Autodesk, Inc. (ADSK) Earnings Call Transcript & Summary

Please welcome to the stage the Senior Vice President of Research at Autodesk, Mike Haley.

Welcome, everybody. For as long as humans have lived, we have been creative. Indeed, to be creative is what makes us human. And nobody embodies creativity like Leonardo da Vinci, a prolific polymath and inventor years ahead of his time. His drawings and ideas persist to this day. Now Leonardo existed at an interesting point in history when there was an abundance of time to think deeply about creating a world that had both function and beauty. Now as the world accelerated, we've often assigned creative beauty to a category of optional or costly. Instead of rewarding beauty, we rewarded function. But at the same time, we've seen new challenges arrive at our shores in the forms of climate change, population density, new energy sources and more, even more reasons to ignore beauty. I have a provocation for you. Just as in Leonardo's time, when there was an abundance of time, I suggest that we will have an abundance of creativity. It's not that we'll have more time. Indeed, in media and entertainment, likely even less in the future, but we will be able to do orders of magnitude more creative work in an AI-powered world. With AI, we see a return to a focus on what matters: melding beauty and function, creating devices that you want to hold because they embody the love a designer put into them or building warm environments where people can thrive and connect, and telling an incredible story that touches your audience. Today, we don't get enough time to be creative. We're all caught up in the muck of getting things done. But these changes will give us more time to be creative. Perhaps one of Leonardo's most iconic images was that of the Vitruvian man. This is a family-friendly event, so please excuse my slightly censored version of the original image. AI is pretty good at removing naughty bits from pictures, by the way. It turns out that -- so AI is powering innovation and increasingly disrupting, as you all know, the entire economy. It's happening in every single industry. I'm going to use the Vitruvian man as a guide to take you on a journey to see how designing in a fully AI world will create this abundance of creative energy. So let's start with how we interact with creativity-enabling technology today. We have become so used to keyboards, mice screens and the restrictive inputs of even the most modern computers that we don't question it. Artists and animators learn to draw and then they move to 3D animation and CAD software and they leave their sketchpads behind. Why shouldn't we be able to talk with our computers while quickly sketching out a rough idea and then have a discussion with the computer about that idea? Rather, we are accustomed to reducing ourselves to the level of the computers' capabilities. I suggest that AI is ushering in a new era where our senses and our natural physical actions, much like in Leonardo's times, are what we will use to be more naturally creative. And this is great news for creativity as it means more accessible tools powering greater and greater expression. I'm going to share some examples of our research to illustrate my thesis. We have a rich history of original research at Autodesk. Our industrial research lab is over 20 years old with over 600 research publications, 150 of which were published here at SIGGRAPH or other ACM conferences. Now we've published over 90 scientific papers on AI for CAD and geometry. We strive to share discoveries that advance all the industries we serve. And there is an advantage to having a broad perspective. We are not focused purely on media and entertainment, but on all creators, artists and designers across many industries. As our interest and expertise in artificial intelligence grew, we established a formal AI lab in 2018 to accelerate our fundamental and applied research in AI and machine learning. And last year, we announced our first experimental generative AI model, Project Bernini. Let's take a look at this research to see how creativity might function in an AI-mediated world. In this example, I took my iPad and sketched a car. I don't have great skills, but it took my sketch and it understood the shape of the car in 3 dimensions. It knew I would need 4 wheels. It offered a hollow body, but it also gave an option with a solid shape. It isn't perfect, but that only took a few seconds to generate. In one instance, the model understood that I might want side mirrors. Here, I can communicate with the computer through the very natural method of sketching. Keep in mind that this hasn't been trained on cars. It doesn't understand the discipline of car manufacturing. Instead, it's been trained on shape geometry. We are moving far beyond this now in our research. But already, we are beginning to find that something like simple sketching can fundamentally change the way you create geometry. Natural interactions like these are exciting, they're really exciting. But what if we could go further? The human mind, although amazing, is no match for the computational capabilities of modern computers. And in a world where most design problems are inherently systems problems, involving dozens or even hundreds of variables, perhaps we can do better. AI augmented design can go beyond the interface of the computer to reasoning about the entire solution space, being a brain partner that can explore design variations or instantly test out a range of engineering options. The next example I'm going to show you is from Autodesk's research team as well. This is a site in West Oakland in the Bay Area. It's a site right next to the highway. And as you can see, it's been -- you can see right next to the highway then, it has been set aside for affordable housing. The state of California also wants to make this a low-carbon site. They also want to build it cheaper and they want to build it in half the time that it typically requires. So not too many constraints, right? Kind of faster, cheaper, better sort of sounds like most film and game productions, right? So we started working with a couple of partners to figure out how we could apply AI to this. We built the first AI tool that could figure out all the possible ways to lay out the buildings on that property. You're looking at the different balances of green spaces and orientations of the buildings and more. The software is also considering highway noise, building heights, the accessibility and some materials. You can see all the different sort of balances in that radar chart there. The AI was able to look across all those elements and gather all these variables at the same time. Now this is exactly the kind of thing that's typically hard for a designer to do. Your brain cannot extend across all these variables simultaneously, but the AI could optimize across all of them. Then we started working with a prefabricated module company for this entire building because if you want to do it fast and you wanted to do it cheaply, prefabrication is probably the way to go. So then we went with this company that could build things 40% to 50% faster and make them less expensive. That's their site over there on Mare Island in the North Bay. We could take the fundamental design objectives and explore the full space. Here, you see it exploring the different configurations of the modules into the different buildings. The colors represent different functions of types of modules, 1-bedroom apartment, 2-bedroom apartments, kitchens, stairwells, et cetera. Now notice on the right, it's also computing embodied carbon. It's also optimizing the cost. So now you get a representation that the designer can quickly understand the trade-offs of the variables we just discussed. Currently under construction, here is an artist rendering of the site of what it will look like when it's complete. Now just as a film or game director guides a production, so does an architect, directing many teams and moving parts to match their vision. And if a language model can predict the next word, why not predict the character's next move. So let's turn our attention to your pursuits, your artistry. As AI learns to model behavior from data, it can learn to predict character behavior. We hope to create the next generation of animation software that increases the artist's time to be creative. Here's an example of an animator using Maya to create a simple walk cycle using a process that is not too far from the stop motion or traditional animation. They're grabbing rig controllers. They're posing the character. They're keying those poses. They're going back and forth between their reference and the graph editor, adding more and more poses, playblasting, refining their key frames. I'm not going to play the entire time lapse, but this was a time-consuming exercise for the artist, all to get what's essentially 2 steps that are copied into a walk cycle. So even though this process has been entirely digitized in software rather than with pen and paper, it remains tedious. It takes a lot of effort to achieve that natural movement. In contrast, here's an art-directed animation that was completed in a few minutes in Maya on a production-ready rig with only a handful of key frames. Again, the artist completed this in just a few minutes, and this was done using the work that I'm going to share with you today and a new Maya tool called MotionMaker. We trained this learning tool exclusively on motion capture data from human and canine actors that Autodesk produced. MotionMaker quickly animates characters based on the instructions the animator gives it, including the points in 3D space that the character travels through and within a given time frame. Throughout this process, the artist maintains control over the animation, and they can adjust and tweak the character's initial motion path and rigging. The software rapidly offers an artist a starting point to refine and then enhance. MotionMaker eliminates that tedium. It accelerates your work and it increases your capacity to be more creative, thinking for more options quickly. I hope this gives you some idea of how creativity might play out in the age of AI, how AI may enable interaction with computers that include all of our natural forms of communication. Everything I've shown you is active research here at Autodesk, and MotionMaker is now available in Maya. At some level, like Leonardo da Vinci, we are all researchers: exploring questions, testing hypotheses and sharing discoveries. And now AI will help anybody with a story to tell, quickly explore their characters, test their ideas and share their vision. AI is removing the technical obstacles to human creativity. No more keyboard and mouse thinking, you work and express however you feel most comfortable, and AI meets you there. I truly believe that AI will be transformative for all of you, bringing even more creativity and beauty to your work. AI is another technology. It's another tool. If Leonardo were alive today, I have no doubt that he would be a power user of modern AI tools, allowing his creativity to flow to entirely new levels by leveraging these new natural tools and extending his reasoning capability. Perhaps he might draw the Vitruvian man a little differently. And just as Autodesk concluded canine characters in our MotionMaker tool, perhaps Leonardo would have conceived a Vitruvian dog. Thank you very much. Now it's my pleasure to introduce my colleague from Montreal and Autodesk's talented strategist for Media and Entertainment, Maurice Patel.

