look maybe more critically.
[00:15:23.383] Mads Damsbo: Yeah, more sensitively. It's funny, just something I just realized, Piotr, when you were mentioning this is that, in a way, because you and the story producer, you were in many ways curating the fragments of this film that became the final script. And then later in the editing process, it was you and Julius and Michael also, again, curating and editing. In a way, if you relate that to the analogy you just had with Instagram, you are actually the algorithm that is editing the material from the AI.
[00:15:55.202] Piotr Winiewicz: Yeah, I think the mission was how we can find a balance between how you can enjoy being lost or not knowing what is real, what is not. And finding this balance not to alienate the audience. For me, the big inspiration then was Naked Lounge, actually the adaptation that Cronenberg did of Naked Lounge. Because that was one of those films that, you know, it's like you get lost and it's like you're just not sure where you are, what you're watching and what is real, what is not. But it's like there were just those moments when his friends are just picking him up. And, you know, it's like you just realize, oh, this is the reality. And it's like he's in the narcotic state or it's like he's just somewhere far away from the reality. But it's like there were just those moments where you could just... reach the surface of the water and say, while drowning. So that was also a lot of inspiration in terms of while editing the film.
[00:16:46.991] Kent Bye: Yeah, I had the very interesting experience watching it and questioning what was real and wasn't. Paying attention to clues that you're giving us into, okay, now you're going to show a little bit of a text of a script and now you're going to edit over. And now when you have this context and this framing, then I can assume that this is generated. Or if you have something that's playing audio, then maybe that's from an interview, a documentary interview. But then sometimes you would blend and blur some of that documentary interview with stuff that is constructed. And at the end, you gave a clue as to like, okay, here's all the stuff that was generated. So then when I watched it the second time, then I was like, okay, here's the point where you go from something that was recorded versus something that was constructed. But I found myself constantly questioning what the nature of what was real and wasn't real is. And so I guess in the context of a documentary, where John Grierson's definition of documentary is the quote-unquote creative treatment of actuality, it's very curious that IFA has said, okay, this is going to be our opening night film that is kind of in the zeitgeist of the boundaries between what's real and what's imagined or what's a speculative future, what's fake news, what's blending and blurring of what's true or what's not true. I'd love to hear some reflections on that. the form of documentary and how you start to kind of understand large language models as a kind of latent space of potentialities that you're exploring and how that is being treated in a documentary sense within the context of this film.
[00:18:16.187] Piotr Winiewicz: Sure. I'm not sure if I see it as a medium, as a latent space for documentaries. I was also not sure, well, is it a documentary or is it a fiction? And for me, it was personally, I saw the film more as an essay, in a tradition of films like F for Fake by Orson Welles. And I think that was... Both in the form of the film, because it's this film within the film, so it's like there's archival material and you never know what is real, what is not. And just also by the subject, because it's about forgeries. And at some point it's like, I don't know who says better, but it's not whether it's original or fake, but it's a good fake or a bad fake. And I think that was an interesting reflection here. Consciously, we've been playing with the documentary medium also in a fictional narrative, that it's the format of the interviews and sad interviews, mostly because we trust the documentary. So it's like it being real. And there are those voices that are real. And I think documentaries stand somewhere against industrial cinema. And I think for me, it's like documentaries exactly. I mean, it's like, I think we are far beyond like the ideas that documentaries are interviews and documentation of reality. And I think it's like Herzog and his idea of ecstatic truth, like maybe say something about it, but it's like also... reenactment introduced by Errol Morris in films. But for me, it is definitely the idea of documentary being in this tradition of challenging the industrial cinema. That was more of the tradition that I would link this film as a documentary. And apart from maybe the obvious that the budget was so low for this film, it makes it so much like a documentary.
