Master Your Sales Pipeline with Matt Heinz, Founder & President of Heinz Marketing

(Intro Narrator at 00:00:00) Today, we're talking to Paul Salvini, Global CTO at DNEG, about the technology of the VFX industry. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:16) Paul?

(Paul Salvini at 00:00:17) How are you, Joel?

(Joel Beasley at 00:00:18) Dude, this is good. This is exciting. You're in Canada today?

(Paul Salvini at 00:00:22) That's right, just outside of Toronto in Kitchener-Waterloo.

(Joel Beasley at 00:00:25) Is that where the headquarters is?

(Paul Salvini at 00:00:27) No, just where home is for me. And so work from home most of the time and travel around the world where we need to be.

(Joel Beasley at 00:00:35) How do you say the name of the company?

(Paul Salvini at 00:00:37) DNEG.

(Joel Beasley at 00:00:38) DNEG. Is it an acronym?

(Paul Salvini at 00:00:40) It used to be Double Negative. That's where it originated, but everybody in the industry called it DNEG. And so that's why DNEG ended up being the brand.

(Joel Beasley at 00:00:51) Well, I'm a big movie fan, and I just found out—I looked all through your roster—you guys have done the VFX for Dune Two, Oppenheimer, and basically every major movie that's been out in the past couple years. I just stopped reading when I got to a certain point. I was like, they work on every single movie. It's crazy. Can you tell me about how you describe the company?

(Paul Salvini at 00:01:15) Yeah, absolutely. So it's such an amazing company. It's very much a creative partnership company. We work with some of the leaders in the industry in terms of directors and filmmakers to really bring great stories to life. And so our role is to bring together technology and artists to create stories and to be able to tell stories that probably couldn't be filmed and couldn't be told any other way. So that's our role.

(Joel Beasley at 00:01:43) So you heard the episode I did with Nathan on movie theaters, and then we decided, reached out, and we're like, all right, let's set up a conversation with these VFX people. What do you think is going to happen to cinemas long-term?

(Paul Salvini at 00:01:58) Well, I sure hope they have a healthy and successful long-term future. I think there's something very magical and special about the cinema experience or any sort of shared experience, shared visual experience. Before I went to DNEG, I was actually the CTO at Christie, which was a manufacturer of projection and other visual systems, especially in the cinema world. And so for me, really thinking about the future of cinema and what that experience can be and really how we consume and enjoy content—I was saying to Josh earlier that I still go to the cinema as much as I used to. It's just I also consume media in every other imaginable form as well. And so I think that's pretty exciting. But for me, nothing beats seeing something on the big screen the way it was sort of meant to be seen.

(Joel Beasley at 00:02:50) Have you built a cinema in your house essentially?

(Paul Salvini at 00:02:53) I do have one, yeah, as a matter of fact. But I still love to go out, and it's nice to see something with other people as well and hear their reactions and to be in that environment. And obviously, as your guest said, the actual cinemas are becoming much more like your living room or your basement, where the seating is very comfortable. The food options have grown and things like that. So it's quite an experience now out on the town.

(Joel Beasley at 00:03:22) Since I did that interview with Nathan, I listened to an interview with Elon Musk, and he was talking about his kids. He has one kid that I believe he says is autistic to some degree. But one of his kids said, "Hey, Dad, I noticed that the reason why we go out to eat is not because of the food. It's because we like to be around strangers in the restaurant, because you can make the best food at your house if you're right." But you still choose to go out and have the restaurant experience because it's nice to be around strangers. And I think the same is true for the movies.

(Paul Salvini at 00:03:56) It is, absolutely. And I think that's exactly the feeling you get—that it's a community thing. People are going to experience and enjoy something together. So it's nice to do that. But obviously I'm grateful for both. I mean, there are times where you need to keep the kids occupied and nothing better than letting them watch some content on an iPad or something like that, or when you're on an airplane on a long flight to be able to watch something. So everything has its time and its place. But yeah, I sure hope the theater experience continues for many years in the future.

(Joel Beasley at 00:04:27) There's a lot of fear there in the industry, and the fear tends to be that if we stream to home, people aren't going to come out. I think there's truth to that. I'm sure they can see it in the numbers. But I think there's a certain amount of confidence that if you say, "Hey, we're going to create such a great experience that people are going to desire to just engage with this experience," and then adjust your business models accordingly, that's going to work long-term—just like my kids and their horse riding lessons every week.

(Paul Salvini at 00:04:54) Exactly. Yeah, no point, right? So I think it'll become more of a special experience. You'll do it on films that you know truly matter and have been built for that kind of experience and warrant that kind of experience. And then there'll be other things that, you know, better make sense to watch at home or watch in a different way, whether it's sort of an immersive experience, a VR experience, those sorts of things. So I think the future of content is a super interesting one in terms of how people will enjoy the stories that get created.