Thank you, Mike. Hello, everyone. Every story starts with someone imagining what can be, not just what is. The story starts in your head where your imagination is limitless, but getting it to the screen requires talent, tools and budget, all of which limit what you can do. Yet technology can help you overcome such limitations. Take computers. One story that's always resonated with me is Pixar's Luxo Jr., not just because it was made by computers, but because it's so engaging. I mean, who knew a lamp could be so expressive. Using computers, the talented artists at Pixar then went on to craft the world's first fully CGI feature film, Toy Story. Another highly imaginative story made possible by technology was James Cameron's Avatar. To make it possible, teams at Weta, Lightstorm and Autodesk collaborated to develop the world's first virtual production tools and help bring the Na'vi to life. And it's not just large studios using new technology to tell their stories. I was recently captivated by an upcoming game from a husband and wife duo, Rachel and Aneel Glendenning, and their team at Friday Sundae, a small game studio in England. The original composition of their '80s punk and ska music was what really drew me in because I was a kid in the '80s. It really captivated me. Today, technologies like Unreal Engine and Maya are making it easier for authors like Rachel and Aneel to tell stories with unprecedented levels of creativity. New technology, whether it's computers, real-time engines, help unlock stories. I believe this will also be true of AI. It's arriving right at the point where the industry needs change. It's no secret, our industry is under intense pressure. Since COVID, studios have lost an astounding $30 billion in streaming content. And across just 2 generations of console, the cost of making a AAA game has gone up tenfold from $20 million to $200 million. If those costs were passed on to a consumer, we would be paying $600 a gain. This has led to a vicious cycle of make it or break it with crunch times and layoffs, which just isn't sustainable. The industry needs to change, and it will, whether we want it to or not. Technology will drive that change. You can see it already with procedurally generated content, real-time pipelines and of course, machine learning or AI. What will AI mean for us, for artists? Well, I believe AI will change the industry, just like computers did 30 years ago. Remember, Toy Story. Its success disrupted an entire industry. Hand-drawn 2D animation became computer-generated 3D animation. I myself have changed what I do several times from mining to teaching English to computer graphics. Change is uncomfortable, and it can be downright scary. So let me tell you a little bit about my experience with change. When I started working in visual effects, computers were expensive and limited in what they could do. Back then, most visual effects were practical, miniature models, animatronics, optical compositing. And my first job was teaching people how to use computers instead. So some computers were scary, but quite frankly, they seemed like cheats. I mean, where's the skill in pointing and clicking. To them, it did not feel like a real craft. Yet fast forward to today and now, not using a computer is inconceivable. And while practical effects still exist, computers have propelled visual effects from a niche craft to a career for hundreds and thousands of artists. Teaching taught me that it's the tool, not the task that makes the artist. It's the talent. [ Sinon ] users became Shake users. Shake users became Nuke users. The tools disappeared, but the artists remained. Technology may change the way we work, but it also creates new work and often more of it, which is why I am an AI optimist. I believe in human adaptability, but also in AI's limits as an engineered technology. There are 3 reasons why I believe artists can have a future in a world of AI. One, AI is not human. It can never experience smelling cherry blossoms in the rain. It hallucinates, but it does not dream. It has no emotions, no awareness, no imagination. It only has patterns. It has learned from the data it was trained on. Yet good stories connect precisely because they are born from our imagination and forged by our perspectives and emotions, which is why I believe that to tell good stories, AI will need to be directed by artists. It's not sentient, it's just a tool, albeit a powerful one. Second, art is a craft. We all know nothing is ever final. It's never perfect. My wife is a jeweler. Before she casts in silver or gold, she sculpts in wax. She will spend days, sometimes weeks, fine-tuning a design before she's ready to cast it. To her, it is never perfect. Crafting is a human process of refining a real-world result towards an ideal imaginary state. AI cannot do that. It has no notion of what is ideal, but it can give you more options to hone your craft. I envision a future where content creators will work with AI, much like a movie director works with the cast of actors. Endless takes trying to coach a perfect performance followed by a lot of fix it in post. Yes, AI can and does create a lot of work to fixing things. Finally, I believe AI empowers you by making it easier and cheaper to create content, whether it's for a movie or video game. Smaller teams with smaller budgets will be able to bring bigger projects to life with fewer limitations. AI will become an indispensable tool for all artists, which is why Autodesk has been researching and developing AI tools for more than a decade. Six years ago, we added our first machine learning tools to Autodesk Flame, like Flame's ML Z-Depth Generator, which lets studios like MangoFx isolate an actor and extract depth contours. Here's a cool example where they cast a shadow of a 3D alien onto a person delivering a pizza. MangoFx regularly uses Flame's ML tools on shows like Severance and Pachinko to streamline complex tasks like morphing frames. The founder, Sean Finley says that it cuts down on the guesswork, helping them to get cleaner results in minutes and freeing the team to focus more on creativity. That's creative abundance with the artist in control. The pace of innovation in AI is accelerating, which is why we are adding more and more AI tools to the products you use every day, so you don't get left behind, like MotionMaker, the new AI tool in Maya that Mike talked about earlier. It's already helping incredible animators like Eddie Chew, the owner of Griffin Animation Studios, to do the work they need to do. They did all this in roughly 5 weeks. That's 1/3 of the time it would normally take. We are committed to building AI tools like MotionMaker to help you get more out of the products that you use every day. We know you will be expected to do more for less. It's just how the industry works. Autodesk's goal is to give you the means to do so using the best that AI can offer. I talked about my belief that AI will help more artists tell more stories. Well, we want to make that future possible with Flow Studio. Originally developed by Wonder Dynamics, Flow Studio is designed to be an easy-to-use and accessible AI platform for 3D animation and visual effects for newcomers and industry veterans alike. Wonder Dynamics' approach to developing AI is very similar to Autodesk's: building tools that artists can control and direct. Wonder Dynamics was founded by visual effects supervisor and filmmaker, Nikola Todorovic, and actor and producer, Tye Sheridan, who you will meet in a few minutes. They dream of making their own sci-fi movies and saw firsthand just how hard it is to do limited by budgets and technology, but they never gave up on their dreams. And so Flow Studio was born. Take a look. [Presentation]