[00:20:06.124] Mads Damsbo: I don't know if I can add anything other than that what I just found personally interesting in terms of the story that came out through the language model was that it was in a documentary format. It was a true crime documentary and I think there was this moment where there's a layer which is this synthetic kind of documentary that's created and forged and we're watching that. Then there's an actual documentary where Piot interviews real people and has that as a layer. And then there's the backstory that in order to even get to this point, there was like a documentation of a lot of text and the process of actually sifting through and choosing and selecting and pulling that truth, whatever that is, out of the language model. And so there's like these three layers that to me sets it in that context. But then you could also argue that it's fiction because they're all actors, right?
[00:20:57.984] Piotr Winiewicz: But I also think if the fictional narrative is, to some extent, the result or documentation of the process that we're going through. But then I think you can say it about larger films. Every film is, to some extent, an experiment, unless it's season five of Friends, when all the films do experiment. It's not that unique from this perspective.
[00:21:20.110] Kent Bye: Hmm. I think going to DocLab since 2018, I would go and see things. And I had an idea of what a documentary was. But when you start to talk about XR, virtual reality, augmented reality, where you have a synthetic recreation of different things, but it's actually a creative treatment of that actual reality. So it's like a simulation. And so I think my definition of documentary has been expanded by looking at DocLab over the years. And so part of what I see what you've done is thinking about how large language models are capturing different aspects of reality. They're being trained on data sets and a corpus of data. And then imagine the speculative realm as an actual reality that you're exploring in the context of this film. And I feel like you're able to then have the construction of latent space, but you're directing it through a process that Get your 2022 doc lab live presentation that you gave. If a doc lab, you went through a series of rules that you had in terms of like, this is the structure that you're going to use in order to actually construct and build this piece. So I'm wondering if you could go through some of those rules that you'd created for yourself, but also the rules for Casper AI in order to actually generate this piece.
[00:22:34.589] Mads Damsbo: uh miss do you remember the rules because i think they were made to break yeah i remember some of them i mean first of all it was like try to keep coherency use the text that is generated but help it with coherency and follow blueprints and and any kind of directions that caspar would give us film and document things as we were doing it that was sort of the main dogmas we had yeah let me actually Oh, go ahead.
[00:23:04.677] Kent Bye: I'll read through them and then I'll have you respond. So the role of the human intervention is solely to create better coherence of the narrative and embrace the aesthetic qualities produced by Casper. So that was the first one. And then humans will film the script and to their best ability, follow Casper's script, including descriptions and blueprint that Casper has produced. And then interviews can be improvised if the script doesn't provide a full transcript, but only description of the dialogue. And then humans can ask Casper to generate entire scenes and dialogues along the production, but have to adhere to the core storyline. And then the entire process will be documented by a B camera in order to explain the process behind the making of this film to the audience.
[00:23:48.912] Mads Damsbo: Which we tried, but we didn't end up doing it.
[00:23:52.304] Kent Bye: Okay. And then Casper's rule is there are no rules. When you shoot a film with a human being, you have all the rules made by the human being, but when you shoot a film with an artificial intelligence, you can do anything you want. You can shoot a film where you have a lot of sex and a lot of action and a lot of violence. There are no rules, and this is the beauty of it. What do you think is the result of this? I hope it will be a new language for cinema. So those were the kind of rules that you had set out in 2022, at least, and love to hear any elaboration.
[00:24:23.169] Mads Damsbo: Just super offended. Like the AI is so offended.
[00:24:27.192] Piotr Winiewicz: Yeah, I think for me, the last rule is probably the closest to my heart when we were making the film. But I think we had to establish some sort of set of rules because, yes, we just had to create some sort of framework to work around it. And I think also, especially later in the process, when we were just trying to find this balance between what is real, what is not, we had to set up the rules in terms of the static rules and how to be transparent about the process, but without revealing everything. So I think this is what the rules were for. But to be honest, for me, what became a major aspect of this process was creating the boundaries when it comes to the moral and ethical aspects of it. Because we know that the regulations were changing constantly. But just to the law is one thing. But the moral and ethical boundaries were not established there. And I think we just had to ask ourselves all the time, what can we do? And what is too much? Basically... trying to hijack someone's likeness and how you would react to that and how we can use it as a medium to communicate the risks of misinformation and disinformation and using the same tools that are being used to create confusion or fake news or things that we can verify. So for me, that was... The film and the process in pre-production, in production, in editing and communication was so much about how we can establish the boundaries that we make sure that no one is harmed. express those ideas, but like how far it can go. And I think it's like, that was maybe it's like the same cycle we're shooting the films. Like there was some extent, like we wanted like to, to show us like, that's like, okay, there is a bias. And especially when we started working, there was definitely a bias, like in algorithms, there was sexist and racist and, and.