(Joel Beasley at 00:05:29) That was an unusual episode I did. It got a lot more feedback than other episodes that we've done. And so I thought maybe there's just a lot of movie fans out there. And one of the things that people really enjoyed was how open and honest and willing Nathan was to engage in the conversation about his business going out of business—in the sense of just an industry consolidating and behavior patterns changing. To that point, I figured, why not ask the tough questions to the smart people? What are you guys going to do with Sora? Like, how's that going to impact your business?

(Paul Salvini at 00:06:07) Yeah. So, well, the whole issue of AI is a super interesting one. I think it'll be absolutely transformative for content creation in general. I think probably more so in the consumer space and the prosumer space. I mean, AI is—you never want to say never about things replacing other things when it comes to technology, but it often, you know, it really takes time. What we do is not just create images in a sort of random way and say, "Yeah, good enough." We really need to bring stories to life in a very exacting way. And if you see the time and effort that goes into every detail in a high-end film, whether it's a Dune or an Oppenheimer—I mean, these are very serious endeavors in terms of bringing a story to life for people to enjoy. And you know, there's a place in the world for "good enough," and there's a place in the world where you aim for perfection. And I think it's a reasonable question to ask, but I guess it reminds me a lot of my own journey in photography, right? Because in the early days, I started out as a photographer. That was what I was going to do for a living. And you know, I had my film camera, of course, at the time, and my big clunky lenses. And over time, you started to see more and more smaller format cameras emerging until eventually everybody was carrying on their smartphones, you know, a camera with incredible capabilities. But I don't think it ever replaced the idea of professional photography and what you're able to do when you think about how do you capture an experience or capture an event and deliver it to people. So if you look at, whether it's a live sporting event or whatever, you know, there's not a bunch of people on the sideline with smartphones. It's serious people with serious leading-edge technology, finding the best way to capture that experience and deliver it to their audiences. And so I think the same thing holds true with AI—incredibly powerful set of tools. We have a fairly large AI research and development team at DNEG, but we're not looking at how to replace people or things like that. We're looking for how can we use AI to improve the process, either to reduce some of the mundane things that we have to do, increase our productivity, increase the quality of what gets produced. And that's where AI really, really shines. It's allowing us to focus on the things that truly are the art form, that truly are the craft of creating a product of the quality that you see in something like Dune or Dune Two and Oppenheimer. So that's an important thing, very critical to our industry. And I think it'll change a lot of things that we do. But for sure, it's the kind of technology that will have a huge impact in the consumer and prosumer world and moderate eventual impact in the high-end world.

(Joel Beasley at 00:09:17) One of the things I've noticed doing these interviews is that people really don't want to say that they replace people. Like, everybody doesn't want to say that. And I think it's fascinating. And so I found that, you know, it's not a popular thing to say. No one wants that sound bite out there. And that's also not exactly what we're doing. What we're doing is people are a collection of repeatable processes that they execute on, and we're trying to figure out which repeatable processes—because you don't want to try to replace the human because that's not a solid endeavor, because the human does many different things, and they will all be on this spectrum and gradient of when they can be automated by what technology. So you want to look at, you know, if I've got a group of people—let's just call them editors because I don't know the slang. Let's just say editors. And I do a poll across my company and I say, "Where do you spend your most time?" as one set of questions. And then, "What are the things you least enjoy doing?" And overlay those for opportunities for potential—like, you know, what are the most tedious things that you don't enjoy doing and where do you spend the most time? And then that's how we have—at our company, and we're small, we're less than 20 people, you guys are like thousands of people—but that's how we have figured out how to leverage and which opportunities to pursue. Are you guys doing anything like that?

