Fantastic Futures 2024 - Day 1 - Session 13

This transcript was generated by NFSA Bowerbird and may contain errors.

And I would like to start by acknowledging that Aboriginal land is a memory land and it always was and it always will be Aboriginal land. My name is Ripley, I'm a curator here at the National Sound and Film Archive of Australia and I'm part of the research and development team here. The focus of the Fantastic Futures conference is exploring the opportunities and challenges that AI poses to the cultural sector. I'm very excited to share some work that we've been doing as institutional collaboration that sits in this space. You are about to watch a great short little film that shares the collaborative project that we've worked on together, and it's good fun. As many of you know, there's a pervasive discourse of fear and apprehension surrounding the application of machine learning and AI technologies, especially in the context of cultural creation, curation, and the archives. While it's important not to diminish or delegitimize these concerns, this project really revealed to us and demystified the application of these tools can actually take place in really creative, interesting, and intriguing ways. It's not all doom and gloom. So we'll throw on our little film for you. We've brought together the National Institute of Dramatic Arts, the School of Cybernetics at ANU, with the National Film and Sound Archive because all of us are really concerned about what the future might look like when it comes to the role of AI in the creation of culture. We don't want to let it happen to us. We really want to be part of designing that future together. This workshop is an understanding of one another's expertise and craft and to see how machine learning could be used in this project. Where one is sort of the keeper of culture, one is a maker of culture and then there's a school of cybernetics as well where their interest is really in sort of like how humanity and technology intersect. The NFSA itself has been thinking about how it can use machine learning to make its items more discoverable. There's this really beautiful Venn diagram that happens when you put these three institutions together, which are really guardians and creators of culture. That overlapping interest kind of was the seed of what has become these workshops and these incredible outputs. So there's a mixture of creative practitioners, curators, technologists, designers, theoreticians, academics, and then people who just are really passionate about doing good work now, so we're prepared for what might come. Here we are again, AI is the thing. There is quite a lot of fear around artificial intelligence in general from creatives. Yeah, there's a lot of messed up stuff that's going on so far in the training of the models that are out there. And this is very legitimate fears. We're not trying to stop those concerns. We want to be able to interrogate this technology but we also want to be able to find uses for this technology that feel sound and ethical because it's not going away. But in an ideal world we also have some ideas that are generated out of this workshop. The creative process has been a two-day workshop split over three days. And I think that it's been really rich in terms of slowly stepping people through lots of different exercises and getting people to all be on the same page and generate really interesting results. You're just free to come up with crazy, wonderful, brilliant, kooky, quirky ideas. You'd be able to step into that. You could do that through virtual reality. All of this material will create this sort of highly experimental and interesting video. There is emotionality to all of these objects except for this one. Bringing together these people in this context has led to things that none of us expected and we have all been so excited by what we have found at the National Film and Sound Archive. The people, the things, the film gives us the opportunity to build something together that we can then share with other people so that they can experience it in person. We've had these kind of concrete objects from the archive to look at, the films, the articles that we all kind of got yesterday, but then we've got this vision of how can we transform them, what stories are in there, what stories are missing, and then just mashing together various ideas about how we can go forward. I think one of the most important things I'm going to be taking away is just being the power of collaboration and how interesting and exciting the unknown can be as well that it's not necessarily something to be feared but when we can understand it a lot better we can harness it. Even just being able to be in a room with people who actually know how to put it into practice you realise that it's not this big scary thing that you see in movies you know it can be a thing of practical use to be able to help human beings and really the issue that we should be focusing on is how do we take ownership of that. And kind of demystifying it as sort of a tool rather than sort of a big scary unknown. It's up to us to be part of the conversation and shape how those technologies are used and implemented. to keep the human at the center of it all. I'll never lose sight of the human that will be using these technologies. And I think everyone that I've met in this experience is really, really conscious of that. There's actually a lot of really great, ethical, creative people