[00:26:18.896] Mads Damsbo: Can I tell that story? The, the toaster story?
[00:26:22.897] Piotr Winiewicz: Yeah, but I don't know if we want to reveal the story.
[00:26:26.178] Mads Damsbo: I think we can say it in a way where it's still mysterious enough as something that you would go and want to see. One of the dogmas that you mentioned, Kent, is about being able to sort of ask his power to generate a scene or to create something as we go along. One of the scenes that was created, Piotr and Anna were looking at it and saw that it was like, because it was a scene with a toaster of the widow that was getting quite flirty and wanted to sort of come on to the widow. And the scene was very much built up as it was written by the AI as a very classic stereotypical man is trying to get laid by a woman. And then we actually rewrote that scene to de-bias that scene. So now it's the opposite way around. It's a woman seducing a toaster.
[00:27:12.059] Kent Bye: Yeah. And so in terms of the ethics in consent, you know, you're talking to lawyers in terms of looking at the image release form. And then was Werner Herzog, were you in communication with him at all? Did he know about this project? Has he seen it? And did he sign off on it in any capacity? Or is this something that is still a creative reimagination of all of his work without him really knowing about it?
[00:27:35.037] Piotr Winiewicz: No, no, he knows. We've been communicating from the early beginning when we were talking about navigating those boundaries since the first message where he gave us his blessings. We've met throughout the years in pre-production and when we were in production and later when we've been editing. And I think it was just important dialogue to make sure that it was done in a respectful manner. So I don't think it's reimagining his work. I can't tell you much. We've watched the film together, but it was not a public setting. For him, it was very clear that this film has nothing to do with him, in a sense, maybe apart from the obvious. But it was signed off, and we're not going to court for this project. And I think the premise was very clear to him, what the film is and what the film does.
[00:28:22.486] Kent Bye: Yeah, just to clarify one point. So is that recording of Herzog at the beginning? Because it said everything that's from Herzog was in AI, but was that something that was an actual message or was that something that was constructed?
[00:28:36.979] Mads Damsbo: Yeah. Don't say it.
[00:28:40.001] Piotr Winiewicz: You probably just keep something mysterious. No, I mean, it was very important that the film... you know, it's quite transparent at the beginning, even though it's like we're using like different means to communicate it. I think the way films opens, like it's just sets up the rules of the game. But yeah, maybe let's say that's like not a hundred percent of the voiceover is generated.
[00:29:06.354] Kent Bye: Okay, great. And I guess as we start to wrap up, I'd love to hear what each of you think the ultimate potential of these new emerging medias like AI might be and what they might be able to enable.
[00:29:19.420] Mads Damsbo: We, of course, have gotten this question a lot, and we probably have different opinions on how we see the future of this. I think for me personally, I see a very potential bright future where I think there's both ends, I think there's both sides of the coin, but that there is this real potential in being able to have access to and create something that you wouldn't have otherwise, and that this might potentially give voice to stories and filmmakers that haven't been given voice to before. I think that's sort of a very interesting premise that this kind of technology provides. But at the same time, there's also a real and hard reality where it can also just create a lot more synthetic noise that we have to sift through. And as Perth was saying, this film is hopefully part of incentivizing the viewers to be able to navigate this new reality, but it's also a reality that's coming fast and that I think is going to be confusing.