I wish I could dance like that. One of Flow Studio's core capabilities is AI motion capture, and it won't be capturing me anytime soon. You upload a video with an actor. Select a 3D character and automatically apply that character to the actor. And Flow Studio integrates it with hand, face and body tracking. No mocap suit and no green screen. Anyone can do it, even me. It's easy for nonexperts who want to get into 3D and useful for experts who want to quickly previs of a shot. What I like most about Flow Studio though, is that it is AI that's editable, directable and controllable. You can export selected project elements like clean plates, camera tracks and alpha masks to Nuke or export a 3DC to Maya, Blender or Unreal. You get to control every little detail. Just like Voxel Studio, who built a pipeline with Flow Studio and Maya. Voxel faced real pressure when Superman and Lois returned to production after the strikes. The final season needed high-end visuals on a tighter schedule and smaller budget. We've all heard that before. Here's how they pressed forward. [Presentation]

Flow Studio, great work, isn't it? Flow Studio helped Voxel deliver the shots on time and on budget. Maya gave them the quality the client demanded. And together, they were unbeatable. But Flow isn't just for studios. It's for anyone aiming for cinematic results without a big budget or a large technical team, like emerging filmmakers and content creators. Many are already exploring what it can do, like Giacomo, a 3D artist from Italy who used Flow Studio with Maya at a school called BigRock to turn a basketball player into a robot. Or Farhad and Faraz from Bad Decision Studio. They use Flow Studio for storyboarding. They shot a fight scene on an iPhone, uploaded it into Flow Studio to generate mocap data and brought it into Unreal Engine to generate the previs of the cinematic short called Haruki. Or Mike, better known as AI Video School on YouTube. He's passionate about making good movies but has no visual effects background. Flow Studio opened a whole new world of storytelling for him. I love seeing what the creator economy can do with Flow Studio. Earlier, Mike spoke about an era of creative abundance, and we don't want that abundance just to be reserved for those with big budgets because storytelling should not be gatekept. So today, we're doing something about it. You can create stunning visual effects with AI you control for free with Flow Studio. Yes. We're introducing new lower-cost pricing for Flow, including a free tier. So whether you're starting out with a story in your head and the phone in your hand or you're an indie filmmaker trying to get a pitch ready, Flow Studio is a great way for you to get your stories out there. Our free pricing is actually a statement of belief that creativity belongs to everyone, that the next great story could come from anywhere and that AI, when built right, can help you tell it. Every time I come to SIGGRAPH, and I won't say how many years that is, it's been a lot, I'm reminded that we are part of a wonderful industry, a community of creatives and risk-takers and dreamers. You have the skills, you have the imagination. And with AI, you will have tools to break down the limits on your imagination and the stories you want to tell. I can't wait to see what you create. Thank you. And up next, we have an amazing group of writers, directors, artists and filmmakers. They're here to share their perspectives on AI and how AI is reshaping the way they tell stories. So please join me in welcoming to the stage your moderator, award-winning author and journalist, Carolyn Giardina. Thank you.

Thank you, Maurice. I have actually known Maurice for many years, during which time we have had countless conversations about the impact of technology on storytelling and Autodesk's support of the creative process. Today, we're at a critical inflection point and at an important SIGGRAPH, and I am thrilled to be here to help examine the next stage in this journey. Our first speaker is Nikola Todorovic, Co-Founder of Wonder Dynamics, an Autodesk company since 2024. An entrepreneur, visual effect supervisor and award-winning filmmaker, Nikola spent most of his career working at the intersection of film and technology. This led him to dream up Wonder Dynamics and ultimately create what is now Autodesk Flow Studio with fellow Co-Founder, Tye Sheridan. And I'd like to introduce, in addition to his visionary work at Wonder Dynamics, Tye Sheridan, our next speaker, has enjoyed an impressive career as an actor and producer. He started in Steven Spielberg's Ready Player One and has worked with directors ranging from Paul Schrader to George Clooney. And I'd like to introduce our special guest, acclaimed writer, producer and director, David S. Goyer. His breakthrough came in 1998 with Dark City and Blade. He's also known for collaborating with Christopher Nolan, writing the screenplay for Man of Steel and co-writing on the Dark Knight Trilogy. In the realm of video games, he contributed the story for Call of Duty: Black Ops and various sequels. Presently, Goyer is serving as executive producer and showrunner -- he executive produced Murderbot for Apple TV and co-wrote The Pilot, and serves as executive producer for Netflix' Sandman. In film, your upcoming work with your production company, Phantom Four Films, is an announced reimagining of The Blob. And recently, Goyer created and serves as Executive Creative Director for the blockchain franchise, Emergence, built on the Incention and Story platforms.

So now that we've learned about the latest advances in Autodesk Flow Studio, let's delve into what AI tools such as these for storytelling and storytellers. With AI, we are at an inflection point in entertainment where creators are being asked to do more, faster and with fewer resources. But alongside these pressures, these new tools are emerging that can unlock entirely new forms of storytelling. So for the group, let's start. With this in mind, what excites you the most about this potential? And what are the areas that you feel needs attention?

I think for me, the most exciting thing is kind of the reason why Tye and I started this in the first place was I think the barrier of entry always in our industry has been quite great. I always call it kind of you need permission from someone to tell a story. And I think this AI boom and revolution is really opening kind of what I call permissionless storytelling. Whether you need to be in Los Angeles, a known producer, get some funding, it's quite hard to get your project green list. So to me, that's really exciting. And then also on top of that, I think what's really exciting, I think we're going to see more risk being taken in some of the stories. Unfortunately, if you have films, especially my favorite genre, which David knows really well, sci-fi, usually, those budgets are so high and unfortunately, forces studios to be a bit more conservative when they're making films because they need an 8-year-old and a 50-year-old to be able to enjoy it. So I'm excited about seeing this indie movement going in genres they haven't been before.