(Paul Salvini at 00:10:32) Yeah, very similar. We look at where we spend our time. We look at the tasks that really don't add value to the—if you know what I mean—they have to be done, but they don't add artistic value. So, you know, remove this wire, remove this object. Hey, there was an air conditioner on the side of that building that probably wasn't there in 1950—let's remove it. And so tasks like that are going to be fantastic. And I think you're exactly right. It's not about replacing the person, it's about replacing certain tasks or significantly improving the ability to very quickly do certain tasks. The entire industry of computer graphics and digital filmmaking, if you will, is replacing the traditional way of making films, right? So it's hard to say as a technologist that, "Ah, yeah, no jobs are going to be affected." That's certainly not the case. But it's not that—the nature of the jobs are just going to change over time. And you know, precisely what each individual is going to do is going to change. But I also think the thing that's important to note is the volume of work that we're going to be asked to do as an industry is going to increase over time. And that's been consistent since I started. I mean, I started in the mid-'90s where a big visual effects show might have been 50 shots, something like that. It's like, "Hey, can you extend the set? Can you remove these elements that don't need to be there, like that wire and these people in the background of that car? And then, oh yeah, can you add this monster in this particular scene? Because there's no other way we can do that. We tried to film it with a person in a suit and it didn't look good. So, you know, we'll need to add that digitally." That was visual effects largely in the mid-'90s. And if you look at the types of things we do today, you can watch a film, even a rom-com, and think, "No visual effects in that." Then you sit and you wait for the credits, you see the list of hundreds, if not a thousand visual effects artists, you think, "Oh, maybe there were some visual effects in that." And the reality is the scope of what we do has changed over time. And we're doing so much more today digitally versus what would have been done traditionally, just because it gives incredible flexibility to filmmakers and storytellers to do things in the digital realm.

(Joel Beasley at 00:12:56) I'm trying to piece together this industry. I don't spend a lot of time here, but I'm a big fan of it in my personal life. Like, I typically don't watch a lot of TV shows, but I like movies. I tend to find that captures my attention—that format of story and the completion of it and the time. I like it a lot. But what I'm curious about is—I'm going to share with you what I see in my mind, and you tell me how far off I am in reality. There's industry, which is like the big guys doing it well, right? Like DNEG and Animal Logic. I don't know if they're a competitor, but I got to know a couple people over there. But you've got the big guys that are doing stuff really well. And then you've got, you know, that YouTuber that's piecing together trailers with Sora prompts that look like commercial grade. And so what I would imagine is happening is that these are both progressing together, right? And they're both progressing. And industry is looking at the technologies and figuring out how to improve inefficiencies, while the cottage industry, I guess you could say, is just figuring out how to do something cool. And at some point, there has to be some system to monitor what's happening to know when there's a prime for an acquisition or when there's a prime for a serious process change. And I was hoping you could tell me—does that exist, and what does it actually look like? How do you watch cottage industry and monitor their advancements?

(Paul Salvini at 00:14:22) Yeah, I think we always have an eye on what the consumer world and the prosumer world is doing, because I think in many ways that it informs what our world could be like if those tools progress to a certain level of sophistication. But you know, the things like the demos you see coming out of the AI tools today, even the video tools—they are incredible. There's no question about it. You know, they have an amazing look, and yeah, with some very simple prompts, you can produce something that is very sophisticated, that would take an incredible amount of time and effort to replicate. But the challenge is that often doesn't give you exactly what you want. And in any sort of either full feature film or TV series—you look at a Last of Us or the Masters of the Air that we're doing, or a Lord of the Rings series—you know, it's not about just a single scene with a certain prompt. You've got characters that have to be developed. You have locations that have to be modified over time in a consistent way, right? You have all of the elements that have to come together in terms of telling a great story. And you need that consistency from scene to scene, shot to shot, sequence to sequence, over time. So there's a pretty huge gap between creating, you know, a short piece that looks amazing and creating something that very specifically fills a purpose. And, and hey, I'm guilty of the same. Every year I do Christmas cards for people. And this year I decided to do all of my friends' dogs in a Christmas theme. And of course I used AI and fired them out as quickly as could be done. And it was a lot of fun for me. It was super creative.

(Paul Salvini at 00:16:17) It wasn't a huge workload, but I didn't use the tools of our industry to do that because in that case, the goal was to create something simple, fun. In fact, if it looked terrible in some ways, if it was kind of campy, that was actually added to the joy of it. It's like, wow, that hat doesn't fit at all. And so there's an example where the tools are amazing, but we certainly wouldn't use them in that state in the world we are. So we do have to watch because there are people creating elements that are super exciting.

(Paul Salvini at 00:16:51) So if we can start to use AI to learn, for example, how certain creatures move and walk, that would be very useful. I think if you want to look at the way that the film world is going to evolve, it's about, you know, the first step is a bit of a separation between the performance that we capture and the look of what people are going to see when they watch the end product. And that already happens. You'll have an actor make a performance and bring some other character to life that doesn't look like that. Now that character could be very different, like a monster, or the character could be a younger or an older version of that actor, or it could be the exact look of the actor, but had to be inserted digitally because the lighting was wrong.

(Paul Salvini at 00:17:42) The way it was filmed was wrong, or it had to be put in a scene that was too complicated to film practically. And so there's a bit of a separation already emerging between the creation of a performance and what people actually see. And of course, that's very different from the old days where everything was through the lens of the camera and the performance was what you saw, was a direct capture of the performance that was given. And now we have the ability to combine, blend, modify, change, but it has to be done obviously in a responsible way, and it has to be done in support of telling a great story.

