Russell Palmer: Empowering Creators - How AI is Shaping the Future of Filmmaking

Speaker A: We're creator, creators, and we're going to save the arts from, uh, these robots. I think people are a lot more receptive, and especially now that we see, you know, Ridley Scott used AI for the rhino in Gladiator 2. You know, no one's going to blacklist him or give him a death threat to keep creating work. Copyright. Founding fathers said in the Constitution this is to incentivize people to continue creating.

Speaker B: Welcome back to the 8 Growth Ventures podcast. Today we're excited to have Russell Palmer, CEO and founder of Cyberfilm AI. Cyber Film is using generative AI to help filmmakers create Hollywood quality movies on an indie budget. Russell's journey from Microsoft to Silicon Valley and now to disrupting the film industry is truly inspiring. In this episode we'll talk about how AI changing filmmaking, the mission behind Cyber Film, and what's next for the future of storytelling. Russell, thank you so much for being on with us today.

Speaker A: Yeah, thanks for having me on. Uh, I'm a fan of the show and uh, yeah, it's a pleasure to be here.

Speaker B: Great. Russell. Like the way we kind of do the podcast for most of our listeners, um, that are the first time tuning in is we kind of go through a chronological order of your past, present and then the future. And what I would love to get started with is a little bit of your kind of past and what's been sort of the foundation and influence of you, uh, going to where you are today and how you see the world. So let's take a little bit step back. You worked at Microsoft for a little while, right? So what year was this roughly? Around what time?

Speaker A: So right after college, uh, so about 2008, I moved from, uh, from Toronto, where I'm from, uh, to Seattle. And I, uh, worked for about eight years at Microsoft in the Office division as ah, sort of a user experience, um, um, program manager. And uh, one of the things I really excelled at was, was using Big Data was very hot at the time. So using big data in addition to usability studies, what are people clicking on? What features are not used? How can we um, improve it based on Microsoft's strategy? Um, this was also when this was before AI was uh, officially granted to the term. It was just starting to be called machine learning and deep learning. Um, I really got into that Microsoft research, um, was doing some amazing work there. We were building, um, these magic features in Excel which would auto you type one thing and it would figure out what you want and apply it to your entire Excel.

Speaker B: Oh, you built that? I love that feature by the way.

Speaker A: Yeah, yeah. So it was, it was great. It was also a pleasure to be there because, um, Microsoft Office, some people don't know, um, is likely where the PM profession was invented. So when I first joined, I was trying to read as much as possible, learn as much as possible in my first job. And I found a memo from someone I think back in the 80s which basically said, you know, we've got business on one side, the engineers on one side, they don't like talking to each other and we're building things that end up not being used. And the customer says, well, this is nice, but it's not what I wanted. So there was a letter to Bill Gates directly which said, I think we need to create a new role to manage the program. And they called it the Program Manager. Now it's called a product Manager. But, um, yeah, so I sort of learned my, uh, my traits at the place it was invented and um, it was great. I just learned so much being there.

Speaker B: Amazing. What do you think was one of the biggest, besides learning the profession and the craft that you become great at? What, what would you say was one of the biggest early lessons on that has still held true after so many years later for you?

Speaker A: Yeah, the, the big one is, I mean, at Microsoft, PM is an entry level position. Uh, they prefer computer scientists, engineers. That's, that's my background. Um, but, um, uh, sorry, I just lost my train of thought there. But, um, yeah, yeah, the, the idea is you just came out of college, you're not managing anyone yet, but here's your team of three developers and two testers and a designer. So what they say is you need to lead through sort of inspiration, um, and positivity and by being smart and having the right ideas. You cannot lead by saying, hey, I manage you, Developer X. Do this no matter what I say, or else I'll fire you or you won't get a promotion. They are managed by a dev manager. So it's really up to you to convince these developers, sometimes who are much older and more experienced than you, that you have the right idea is, and you convince them to build it. If they don't want to build it, they can tell their dev manager, this person doesn't know what they're talking about, you know, get me a new pm. And so you need to lead not by authority, um, by just the quality of your work. And I think even now as a CEO, I still try to do that. Um, I've had some bad managers in my life who do lead by, you know, do What I say, no matter what, even though I think it's wrong and don't speak up in the meeting. Uh, one of Google's surveys, um, of their best performing team said emotional security is one of the most important factors. People need to be able to speak up to authority. And so now as a CEO, I make sure I never say, you know, do this because I'm in charge. It's. Well, tell me your concerns and let me see if I can convince you or we can compromise. So I think.

Speaker B: Or the why, right? Like the why of what you're working on, the vision.

Speaker A: Exactly, exactly.

Speaker B: The impact it's going to make and um, the reason behind it.

Speaker A: Exactly. No developer wants to work on something, they're not convinced why it's going to, you know, be a big hit and get them a promotion. And people want to see impact. No one wants to spend a year building a feature and then find out, as I said, we're measuring data, find out no one uses it. Um, so the why. Yeah, absolutely perfect.

Speaker B: And then you transition from Microsoft to a Samsung subsidiary. Viv Labs. Can you tell us a little bit about, about that? And, and here you started working on Bixby, which is probably the first generation of AI agents. Is, is that, is that a fair way of, uh, explaining it or can you talk a little bit?