I think it also seems like the perfect time in the industry where the distribution landscape has changed so much over the past 10 years. And now what theatrical box office was 10 years ago is not really the same, right? So people are consuming content at home. I think the economics are a lot different in our industry. It forces studios to be, as Nikola said, maybe a bit more conservative. Also, they can't spend as much on certain projects. So I think where we can find places to alleviate pain points and help us produce the same films for faster or cheaper with the same artistic integrity behind it and with the same outcome, that's really what is needed in the industry. So for me, I feel like it serves that purpose. There's a huge opportunity there with these tools.

Yes. I'm thinking of the time, at least with respect to film, when we switched from film to digital. And the first time I sat in on like a digital intermediate or something like that and just jaw dropping at how we could change things so quickly, the look of things. And so I'm really just excited about, I think, the iterative potential for what this can mean and also the ability to eliminate or automate a lot of the drudge work so that, at least in the visual effects field, you can have artists being artists and not just doing rotoscoping or wire removal or something like that.

From film to TV to your games work, how has your approach to storytelling changed with the evolution of technology?

I mean, certainly, because I primarily work in science fiction, so the cost has come down significantly. I think almost anything is possible now. So that's exciting. We could not have made something like Foundation or Sandman, I think, even 20 years ago. It would have been just prohibitively expensive. And all of these tools, including what you guys have developed with Flow, they're going to bring those costs down significantly even more. So it's the kinds of storytelling, I think. And also just allowing, at least in terms of onset, putting more time -- I'm just thinking of how much time we've spent lighting green screens or things like that, how many manhours, and the fact that that's almost completely going away, that's exciting for me as a filmmaker.

And what about how it affects the fact that you're working in the different mediums, being able to tell stories across all of those mediums?

I think that the -- I mean I don't know that it affects working in -- I mean, I suppose it does. Certainly, when I was working in VR early on, that was so new that we were kind of making it up as we went along. And the same is the case with Emergence, which I built with Incention and Story. That's an entirely new kind of storytelling that is utilizing the blockchain and utilizing AI in certain components, which we can talk about later. I think what's exciting about that is we're experimenting with these forms of storytelling, and it's not entirely clear what the use cases are going to be from them. So we're sort of very early days in that right now.

Well, right now, we're going to jump into how AI is supporting stories. And Tye and Nikola, what was the inspiration for developing what is now Autodesk Flow Studio?

It didn't happen overnight. I think it was years of just us being really curious guys. We've got into the industry at a pretty early age. I did, for sure. And I think I was always amazed by the industry, amazed by what people could come together and create together. And so many people could pour blood, sweat and tears into a story and then that story would reach out and touch millions of people all over the earth. That, I became enamored with. And when Nikola and I met, I think we both had that love for storytelling and had always wanted to make our own films. And so when we met, that was really kind of how our relationship started. We started kind of coming up with ideas for films to make. And then we realized all these ideas we were coming up with were going to cost $150 million. And who's going to give us $150 million? I'm not James Cameron. So we thought, well, what technologies can we lean into to help us kind of relieve the cost. So really, that's how we got into technology. We started looking at technology as a means to an end to help us make those films that we had always wanted to make. And I think somewhere along the way, the research that we stumbled upon, we realized that it was much bigger than just us, that there were so many people out there, artists, filmmakers, idea people, who had all the ideas. They just either didn't have the connection to the industry, they didn't have the resources. And when we first started building what's now called Flow Studio, selfishly, we were building it for ourselves using our own films. And then we realized this is bigger than us. And we wanted to share it with artists and other people like us. And that was kind of our journey and the inspiration for it all.

It's clearly very personal to both of you.

Yes. You could say that, for sure.

And today, they just announced the new pricing. Tell us a little bit about that.

Yes, that was really important for us. We started our partnership with Autodesk about a year before acquisition. And when you're a startup, you kind of have a choice of are you going to raise a lot of money and kind of really race to become a big number on paper or run lean. We always believed in running lean. And one of the sacrifices there is we have about 25 models in some of the solutions we have, and it's not cheap to run it on the cloud. So Autodesk really, when we first met the team, they really believed in our vision. And one other thing was how do we make this accessible to everyone. So besides introducing the free tier, which I was really excited about, we also lowered the pricing of one of our lower tiers for self-serving customers. So I think it goes alongside of, as you mentioned, being personal, both Tye and I come from small towns. I come from a really small country, and it took me many years to end up in Los Angeles in the first place, but breaking the industry after that as well. So I think I was one of those kids that was 12 years old and watching Video Copilot tutorials to learn visual effects because I wanted to be a filmmaker one day. And so because of things like this, you kind of find your spot in this industry. So as I said, to me, AI really is fighting that kind of barrier to entry in our industry. So we want to stay true to that mission. So really what Autodesk is doing on that is something I'm really personally excited about.

Well, the new pricing, it's a bold choice and clearly important to both of you.

Yes, absolutely.

It was really important. I think you've got to stay true to the mission. We started this because -- as I mentioned, we started early, we started about 2016. 2018, we really committed to that pivot. And we said to each other, "What's the worst thing that could happen?" And we said, "The worst thing that can happen, we can fail completely, but at least we're going to learn what the filmmaking of the future is going to be before the rest of the people." So we really kind of saw it as an opportunity, "Hey, let's explore what this really is going to look and what filmmaking is going to look like." And then I do really believe in that democratization of storytelling being extremely, extremely important. It shouldn't be tied to a location or economical status. I think storytelling is -- it sounds naive, but I'm really big optimist long term of why technology exists in the first place. And traditionally, it's always helped us, right, tell stories easier, as you mentioned, from film to digital to streaming platforms. So we've seen the shift in distribution happen with the streaming platform, YouTubes over the world and social media. I think really the purpose of AI is to do that for production because production traditionally hasn't really went down that much. It's actually increased. So I don't think it's a sustainable business model. So this technology kind of comes -- AI in general, I think, comes really as a solution that our industry needs.

And Tye, as an actor and storyteller, for you, what excites you about AI and its potential for supporting performance and character development?

I mean when you save cost elsewhere, you can put more on the screen. Hence, so if that means more reversal time, longer production schedules -- there are things that I think people typically wouldn't think of that may be ancillary to the acting process, but it's really crucial on any film. So I remember asking a producer I was working with one time, I said, "What is the job of a producer?" It's like, "If you had to boil it down, like what does a producer do?"

Not in a mean away, right? You didn't ask him in a mean way, like, "What do you do?"

No, I just said -- I was like 18, 19 years old. I said, "You're a really amazing producer. Obviously, you have an incredible track record. Like, what does a producer do?" He said, "It's simple." He said, "A producer's job is to create the best environment to make films in." And I think we're always trying to figure that out. The world is constantly changing. The industry is constantly changing. Audiences are constantly changing. Taste for stories is constantly changing. And I think AI can really be a huge benefit in all that. Also, if it helps us actors to not have to wear these superhot, sweaty motion capture suits, I think that's a real advantage. But yes, I think it will have an impact in so many places that's not obvious, right, that will, yes, help us put more into the creative process. And that's my hope. I think what we built, obviously, we were really sensitive -- both of us being artists ourselves, we were really sensitive to the creative process and the integrity of filmmaking. So we built tools that would still maintain that, that those things were still core to what we built. And the better filmmaker you are, the better storyteller you are, the better actor you are, the better outcome and result that you're going to have running through Flow Studio. So I still think for us as a whole, in general, as an industry, it's really important that we educate ourselves on AI, everyone, from actors to filmmakers, to producers, writers, financiers, distributors, everyone should really come together and understand what it means for our industry, how it can be beneficial. And I think it's a completely new world. It is the Wild, Wild West right now, but keeping that creative process and the integrity intact is essential, it's crucial.