(Joel Beasley at 00:18:19) Let's deepfake the whole thing. How much time have you spent personally with your hands on the keyboard, writing prompts and generating short form videos through tools like Sora or any of those? There's a couple other ones as well.

(Paul Salvini at 00:18:35) Yeah. Probably a few dozen hours. Okay. Entirely personally. So not for work, just out of curiosity and seeing the level of control that exists.

(Paul Salvini at 00:18:48) So I've done that, you know, maybe fifty hours, something like that. And probably the same number of hours on the text generative tech side of things where it's, you know, looking for ideas, looking to evolve something. And that's pretty incredible too. So the idea of being able to brainstorm and have an AI agent to brainstorm with and to suggest ideas and see a result and you say, oh, that's a really good point. You know, I don't like the other nine points, but that one I never thought about that.

(Paul Salvini at 00:19:21) And so, you know, I find it's an amazing source for inspiration. But not necessarily today, not necessarily just taking the end result and saying, ah, great, done.

(Joel Beasley at 00:19:34) Oh, yeah. It's like a curator. Right? Where it does the cognitive load of creativity, and then you can pick the best pieces.

(Joel Beasley at 00:19:44) That's how we use it very successfully at our company. I haven't spent fifty hours on the video prompts, but I did do a couple interviews, and I don't know, Josh, did that interview air that we did with the guy who said that it's gonna be two to three years until movies will be generated with prompts? Did that air yet? No. It hadn't aired yet.

(Joel Beasley at 00:20:08) Alright. So we talked to, like, Sora people. We talked to a couple different people that are building these tools, and there's obviously some big problems in them. For example, what you said is persistent. The aspect that you're gonna have, like, a character, and that's gonna persist through multiple scenes, and you need the same creature to look the same in multiple scenes. So there's a couple big issues that are preventing the short little clips from turning into long movies.

(Joel Beasley at 00:20:35) However, they believe that that will be solved in about three years.

(Paul Salvini at 00:20:40) Yeah. I think the right way to solve it is maybe, again, breaking it down into, you know, what is the story that you want to bring to life? And so if you think of a story as first, there's a location or a set of locations that you want to, you know, that you would normally film in if you were filming a story. And so you need to have those locations created and that's easy enough. And if it's a fictitious location, fantastic. If it's a real world location, probably also possible given the amount of information about real world locations.

(Paul Salvini at 00:21:12) You know, typically again, for us, like if we did Last Night in Soho, you couldn't just film downtown London the way it is today. You really had to restore it to, you know, 1950s, 1960s look, and again, there's great AI tools that will give you the style or look of that. And they might not be exactly right. And so again, depends on the level of authenticity you're looking for, right, or the level of reality you're looking for. So you've got your locations, then you have all of the different props that you're going to use. So it can be, you know, fixed props, moving props, things like that, vehicles, whatever it is. And then you have your characters, the creatures, the humans, everything else that are going to appear in this piece of content.

(Paul Salvini at 00:21:45) You know, I think the right approach is to have AI able to help in the creation or in some cases entirely create those things. And those things have to have sufficient intelligence and an ability to be able to perform the way a human would perform or whatever else would perform. And that'll be the right approach. And when that happens, and I don't think that's three years, but when that happens, then you have the ability to do the kind of work that we do, heavily enabled or empowered by AI. But I would still say, you know, again, probably you want to tell a very specific story and you want to tell it with purpose.

(Paul Salvini at 00:22:49) And could AI generate that story maybe. And, you know, and pitch it and get people to agree. It's going to be a fantastic story and curate exactly the right combination of actors and individuals to appear in it. Sure. But I don't think that's a three year thing again.

(Paul Salvini at 00:23:08) I think we're going to see an incredible advance in what is possible, but there's still going to be a pretty big gap between that and what ends up being used. And it's like that in every industry. I mean, we've talked about having self-driving cars for many, many years now and wow. Is it ever amazing? What I say we're quite there yet?

(Paul Salvini at 00:23:27) Oh, not really. I mean, you know, like, we're...

(Joel Beasley at 00:23:30) We're definitely there.

(Paul Salvini at 00:23:31) We're definitely there.

(Joel Beasley at 00:23:31) This is something somebody said yesterday too. They're like, we're not quite there. And I said, look. I appreciate that we're not to perfection, but I have physically gotten in a Tesla, pressed a button, not touched anything, and arrived at the destination. And by my standards, we're there.