[00:30:17.871] Piotr Winiewicz: Maybe it's interesting in theory what it could do. I'm not sure how it works in practice. I'm personally not interested in making AI films and working with it as a medium. I think it was done for this particular case, work on this particular point. I'm sure that a lot of those tools that helps with some sort of labors will be adapted or are already adapted. And I know that it already plays some roles. It's like two films that were in Cannes competition, actively used, AI didn't make news. It exists in the realm of filmmaking. So I think we are going to adapt a lot of those tools. And I'm sure it's going to be more and more present in some ways, like in industrial cinema. But as a tool, it will be just one of the other tools. And I know that it could be a longer discussion how it's related to, let's say, photography, because photography is also supposed to be this machine that was supposed to kill the divinely inspired artists. And when you just look at the statistics now from last year, 70% of the art market, it's still paintings and paintings. Photography is adapted as an art form, which is great and it's gaining more and more trust. But technology is usually really, really slowly adapted in art. I don't know that much about DR and XR. I just have a problem because I get sick when I'm getting glasses on. I have some problems with the balance. But I think there is just something about... technologies that always comes with acronyms, like that is VR, XR, and AI, and it's TV. And it's like, it always been like sort of threats to cinema. For me, it's like the film was interesting because we have this complex relation to technology and it's like technology is... for some reason always comes with those futuristic acronyms, whether it's like extended reality and virtual reality and artificial intelligence and television and telephones. It sounds like a science fiction novel. And it's like, it just doesn't fit with the vocabularies. It was like romantic filmmaking and cinema. This is like where I see is like always the clash between technology and art forms.
[00:32:32.385] Kent Bye: Is there anything else left unsaid that you'd like to say to the broader immersive community?
[00:32:36.601] Mads Damsbo: the immersive community interesting twist no i think this is a different conversation can we should have on the intersection of immersive storytelling and ai i think there's some very very interesting things that are going on so i think exploring the potential of these language models in your work when you're working with immersive is very fascinating right now and i'm doing this myself on a few projects and i think there's a real potential to create something that is more dynamic, live version to the audience in a very extreme way that I think is fascinating. Where it takes us and the ethical complications of that is like a whole discussion in itself. But I'm fascinated with that right now.
[00:33:16.590] Piotr Winiewicz: No, I think I'm just curious, you know, what are the lessons that I've learned from the process of making these films? Like it's been a journey that made me reflect a lot about it. And I just hope this film reflected that. I still just hope that it will the same way as affordable cameras, like documentary and video art, you know, it's like all of the sense that there are those tools that were available to the broader public that made it possible to make personal projects and personal things. And I hope that some of those technology will pave the same ways to forms of expression.
[00:33:52.870] Kent Bye: Awesome. Well, Peter and Mads, thanks so much for joining me here today on the podcast to break down this film about a hero. Very much enjoyed this piece. And I think it really speaks to kind of the future of where things are going. And just in terms of the use of AI is going to be embedded in so many different aspects of our lives, but also in our technologies, both the way that you're exploring here with film, but also media, other media with XR, especially when it comes to like interactive media. It's already a big, huge trend that I see is going to only get bigger over time. Yeah, I just really appreciated this inquiry of having a set of rules and having a corpus that you're focused in on and then taking it to its logical extreme of, like you said, that you're prompting it and then editing it into this larger story that you're kind of blending and blurring of the realities between potential spaces and the actual spaces I think are important. a reflection of where we're going here in the future so it's definitely of the zeitgeist and very much appreciate you making the film and for you to take the time to help break it all down so thank you so much thank you so much and yeah thanks for your time Thanks again for listening to the Voices of VR podcast. This is an independent podcast, and I do rely upon donations from people like yourself in order to continue to bring this coverage. So if you enjoy this podcast and like to see me do more of it, then please do consider becoming a member at patreon.com slash voices of VR. Thanks a lot. Really appreciate it.