I'm just curious, you obviously started in 2018, a sci-fi film, Ready Player One. How do you think that experience would have been different had the tools that are available today been available at that time?

Yes, it's interesting. I mean the first things that come to mind are that film largely takes place in a completely animated world, in a virtual world called the oasis. And I think there were so many questions about what the oasis actually looks like. And it's very descriptive in the book that it's based on. But still, I remember in the early production process, sharing, there was constant concept dart going around. It's more like this, it's more like that. And the concept dart at the time, I'm assuming a lot of artists were making this concept dart, but imagine, if 90% of it was going in the wrong direction, it's a waste of time and it's a waste of effort whereas if you took the iterative process from, let's just say, generative tools to generate concepts, to understand what direction you are going in and then get artists to delve into that, I mean I think that's a way that it could be used. I mean also, not everyone can afford performance capture systems, the Vicon motion capture system. So using image-based solutions like Flow Studio or something like that for the animation, I could see being very viable. Again, you wouldn't have to wear the sweaty motion capture suits. But yes, I'm sure it could be used in a variety of different ways. Those are just a couple of ways -- a couple of things that come to mind.

David, let's go back to what you started to talk about earlier. You were involved in development of Incention and experimenting with these tools. What did you learn from this experience about AI and where it adds value and what guardrails might be needed?

Well, first of all, I mean, I agree with Tye that there's a lot of -- on one side, you've got all these people saying AI is going to save the world and revolutionize the industry. And then you've got the doomsayers. And it's going to be somewhere in the middle. It's going to end -- and I think that the -- what's not helpful, particularly in our industry are the people that are just sticking their heads in the sand. So I completely agree with the idea of people educating themselves and using these tools and experimenting with these tools. There's a lot of misinformation even amongst the studios in terms of the capabilities of these tools. I think with Incention and Emergence, what we were attempting to do was, if you look at how franchises are created, it's really from the top down, something you alluded to. And so a lot of these big franchises have either been created from or the IP was bought by these legacy companies that, in some cases, are 100 years old. And it's very hard to come up with -- to innovate and come up with new franchises sometimes from the top down. So we were looking at a way to also see if we could democratize the creation of these franchises. And so in our case, I wrote a bible, and then we trained that bible on a large language model. And we had a bunch of artists do concept art that I worked with. And now people that engage with Emergence can use the model as sort of a copilot to create new characters and new stories within that universe. And then the long-term plan is to allow those creations through an editorial board to flow upstream and become canon and then allow the creators to monetize the part of their creation as we move into animation or live action or what have you. So it was a way to sort of -- if you think about fan fiction or how much fan fiction there is with Marvel or Star Wars, there's no way for the people working in the fan fiction world, some of who are hugely talented, there's no way for any of those ideas to ever flow upstream and become part of these kind of bigger franchises. And so we were trying to crack that and using AI as, I guess, a copilot to help with that.

It sounds like this could be a whole new area.

I think it will be. I mean lots of different people are sort of experimenting with this.

Let's talk about the potential of new creative tools and efficiencies that we've talked about with AI in your various roles. What is the impact so far are you finding as a writer?

As a writer? As a joke, we tried to have ChatGPT write a foundation script, and that didn't necessarily work very well. But it's helpful in terms of -- I'll give you a classic example. I wrote like a 3-page monologue for a character. And it was cool, but I knew it needed to be shorter. So I could just take my 3-page monologue and put it into one of the models and say, okay, give me 5 versions of this, 2 pages long. And it could kick them out whatever in 30 seconds. And they weren't perfect. Again, it was my words edited down, but it gave me -- I might have spent an hour rewriting that monologue. And now by tweaking this or that, I was able to do it in 2 minutes, something like that. That's incredibly helpful. Or creating -- sometimes like on Foundation, we've got spaceships and this and that. And so can you suggest a pilot chatter. Just the stuff that you wouldn't put in the script, that you would put in a loop group or ADR, that's incredibly helpful. You need to edit it, you need to look it over, but that's incredibly helpful. I could see how it would be really helpful with localization as well. That's not specifically for writing, but all of these big streamers and studios need to localize their products, and I could see that being immensely helpful.

Well, my next question was about the potential as a producer, and it sounds like that's one of them.

I mean that's definitely one of them. But you were alluding to -- I mean, again, on Foundation, we've got a lot of situations where -- I mean, we have some actual sets. We've got sets that are 50% virtual, 50% real, and we've got sets that are 100% virtual. And it's hard for the actors to visualize, if they're in a completely virtual set, what it's going to look like or what the conditions will be like. And I think if they're just guessing or if the creators don't know what it's going to look like, then those performances aren't necessarily going to track. There's going to be something kind of unreal about it because the actors don't know what they're acting against. And so I think real-time visualization, for sure -- I was on a project where we spent over $2 million just rigging green screens. And the idea that that's something that we could eliminate and potentially have more shooting days for what have you, that's absolutely massive.

And then it sounds like that goes into also the impact on the director.

For sure. I'm really curious also -- I could see ways that it could go wrong, but I'm curious about the advent of AI and cameras, following, focusing and adjusting exposure. I could see how that could get very generic and create a lot of slop, but I could also see how that would be a really interesting tool as well. So that's something I'm interested in experimenting with.

What are the areas that you think need to be protected as we move into AI?

Well, obviously, one of them is copyright. I think with all of this technology hand in hand, there needs to be -- this is going to be -- not just for film, but I think the AI revolution is going to be like the industrial revolution times 10. I mean it's going to change everything in our world, and it's going to change everything in our world very, very quickly. And there's going to be amazing ways that's going to be beneficial, and there's going to be a lot of unintended consequences as well. So I think the idea of education, of training, of allowing people who've spent 20 years doing a craft in one way and then training them to -- I look at the way, frankly, the camera crews have adapted to digital. I mean that's something that you're going to, I think, start to see in all the division of labor within film.

Something we were talking about upstairs before was the impact on characters and development of characters. What are your thoughts in this area?

On AI creating characters?

For assisting you in creating character.