(Paul Salvini at 00:23:47) Yeah. Yeah. So it's amazing. But I'm just saying that in terms of actually changing the industry and things, and of course, you know, I'm an early adopter. I got a self-driving Tesla, the millisecond they were available to me here in Canada, you know, waiting for each beta version. And no question. It is incredible, but you get a nice snowstorm here, up here in Canada and you can't see the lines on the road. I can barely drive and clearly it can't drive, and you can't blame it for not being able to drive. It's like, that's a difficult task.

(Paul Salvini at 00:24:19) But again, so getting to where we are today, which is like clear day, nice road, you know, fairly well-known track. Yeah. We're there. Yeah. And it's incredible.

(Paul Salvini at 00:24:31) But all I'm saying is, you know, there's a reason why we're not there there. And that's that final P, same as the airline industry. Right? So my son's, you know, going to be training to be a pilot, a professional pilot. And it's like, wow, well, how does that job still exist? Like we've had the ability to, you know, remotely control planes for a long, long time now. Planes can, when things go well, fly themselves and land themselves and taxi themselves to the gate. And yet still an incredible demand and it doesn't seem likely to go away and you think, but we're so close. How is it right when we're so close? How is it that that industry, you know, is still something where it seems like in the at least near future for the next ten, fifteen, twenty years, you know, there's still this incredibly strong demand for real pilots?

(Paul Salvini at 00:25:28) And so it's interesting, right? That final, like, we should never underestimate the challenge of that final, whether it's 1% or 5% or whatever it is that gets us to, you know, to perfection.

(Joel Beasley at 00:25:42) Well, I fully agree by the way. Do you think actors are gonna, when are we gonna replace actors?

(Paul Salvini at 00:25:47) Well, it's a, you know, that's another great question. I don't think we will. You know, what is an actor, I guess, is what I would ask, in the sense that, is it, is it someone's, is it their fame, right? Is it the fact that that individual is well-known, has a brand, has a following? If you're saying, are we going to replace that? I doubt it.

(Paul Salvini at 00:26:13) Because you could take somebody that's maybe even better looking than that individual, but maybe doesn't have the same following, probably wouldn't have the same box office result as a film that has somebody that's well-known. And then you talk about the performance. Well, let's say, okay, we love the look of this person, but they're not the best actor in the world. And there are some incredible actors that sadly got shortchanged in the looks department. Right.

(Paul Salvini at 00:26:41) So at what point do you say, hey, what if we can combine a look that we want with a performance that we want, etcetera, etcetera. But what happens when that isn't a real person where people can't, you know, watch the Oscars to see that person or understand what's going on in their lives or follow them the way that they're real. Right. So that's where it becomes challenging to say, yeah, sure. The technology could exist to absolutely replace, if you will, what you see on the screen with something that is either looks exactly like you, but isn't you, or is you, but doesn't look at all like you, etcetera.

(Paul Salvini at 00:27:25) So we have all of that capability, even today. Now the question is, well, what do we do with that? And I think it's exactly like what you said about the pilots and, and first of all, I'm a big proponent, obviously. Pilots—I'm a commercial pilot myself. So I hope that never goes away. Like, and obviously we're there when things don't go well.

(Paul Salvini at 00:27:47) Right. It's the exceptions that you want to be able to handle. And so, but it's the same question about that in, in that case, it was well, what does, what are the people in the back think about there not being a pilot there? Are they comfortable with that? And I think the same thing for the industry would be, well, what do audiences say about that?

(Paul Salvini at 00:28:04) Are they willing to take an AI-generated character that, you know, no one knows where the performance came from or it doesn't really match the visual look of anyone that anyone knows. Is that a good thing or a bad thing? And I'd say, not sure. I think what's fascinating about it is it does open the door to things like, hey, we could create a James Bond series where James Bond always looks the same. Right.

(Paul Salvini at 00:28:31) And that's always been a challenge is that actors age, you have to change actors, preferences change over time and things like that. So the ability to have that level of control over the visual appearance and decouple that from performance is incredibly powerful. What our industry chooses to do with that will be very, very interesting because I think it does open the door to some really incredible and powerful abilities and capabilities. And so yeah.

(Joel Beasley at 00:29:04) I think we'll end up somewhere in the middle. For example, there's already influencers making $100k a year that are just digital fake people. And there they don't even hide it. They'll do Victoria's Secret brand deals. They'll do all types of deals, and they talk about it and publish about these fake characters that they've made.

(Joel Beasley at 00:29:22) And you should see in the comments, there are some people that really don't like it. But then there's also people that think it's the coolest thing in the world. And to your point of AI-generated stuff, I mean, cartoons. Right? Like, I like Bluey.