who are using it to kind of better our society. And that makes me hopeful. It really feels like this is a unique circumstance that hasn't really been done before and I think that there is going to be a lot of things that can be taken back to each institution. What we're really hoping is the start of a really beautiful collaboration and friendship. I'm Beth, I'm the head of NIDA Future Centre. I'm sorry it's really awkward because you just saw a lot of me, so I promise I will make this very short and very sweet. So the NIDA Future Centre has a mandate to educate, explore and experiment with emerging technology and future thinking. We want to be active participants in the development of these technologies and give the next generation of creative practitioners the thinking and the tools that are needed to create a positive and meaningful future for the entertainment industry. Now you saw that we did the three-day workshop and there were a lot of great ideas that came out of it, but the most important thing for us was just this deep desire for all three institutions to continue collaborating. So we decided to see if we could take one of our seeds of an idea and turn it into reality. During the workshop, we had the opportunity to tour the NFSA's conservation labs. That, in combination with reviewing archival films and images, led to a discussion on how AI and machine learning might be used to help restore damaged or even contextualize archival films and media which have little to no information. However, how might we as curators, developers, conservationists, and creative practitioners help to bring these archives back to the forefront of the Australian consciousness, restore meaning, and weave new stories with them? What we are working on now is an interactive installation entitled Imaginative Restoration, a physical and playful embodiment of this concept of restoration and augmentation of archival film. Ripley and their team were tasked with finding the job of finding a collection of badly damaged films to act as a foundation of our project. What emerged was a collection of disintegrating scenes of the Australian landscape. We used these films to ask ourselves, could humans working with AI add to these scenes to provide context and meaning? Could they create their own meanings? Being from NIDA, we wanted to set the scene a little, a story to tell, as it were. And yes, we are all that dramatic. NIDA's team of writers, of which Lily, one of our writers, is here today, video designers and prop makers, came up with the following story world and physical manifestation. Humanity has retreated underground to escape increasingly inhospitable surface conditions. Here, in subterranean grottoes, the storytellers safeguard fragments of the past. But they don't merely preserve these artifacts. They breathe new life into them through a process called imaginative restoration. As a storyteller in the rewilding division, you work to dream up and repopulate the scenes by hand drawing the creatures you can imagine. You will see them enter the footage of the film, adding colour to the black and white scenes of the past. Hi, I'm Charlotte. I'm a researcher and PhD student at the School of Cybernetics. Together with Ben Swift, I've been developing an interactive experience that weaves together the various elements we've just discussed. Ah, this is a great shot of Marcelo working on our sandstone grotto. We've been building out a machine learning application using open source models that take user input in the form of a hand-drawn sketch, reinterprets that input and interpolates it into archival footage. Say a user sketches a butterfly. Our app first recognizes that sketch as a butterfly and then uses that drawing as input into a text prompt to generate a new version of the sketch. This new version of the sketch is then released into the footage alongside the input of previous users. And you've just seen a preview of what that looks like, though of course we're still wrangling the technology. Putting CUDA versions aside, one of the greatest challenges for us has been resisting the tendency of generative AI to over-determine everything it touches. Our collaborators from NIDA and the NFSA looked to us from the School of Cybernetics to demystify AI, to explain the ideas and concepts lurking behind technical jargon. And likewise, we look to them to guide us through the art and science of curation and storytelling. And a big part of this project has been about finding a shared language to negotiate this really difficult balancing act of genuine collaboration. And it has required us to look for shared experiences and concepts. Here's that fly. Western philosophy and critical theory is famously suspicious of theatricality. as it is of technology, and also of the forms of mechanical reproduction that the NFSA is dedicated to preserving. I like to think that the emergent campiness of our collaboration, along with the strong queer contingent on the collaboration crew, is the result of the shared theatricality of stage props and generative prompts, which both invite suspicion and derision, as well as pleasure and enjoyment. We're still working on the project. It's due to launch in November 2024, at which point it will spend some time at each of our institutions, and we really hope that some of you will have an opportunity to interact with it then. I think I'm the last thing between you and a drink, so with that I will leave it. Thank you very much.