Assisting in creating characters. Look, right now, I will say this, we started using -- I think my supervisor for Foundation is somewhere in the audience, Chris. But we started using, for instance, mid-journey, I think while we were shooting Season 2 of Foundation. And we would employ a ton of concept artists, whom I love. But we started also very early on putting prompts in the mid-journey. And so two things would happen. One is Chris and I were able to sort of iterate very, very, very quickly and then give some of those images to the concept artists. And so at least they weren't -- we were kind of slightly narrowing down the goal. But the other thing that would happen that I thought was really interesting is the mistakes. So as we're prompting it, there's all these unintended mistakes that would come out of these. But some of them were amazing mistakes, and there are things that in a million years, we never would have thought of. And that sparked a lot of creativity as well. So there's a ton of ways that it can be immensely helpful. I mean everyone is using it right now, creating look-books and creating reels and cloning voices.

It's interesting you say that because in the early days, we were experimenting with physically plausible engines and, for instance, giving a character like a slightly shorter leg so that they have a limp just naturally, to see like what you could generate -- what kind of character either flaws or just traits you could generate in that regard. It was a pretty interesting.

Yes, my kids do that on Roblox. It's like how can we mess with the physics, just tweak the physics 5%. I'm really curious, one of the things in science fiction is it's really, really expensive and really hard if you are filming a scene like on a planet that has not zero gravity, but like 1/6 the gravity like the moon or half the gravity. And I'm really interested in seeing what AI can do in terms of that because that's really going to change the game as well. And I suspect we'll see a lot of that with AI and video games. So like imagine you're playing Helldivers and you go to a new planet that has half the gravity of Earth and now the physics of everything have completely changed and the way that your weapons works has completely changed. And that's something that I think we're really on the cusp of right now, with AI sort of generating these worlds and these maps.

To pick a past project here, let's use Blade as an example. Is there something that you would like to have been able to do at the time, but maybe you were limited by time or by budget?

Yes. Well, famously, there was an entirely different ending that we were going to do with the character Deacon Frost turning into a blood god. And we spent a couple of million dollars with a VFX company that has since gone out of business, but they could not crack it because they couldn't crack the physics of it. And he was sort of turning into this sort of living whirlwind of blood that could also animate. And we had to completely abandon it and just go for something -- the technology just didn't exist to do that back then. We were foolhardy in sort of attempting it. So I think if we had the tools that we had now, we probably would have succeeded in pulling off that ending. Now the question is whether or not it would have been as good, I don't know, because the ending that we did have for the first Blade film was just something we pulled out of our ass at the last minute.

And it worked. Could you tell us anything that you're currently working on and if you're incorporating these tools at this time?

It hasn't been announced yet, but I'm working on a sort of horror/science fiction project in the kind of what they call an analog horror vein. And we're talking about potentially shooting it with iPhones or Prosumer cameras. And I'm curious to see if we -- because there are some nonhuman characters in it as well, but if we can be using motion capture just using AI to do it, I'm really curious to see how we can -- because that's a very low-fi looking project, deliberately so. But I'm curious to see to the extent that we can employ the latest technology to make it look low-fi as opposed to looking hi-fi. But I can't -- it hasn't been announced yet.

Maybe at SIGGRAPH 2026. And then what about the team? Do you expect to keep your current workflows? Or do you anticipate roles changing?

I mean, look, again, I go back to when we transferred from film to digital, the idea of a digital intermediate or a DI technician was not something that existed. And now that person is like a critical member of the team. I think if you're dealing with virtual environments and you're previewing in real time what those virtual environments are for the cast and the crew, I mean that's something that I think is almost increasingly -- usually, that's like a specialist that just comes in for different -- if you're shooting something on the equivalent of the volume or something like that, somebody comes in just for that sequence, but I could imagine that being a dedicated sort of component of the crew moving forward. I'm also curious to see to the extent that will we need as many lighting technicians. I think you'll see a lot more work being done in prep than we are right now. I think a lot of that a lot of that will be shifting over. But I don't know. I mean, what do you guys think?

Well, for all of you, I mean, obviously, there's a lot of fear out there right now. What advice would you give to someone working in the business right now who is worried about being replaced?

Yes. I think there's obviously a lot of fear on it, but visual effects industry has always kind of been the first forefront of kind of change in technology, right? As I said, we spoke a lot to people when we started developing this early days, and it was really hard to explain to people what is going on. And I think now everybody is more aware. My advice always would be don't kind of try to put your head in the sand and hope this thing is going to go away and the industry will remain as it is. I think it's a wrong approach. But I always advise people it shifting doesn't mean that your titles will shift. There are certain roles. Some of the roles in VFX especially have been sort of shaped because of the tools you used, right? And so that will change quite drastically. But I think more importantly, my advice always would be, don't be scared to go in a field that you don't understand in a sense like always tell artists, and we had this sort of our artists in our companies, is read a research paper. Even if you don't understand the math and physics behind it, it will give you some ideas of how you can combine a couple of maybe open source solutions with some commercial tools and then build on top of that, right? So I think because the industry is shaping -- is changing, I think it's very important that we also build it internally inside of our film industry because certain use cases are very different on the tools that we're seeing, right? There's a really big difference if you're building generative models with a kind of goal of making tools for social media versus with the goal if you're making tools for our industry. Our industry is very iterative in nature. It's all about the details and fine-tuning and where you want your camera to end by millimeter. So much, much different than, for instance, turnaround in social media where you're not going to nitpick the pixels as much, right, because your turnaround time is 1 day versus in film where you're going to work on it and you kind of have to know when to stop when something looks good. So I would say curiosity is really important, but I think it's really important we build it internally, inside of our industry, so then we can shape it how we really want to see it long term. So I guess, fight the fear because we've seen the change in the industry many times. And I think sometimes the hype can be overwhelming and the clickbaits will get you and you're going to think it's the end of the world. But honestly, to me, really it's a means to an end, as Tye said. We're all trying to tell better stories. How we got there will change with time. And one thing I always say is we really don't know what that new -- the baseline has changed, right? The baseline has really changed. But to me, what's really exciting is the ceiling is also going to change. We just don't know what that is. But I'm a firm believer in that we recognize laziness. Let's say the prompts get really well, the text in video gets perfect, and we can make a film in 5 minutes. If we know it took 5 minutes, we're not going to pay for it or like it, right? We're going to want something that's much harder. In my mind, that's why we like art. We like art because it's quite difficult and you're impressed by how do they do that. In Foundation, how did David pull that episode, right? Because it was really difficult, and I wouldn't know how to do that, right? So you respect that, and that's what really attracts us to watch films. So we just don't know what that new ceiling is. We know the baseline is changing because of these tools. But I think if you look at it pessimistically, you just think, okay, my industry is going to become easy and maybe that's the end of it. No, I think we're just going to push it forward, just like you can't explain to someone 50 years ago how Avatar was made. They're not going to understand, but they're watching completely, right? So that's what I'm excited about. We'll see what that new thing is.

I don't think it's that the fear is completely unwarranted. I think that it's warranted in a lot of cases, like as you say, fear of someone who doesn't know your industry, coming in and building tools for your industry, but they don't know your process.