(Joel Beasley at 00:29:34) Because I've got small children, so I see Bluey a lot. But, yeah, you can fall in love with characters that are fully digitally generated. But at the same time, I also am more likely to watch some of the animated movies, like if Kevin Hart's in it, or if, like, somebody that I enjoy is in it.

(Joel Beasley at 00:29:50) And I enjoy them. And, like, why I enjoy people is, in I think there's a lot of similarities to why people enjoy certain music. As you go through stages in your life, heartbreak, growth, difficult moments, and the media is all around you. And as you go through these stages, you'll pick up and identify with some of these people, and maybe they'll have a story. And so there's a lot of dynamic, there's many dynamic things happening as to why we like characters and who they are.

(Joel Beasley at 00:30:17) I do believe that in the next, let's say, fifty years, there is going to be a push for, or the push for real and authentic will continue, but in the form of, I'm gonna be the old man that's like, I watch the movies with actual people, you know? Like the kids these days, you know? So I think that's a really likely path is that we'll tend to want, there will be a market for the organic actor. There will be a market for that. But that doesn't mean you won't have a favorite digital character like Bluey.

(Paul Salvini at 00:30:52) You know? Exactly. Exactly. No. I think you're 100% right.

(Joel Beasley at 00:30:57) So what other topics did we wanna get out there today and discuss?

(Paul Salvini at 00:31:02) You know, I think it's important to maybe talk a little bit about the evolution of what's happened in our industry. I mean, I think that that piece is, you know, for technically-minded people understanding the role of technology in filmmaking. I think that's fascinating. I mean, I, as I said, I grew up wanting to be a photographer. I was in love with the idea of capturing images and moments in time and being able to share those with others.

(Paul Salvini at 00:31:31) But then I went off and did math and computer science at university. And I didn't think those two worlds could and would ever come together. I didn't think a career in computer science could easily lend itself into a world of creative and visually creative work. And so blessed I am to have the role I do in this industry. And I think that's interesting for people.

(Paul Salvini at 00:31:58) I think it's interesting to know that a lot of what we see these days is being supported and created with technological innovations that are underlying it. And as those innovations happen, as it becomes easier and easier and cheaper and cheaper to do things digitally, we open the door to those kinds of experiences being made available to other people. And so whether it's, you know, being able to do a battle sequence with thousands of creatures and characters, that's a pretty incredible thing to be able to watch and bring to the screen and things like that. But really that could only be done with technology today because we don't have the budgets to actually hire 2,000 people. Right.

(Paul Salvini at 00:32:46) Some people do, but to, to put them in. But I think that's what's interesting is to just see how much of a film today is touched digitally and how important technology has been to this industry. Because again, it's one that we sometimes see movies that are very obviously containing visual effects. And I can tell because my friends will contact me and say, hey, did you guys work on such and such? And it's like, yeah, yeah, we did. Or nah, that wasn't us.

Paul Salvini at 00:33:12
And then other times we think we did an incredible job and nobody asks anything. And it's like, hey, why didn't you ask me about such and such? And I was like, oh, I didn't think there were any computer graphics in it. Wow. Well, yeah, there were thousands of shots that were done digitally or touched digitally in some way.

Paul Salvini at 00:33:32
And so I think that's one thing for people to know is just the impact that technology has had, not only on how we capture performances, how we capture images, but how we put all of that together to create that end product. So I think that's really interesting to say. And the other one that I think we touched on is just the whole democratization of these tools, which I think is incredible. Again, everyone can be a photographer now. Everyone can clean up an image. Now it's like, oh, you don't have a lens with a shallow depth of field. Let's just change that after the fact. So being able to do these things and putting tools of that power in the hands of the masses is pretty exciting, pretty rewarding for us.

Paul Salvini at 00:34:03
Maybe the other one I'll touch on is just we've created a new interactive division called DNEG IXP, and that's to take what we do at the very high end of film and to create all kinds of other, whether it's entertainment experiences or VR experiences, things like that, that again leverage content creation. One of the things that's neat is when we work on a movie or a series, like I said, we create these incredible locations and you can put on your VR headsets and walk around these locations and experience them.

Paul Salvini at 00:34:58
So I can think of three different films or TV series that, you know, I'd love to see them in different ways. So one of them would be the Chernobyl TV series. Like, absolutely incredible. People probably didn't think there were any visual effects in that at all, but of course, naturally there had to be. But you can actually—I mean, that entire city and that reactor was built digitally so that you could walk through and see what's going on. And that's something incredibly powerful to be able to experience beyond the way that it was shown in a film or Last Night in Soho, again capturing what London looked like, what Soho within London looked like in its early days, and to be able to just walk around with authenticity. Kind of like if you had a real date slider on Google Street View that could take you back to 1940—like, it's incredible.