I think that's key, that this change needs to be driven by the film industry, the creative industry, and not necessarily by the tech industry because in a lot of -- we were talking about this in the green room. A lot of these products that are being built are amazing products, but people that are building them don't really know anything about film or storytelling necessarily.

I think they go hand in hand, right? Like, the artists need the technologists and the technologists need the artists. And it's educational. It's a two-way street. We have to be communicating with the other side. I think that was probably the most interesting aspect to the whole experience of building Wonder Dynamics is really combining multiple worlds and multiple industries inside one house, inside of one product. And we hired a lot of junior machine learning engineers who didn't know any film terminology. And as I was telling you guys, we were first looking at data sets of motion, but all these data sets, they had camera angles from the top corners of each room. And we knew that wasn't going to work for film because when's the last time you saw that shot in a movie, never, unless it's like an insert of a security camera footage or something like that. So we encourage them to watch films to understand a close-up, a medium shot, a wide shot, a cowboy, a two-shot to understand what films are made of. And so we had artists and we had machine learning engineers, and we got to see that communication within our own company, and we really encouraged it for one side to learn about the other, to be communicating with the other side. And I think what we saw in the film industry during strikes with SAG and with WGA, I think people were making conclusions without having all the necessary information to make the basis for that conclusion. And I think that's the worst thing we can do, as you say, sticking our head in the sand and just ringing the alarm bells is not going to do any good. I think we have to stop and, one, look at the state of our industry; but two, look at the state of these tools, look at the research, what's coming. If you don't understand it, ask someone who knows, they'll be happy to learn about your industry. You could see something about the film industry and filmmaking. And like I said, it's a two-way street. And I think that in terms of the culture, in terms of what we build, the tools that we build, it has to be means, it has to be the foundation, the track that we lay. So I think we've always felt that and want to encourage people in the industry, I guess, both in technology and the film industry, to think that way. It's essential.

What sort of feedback have you been getting so far? And how is it driving your thoughts about where your tools fit in either to new workflows or traditional workflows?

Yes. I think for us, since we started early, we knew there's going to be this hype in AI space. So we kind of tried to be really honest with ourselves and work with artists and realized, "Okay, if AI gets me about 60%, 70% there, let's say, automated, how do I really not turn into a black box," because that's my biggest frustration right now with a lot of tools. There's not a lot of editability. But it makes sense. The research went generative and now we're trying to figure out how to edit it. So for us, it was always, "Let's not try to disrupt the full pipeline. Let's build something that adds in the pipeline." So for us, when we first started, we did this big project with the Russo Brothers, and we've hired a lot of artists and we really asked them, "Okay, if you get 60% there with AI, what data do you need to really push the shot? Just tell us what passes you need. You need your mocap, you need your camera track, you need your point cloud, alpha mask, et cetera. So we really approach it from a standpoint of, "Okay, I have AI, but then I'm going to go in my DCC of choice." That's whether Maya, Blender or other tools where I can really edit and move the camera or animation, correct it and clean it up in a space. I'm a big believer in that combination in 3D space and 2D space. It's really important because you can't prompt everything. You can't prompt performance. You can't really completely prompt the camera move, not to a desired shot that you want. You can prompt a frame or a little bit of a movement, but usually, directors will sometimes change their mind on past 20 and want to change where the character is looking or maybe move a leg a little bit this way, maybe move a camera this way, so that particular choices and finding editability is really important for us. So we approached it with a sense of, "Okay, AI is not completely there yet. So let it work in that pipeline. These problems will get solved and research is getting better." But I'm a big believer in multimodality. You can prompt something, you can have video sources for something. And then you stick to kind of traditional 3D or reinvent how you're moving things in 3D space. So that's sort of how we approach the problem.

Last year, you made a very big decision with the company. What made you feel that Autodesk was the right partner?

A couple of things. I think the vision is really aligned. Diana, who runs media and entertainment inside of Autodesk, we spent a year working with them. So first of all, timing is really important, working with good people. This is our baby. We're very protective of it. But also, I think the vision really aligned and they really gave us full trust and control of the road map, which is rare. For a start-up, you always have this fear where you go in a global company that your vision is going to get killed, and they really kind of gave us so much trust and believed in our vision and we pushed it. And the opposite of that, obviously, with this announcement of free, that was important for us to go in that direction. And then also the knowledge. Obviously, Autodesk with Maya and other tools have been leading the industry for decades. So for us, we knew we were lacking a little bit of knowledge, especially in that 3D space, that we wanted to go forward.

I think we also hadn't had, how many years, 20, 30 years of experience of delivering software to the industry that we were hoping to have an impact on. And I think that was obviously their legacy roots within the industry. It was something that was really beneficial for getting a tool like this out there in the hands of artists and having big companies trust the tools that we built.

Also, they promised us we're going to meet David, and they said, I'll get to direct one episode of Foundation.

That's right. I remember that.

That was in the contract.

You're a liar, right?

Yes. But it's also the ethos. Everything that we believe in it, kind of felt like one family. And the ethos of what we wanted to build and the spirit and the essence of why we built it, it felt very aligned in that way. And yes, I think the coolest thing ever was when we first started talking about a partnership, when we pitched them our 5-year road map, to hear like, "That's great. We want to support that," and that was all that we could ask for. So yes, it's been a great partnership and all the stars have aligned.

So outside of Hollywood, the creator economy is obviously global. David, what do you see as the -- are there unique opportunities or unique risks that artists around the world face?

Yes. I mean look, the opportunities are almost the same as the risk. And what I mean by that is there is a tendency sometimes with these new technologies to get lazy and just to say, "Oh, we can automate this now. We can have AI generate characters from old cloth or stories from old cloth or episodes from old cloth," and I think what you're going to get there, you're already seeing it, is a lot of slop. And I think we'll reach a period where there's a lot of slop being created and then the pendulum is going to swing the other way because my kids, I've got 3 boys, they can already tell whether something is AI generated, even video, the video of the rabbits on the trampoline. My 11-year-old was like, "Oh, yes, that's AI." I mean he could just tell immediately. So people that are digital natives, they're not necessarily going to be fooled by that. So that's one of the perils. And then the opportunity is, okay, I'm interested in people taking tools like yours and then breaking them and using them for something that you hadn't necessarily thought of because that's where the innovation is going to come from. And that's where I think we've seen a lot of innovation even in going from film to digital just throughout the decades.

That was essentially our story because we when we first got into AI, we were looking at research papers from the field of robotics and autonomy for self-driving cars and all of these image perception models is research that's just been done for 20, 30 years. And it wasn't intended to be applied to the film industry, but it could be because what is film? 24 images a second. And people always ask us, "Why did you guys start the company? What was the lightbulb moment?" I don't really say there was a lightbulb moment. But if there was, that was kind of it, looking at that research and realizing that it was going to become the future of our industry. And I think that's really what kind of gave us the courage to take that leap into this crazy new world.

And then as we look ahead, David, why don't we start with you, what advice would you give seasoned professionals and emerging creators about embracing AI, not just as a tool but as a collaborator in the process?

You said it earlier, it's not going away. It's here. Immerse yourself in it, learn about it. The more you learn, the more you can manipulate it, use it as another tool, the more you can drive the movements. So I would just say, like my kids now, it's just like any AI embedded program, just use it, experiment with it, create a facility for it, have become second nature because it's absolutely -- but that is the thing that's exciting. When I was coming up, even making a tiny movie, right, indie movie, would cost $0.5 million or $1 million. But with the tools that we have now, you could absolutely do a feature for $60,000, $20,000. That just didn't exist when we were coming up. Experiment, experiment and education.

I think if you look at the human's relationship with any tool, whether that be a vehicle or a computer or a guitar, when they know that thing super well inside and out, they're able to make something truly unique. And I think that these AI tools, they're really no different than a camera or a sound equipment or an animation engine. I think the better you know those tools, the more you can manipulate them, the more you can stand out, right? And I think like why we love certain filmmakers or why we love certain artists is because they have a unique voice.

Point of view.

And a point of view. And I think the closer you are to those tools, the better you understand them, the more you can manipulate and control your own point of view. And I think that's one and the same. You can look back through the history of time to infinity, and that's true. And I think it's just as true today with these tools.

If you look at the adoption of drone use in filmmaking, I mean, I'm old. When I started, if you wanted to do aerial shots, you had to use helicopters. And they were cumbersome and they were expensive, they're dangerous and there are only so many good helicopter operators. And we use drones all the time now. And we were shooting last year with drones in a sewer tunnel chasing the characters down the sewer tunnel with these tiny, tiny drones. And that's just stuff that wouldn't even have been possible 3 years ago. And it is another tool that allows you to tell a different kind of story.

And to that end, if you were starting today, what's the first type of story you'd want to tell?

Wow, the first type of story. I don't know. Again, I'm really interested in telling a story in an environment with a different kind of gravity. I mean that would be wild, doing the whole show set on the moon and not just having the actors walk really slowly or filming 90 frames per second or something like that. I think that would be cool.

It really comes down to the limitations, right? There are so many times you put something on a paper and then you know you cannot do it. You can't physically do it. You cannot do it maybe because of the budget. I come from the indie space and it's always so frustrated like you have to stay away from CGI, you have to stay away from visual effects because you can't afford it. And then you have to always box yourself in based on locations you can afford, based on what you can shoot, how many shooting days, et cetera. So to me, that's the whole point of AI is like how do we get from imagination to screen without sacrificing a lot of things or changing a lot of things.

There's a location, which I've done before, that you can't fit 200 people in. You can only -- I shot in a location once on a big project, the crew, we skimmed it down to 12 people. That was the only way we could do it, which was fascinating. But I like the technological solve for things like that.

Absolutely. And as you mentioned earlier, to me, what's exciting is film is very specific because you don't get much time to practice, right? How many films you can make in your lifetime, 10, 15, if you're lucky, right? I mean I know some of the greatest filmmakers that still get so nervous before shooting a film just because it's still rare. It's still such an event for you to do it. And just like everything, I think with these kind of tools and generative tools, you get to practice a lot. And then you get to discover how do I make an arc in this film? How do I frame this? How do I tell this story? How do I build the characters? My biggest concern with generative tools is I really do hope we add to the process versus that we're switching into synthetic actors, for instance. I hope we are not going in a future where we're just having actors sitting at home and licensing their likeness and then we're just watching films. I always ask people, "Do you think in 5 years, top 3 box office films will be all synthetic actors?" And it's a hard question to ask.

I think they're going to try it and you're going to get an amount of slop.

Yes. You're going to get a lot of things. But to me, I hope we are not.

I think that would be a good thing. Let's get the slop out of the way, make people hate synthetic actors, the real stuff.

Earlier in this session, we were reminded of Leonardo da Vinci. Do you think these tools will help to identify new voices?

Absolutely. I'm sure they will.

Absolutely. I think the whole point is to turn it into a global, to discover new talent, to me, at least. I mean that's something for me that's -- we're still very much tied to a certain location in the world where you have to be to have these opportunities. Obviously, coming from -- I was born in Bosnia, lived in Croatia and Serbia until I was 20, and then I moved to U.S. to pursue a film career. But before that, it was just a pipe dream. Explaining someone that you -- unfortunately, being kind of in a time and area in the world where there was quite a bit of conflict, trying to tell someone you have a dream of making film just seemed not something realistic at all, right? And so I'm a big believer that the whole -- for me, the whole point of this revolution is that. It's really how do we discover these voices. And we don't know where next Cameron or Steven Spielberg are coming from. But if you really set that baseline to be -- and that's the beauty of it. It's so new. Everybody is starting kind of from the new starting point. So then you'll get to -- we don't know if these kids that are 5, 6 years old now today are playing with it, they're able to produce things. What will they be able to do in their 2025? We just can't imagine yet. And that to me is exciting.

I think also Hollywood prides itself on telling global stories. But the question is, how authentic are they? I mean, are they really -- to someone living on the other side of the globe, this story is about their culture. They feel it's honest. For me, I think the best art is when it's honest, when it's pure, when it's authentic. And I think to have kids born in any crevice of the world, to have the ability to go out and make something and be seen, is such an exciting dream to envision. I mean I think that's also why -- as Nikola said, we both come from really small towns. When I was a kid, I didn't even know that you could make films for a living. I didn't even know that was like a possibility. I knew people made films, but it never clicked until I just happened to be on a film set and see it. And I was enamored by it, and I've never stopped since. I've been chasing that miracle ever since. But yes, the idea that we can educate audiences around the world through more authentic stories around the world that are pure from those artist voices directly, I think, is really, really exciting. And I want to see that vision in my lifetime.

Well, and this is the last comment and question before we wrap this up, as we look ahead, I think one thing we all agree that's so important right now is education, experiment, try these tools, familiarize yourself with these tools. Are there other comments you'd want to make?

I think that, on one hand, you've got the dream that these technologies are going to lower production costs. And that's definitely something that I think needs to happen and probably will happen. But if you look at the movies, particularly even this year, the theatrical movies that are breaking through, they tend to be the ones that are unique or quirky, whether it be Sinners or what Danny Boyle did with 28 Years Later or even Gunn's Superman, which was unashamedly nerdy and odd and kind of comic book-y. And I think that people are still going to create, crave stories that feel unique and quirky and different and authentic. And that's something that technology can't produce. There always has to be the human interface in there, the author, guiding the machine. And so that's why I'm also optimistic about it. And I think that it is going to be the Wild West, and it's going to take a while for things to shake out. But I would just encourage people, every new product that comes out, just experiment with it and try to break it.

Create happy accidents.

Yes.

Awesome. Okay. Before we wrap, just a reminder that Autodesk has a stand in the exhibition hall, and you can check out Autodesk Flow Studio and learn more about the new pricing. And there's also a QR code. Please join me in thanking David, Nikola and Tye.

Thank you.