Paul Salvini at 00:35:57
And it's sort of a shame that those experiences aren't available to the broader world. And you start to see glimpses of it. So, you know, Microsoft on their last Flight Simulator did a Dune release where you can fly the ornithopter around on Arrakis. So it's like, okay, that's pretty cool. So it's things like that, but I think are the byproducts of what we do in the film industry that over time will be made available in other forms of interactive media. And that's pretty exciting as well.

Joel Beasley at 00:36:28
I'm currently waiting on a babysitter so we can go see Dune 2. I'm like, it's incredible data. Babysitter will become available or it will hit streaming services. I just track what's going on.

Paul Salvini at 00:36:41
Yeah. No, that one definitely is one worthy of the big screen. And, yeah, I'd find a babysitter.

Joel Beasley at 00:36:48
Hey, are you guys building—I know we talked a little bit that you're looking for efficiencies and growth, but improving your processes and systems and making the company better. But do you have any Hail Mary projects that you think are the future as far as text prompt video generation that you can talk about publicly, or do you not have any of those?

Paul Salvini at 00:37:08
Well, yeah, we certainly have three areas that we're working in which I think have transformative opportunities for us. One is certainly AI, and we've talked about what that means. But of course, when I talk about AI, it's in the context of creating the highest end possible work. So it's transformative in our workflows, but it's not sort of the text to final video output. It's to create the elements that we need to do the work that we do and the way that we do. But that's still enormously transformative for our industry.

Paul Salvini at 00:37:42
The other one, the next one's real-time. So instead of having to wait hours, 20 hours, 30 hours for a single frame to render, to be able to see what you need to tweak and modify, being able to allow our artists to work much more collaboratively, much more interactively is going to be a huge and disruptive benefit to workflow so that you can really see things come alive and adjust things on the fly. And that comes with the opportunity to do things in a much less linear way, a much more collaborative way. And that's enormously exciting. And then the third one is just some of the work we're doing in creature and character and digital humans, to create just the highest possible quality and fidelity digital humans and all sorts of imaginary creatures and characters as well—everything from photorealistic to highly stylized, cartoony.

Paul Salvini at 00:38:43
And so those three are the three main areas of research where I kind of think we're hitting it out of the park. I'm so excited about the work that's emerging in those areas that will definitely be transformative for our business.

Joel Beasley at 00:38:56
Yeah, a couple years ago, I saw this company came out, I think at CES, where they did these unbelievably real looking characters. Like, I couldn't tell the difference. And then they, I think, almost instantly got bought up and shut down by Samsung or something like that.

Joel Beasley at 00:39:15
But it was pretty cool because I saw it come out. I was doing the show at the time, and we've been doing it for almost 10 years. And so I saw it come out. I was like, oh, this is so neat. And then I looked at their web page, I sent it to my producer and all that. I was like, we're going to do an episode on this. This is going to be so great. And then I think within a month, the site, it was just like shut down. And I think it was just like Samsung project now or something like that. I'm not sure if that's the company, but one of the big companies bought them and essentially moved the project internal for lack of a better reasoning.

Paul Salvini at 00:39:44
Yeah, be interesting to know why. But no, I think we're already doing it, obviously, to be able to replace characters and make sure that we don't put talent into dangerous locations when filming and things like that is essential. So, you know, being able to have credible digital doubles has been a mainstay for us. But more and more being able to do that in a way that withstands, like, full high resolution, four K, lead character, where it's transparent, it's seamless, that's obviously really important to be able to do for a bunch of reasons. And so that's something that we're definitely working on.

Joel Beasley at 00:40:31
I think it was Neon. Have you seen Neon yet?

Paul Salvini at 00:40:34
No, I haven't.

Joel Beasley at 00:40:35
I think they're back, actually.

Paul Salvini at 00:40:37
Oh, okay.

Joel Beasley at 00:40:38
But I just did a quick Google search because it was bothering me. I'm sorry.

Paul Salvini at 00:40:41
No, no problem.

Joel Beasley at 00:40:42
But if you're in the business of making these digital humans, to see Neon, and you might see it and be like, oh, I know them because of how it looks. But man, is it so real. Like, it's very, very real. Now, I haven't interacted with one in real-time. You just see the video where they hit play, and that's a very big difference because you've got milliseconds to figure stuff out to have human reaction. That's an entirely different thing. But as far as staging a video and hitting play and you being like, that's a real person.

Paul Salvini at 00:41:04
Very cool. Yep.

Joel Beasley at 00:41:04
Yeah.

Paul Salvini at 00:41:14
World's getting there, that's for sure. It's making it harder to know. So I think one day you're going to do an interview with me, and it won't be me.

Joel Beasley at 00:41:23
That's the surprise for today's interview.

Paul Salvini at 00:41:25
It might not be me at all.

Joel Beasley at 00:41:29
Well, I was thinking about that too, because that would be really nice to just send my digital twin to a meeting that I didn't have time to go to.

Paul Salvini at 00:41:36
Super convenient.

Joel Beasley at 00:41:37
Right? Yep. Like, why wouldn't we? Why wouldn't we? That is totally a thing.

Paul Salvini at 00:41:43
But wouldn't you keep that as a secret? Because you don't want everyone to have that, or you lose your ability to clone yourself in exciting ways.

Joel Beasley at 00:41:52
I have to build a second clone to monitor the first clone.

Paul Salvini at 00:41:55
That's right.

Joel Beasley at 00:41:56
Yeah. It'll get messed up. By the way, have you seen the—I know I'm asking you, have you seen a lot of stuff, and I'm putting you on the spot a bunch. But I'm just—you're nerdy and cool to talk to, so I'm excited. The autonomous agents, where agents task other agents to get things done and achieve an outcome together?

Paul Salvini at 00:42:14
I have seen demos of that. I've also seen ones where it's collaborative autonomous agents where they have to work together to achieve a task.

Joel Beasley at 00:42:21
That's the one I'm talking about.

Paul Salvini at 00:42:22
It's frightening. Yeah. That's next level scary. And, yeah, where they dispatch a drone of, or a bunch of agents to do a certain task. Yeah, you wonder how close we are to the Android. That's where, again, you know, as a technologist, I love it. I love the new capabilities and emerging possibilities, but it is a little bit terrifying in terms of how then these technologies are going to get used. And you just have to hope that the responsible uses and that the good uses will outweigh the bad, as with all things technological.

Joel Beasley at 00:43:02
Yeah. I actually think it'll expose more of this concept of good and evil inside of our reality to look at them more as like a force at play that can use anything and everything. And I'm only 36, but I have three kids. But as I've been growing up, I've become more interested in that part of life because it starts to emerge as a more important part of your life, especially when you start raising humans. You know, like, wow, there really is good and evil. Yes, we've got these behaviors that we're more likely to lean to, but you can definitely start to see that as you get older.

Paul Salvini at 00:43:37
So you just need a couple robotic babysitters, and then you can go see Dune 2.

Joel Beasley at 00:43:41
That's all we need. That's all we need. I prefer the organic babysitters. Oh, man. This is good. This is good. Okay. Well, hey, I think we covered everything. We talked about a lot of stuff.

Joel Beasley at 00:43:58
Is there any call to action? If people want to go work at Double Negative, how do they go work at DNEG?

Paul Salvini at 00:44:04
Yeah. Well, on the tech side, give me a call for sure. And we're always looking for people that are looking to change the way that films are made and find or change the way the content is made, find better ways to tell better stories. And so those who have that interest, it's a truly incredible and fascinating field. So, yeah, encourage people who are interested in that for sure to reach out. And yeah, look to this industry because I think it'll be really exciting times ahead in terms of what's possible and how we use technology.

Joel Beasley at 00:44:40
I'm also pretty optimistic about your specific—I'm not being nice either because you guys are experts in this storytelling and this digital. So, yeah, the tools will improve, but the behavior patterns of going to the experts with the history, that's not going to change for generations. I mean, people are still going to come to you guys. And, yes, your models might change and the roles within your companies and the tasks that humans do might change. But I think you guys definitely have a long run ahead of you.

Paul Salvini at 00:45:08
Yeah, absolutely. I mean, I always say the job is somebody wakes up in the morning and has a fragmented vision in their head about a story that they want to tell. And the role is, how do we complete that vision first of all, and make it a coherent and awesome vision? And then how do you communicate that to the thousands of people, you know, shape it and then communicate it so that what you end up with is just a truly incredible, amazing product? And I think that's one of the things I love the most about DNEG is that we do partner with our filmmakers and we really do consider ourselves as partners in production where our job isn't just to do work. Our job is to make a better story and strengthen the product and wow audiences. And I think it shows in some of those movies you talked about. I hope you'll see that in Dune 2. It's an incredible story brought to life.

Joel Beasley at 00:46:04
I'll message you after a bit. Excellent. I'll give it the Joel rating. It'll be A-plus. I'm a huge fan of Dune.

Joel Beasley at 00:46:10
All right. Well, we did it, man. We made a podcast. You guys are the leader—you're the world leader in VFX. And I got to hang out with you for a little bit. Thank you so much for doing this.

Paul Salvini at 00:46:19
My pleasure. It was great. Really enjoyed it. Thanks, Joel.

Joel Beasley at 00:46:22
Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.