Speaker A: Absolutely, yeah. Yeah. It's really exciting to sort of be at that early stage. And in fact, the founders of Viv Labs, um, invented Siri. So they were at SRI Samsung Research Institute, I think Siri, that's kind of what the name comes from. Know, sold it to, uh, to Apple, uh, Steve Jobs apparently called them, you know, once it launched in The App Store, 30 nights in a row until, until the founders would come over for dinner, you know, have some wine by the fire. And the. It's funny, they, they had so many Steve Jobs stories, but they said, uh, we didn't want to sell. That's why we weren't picking up his call. But within 30 seconds of sitting down with him, like, not only were we going to sell the company, we would follow this man, you know, into the. Do whatever he says.

Speaker B: Talk about leading by vision, right?

Speaker A: Yeah. Or some call it like the Steve Jobs reality Distortion feel like he just convinces you very quickly and you believe it a hundred percent. So, um, yeah, I mean I started at Microsoft. It was, luckily it was the first job I applied for and I got it and I wanted to move to the US So I was in Seattle, um, but I was there. Yeah, once you're there I think five or six years you transition from the now controversial H1B visa to uh, I got my green card so I could move and switch jobs. And um, yeah, I thought I'd learned enough at Microsoft and contributed office productivity software was not my passion. And after working on machine learning and now seeing AI coming to fruition, wanted to join an AI company. So um, yeah, moved down to San Jose, joined these Siri founders who had left Apple, um, to start what they really wanted, which was Apple mixed with the App Store, which is meaning to say they didn't think Apple was adding features enough to Siri. If you remember, it was good at first and then kind of slowed down. Um, they wanted anyone to be able to. So Spotify could integrate. Hey, you know, here's a capsule, we called them. Um, so you could just say play me the best song from the 80s and it'll do it or play me a song I'd like based on my playlist. It'll do it. Um, and uh, you know, uh, I would drive up to Uber and say, you know, here's our new platform. Do you want to build on it? This will be on there, uh, was an acquisition by Samsung. So this will be this Siri for Samsung called Bixby. And there's a, there was a button on it you could press and say in the morning call my Uber, you know, and it's smart enough to know, okay, you're at home based on the time you're obviously going to work. You usually pick this type of UberX car, you know, let's do it. So yeah, I think it was one of the first, um, AI agents that were extensible, um, and so learned a lot about natural language processing, um, and training models, but also building a feature that would scale. So you know, Samsung is on and multi language. So Samsung iPhone's most popular in the U.S. but Android phones are everywhere, uh, else in the world. And so they put Bixby on their new smart speaker, sort of the Amazon Alexa competitor. They had a smart refrigerator you could talk to. Uh, they had a smart tv, you know. So all Bixby was everywhere hardware. And so it was.

Speaker B: How adopted is Bixby? Like when you at that time, did it get quickly adopted or was it

Speaker A: God, yeah, immediately adopted because people on their Samsung devices, they didn't want to switch, but they wanted something like Siri. It was so hot at the time and Alexa. And so as uh, soon as it came out, you know, we went from zero to one in terms of, you know, a startup with an idea to an acquisition launching on millions of devices worldwide. And so, yeah, we had millions, tens of millions of users almost from the first week. Um, and so that was um, that was great. And the best part was I worked with um, Adam Chyer, who uh, is the opposite of the manager I mentioned before. The nicest guy in the world leads with positivity. Everyone loves him and he had just his track record. You know, his first startup sold to Apple. His second sold to Samsung. His third, uh, that he just worked on after, uh, sold to Airbnb. I mean the guy is just uh, amazing. So I, I sort of learned like you know, a group. I was there when there was you know, five people in a group of 50 and then more. So I saw if you have the right people, the right team, the right idea, you don't give up on yourself and you're sort of, you have a good product, um, uh, anyone can build a startup. And so that's what really um, planted the seed in my head. Okay, I not only want to work on AI, uh, I want to do a startup now that I'm in Silicon Valley.

Speaker B: How old are you at this point in the journey?

Speaker A: I just turned, I just turned 40. So yeah, 15 plus year career so far.

Speaker B: Amazing. And then, so what transition at this point? You had an incredible experience obviously. Uh, you initially worked at Microsoft where PM was invented and you kind of got a great experience there. And then you transitioned to this incredible opportunity at ah, Web Labs and releasing a mass product and really getting, you know, so here you are in your career. What changed for you? Ah, when did you, at the initial thoughts of cyber film AI start coming to fruition? And when did you say, you know, enough is enough time for me to really go after this and start building this?

Speaker A: Yeah. And um, I mean I love Canada, I'm proud Canadian, just got my dual citizenship, proudly. Um, but um, I was under the impression this was before the pandemic and it was um, easier to work remote M. And so I thought I had to be in Silicon Valley to meet the right people. I mean step one as a co founder, maybe even before you have an idea. And um, so I was trying to just be around people. I actually, after, you know, a few years at Viv Labs, um, got, got recruited to JP Morgan, had just opened up a new AI lab, built a huge new campus on, on Stanford Research park. And the team was.

Speaker B: You've been at the skunk works of all the major companies?

Speaker A: Yeah, I guess you could say that. Well, and, and it's interesting Too, you know, biggest software company in the world, Microsoft, biggest hardware company in the world, Samsung. And now biggest, you know, oldest bank in America. Um, and so it, and it's for inspirational CEOs, you know, Bill Gates, Jamie Dimon. Like, people I'm learning from are just incredible. Uh, I wasn't quite prepared yet and I hadn't met the right people. Um, and so this team, I just couldn't turn it down, you know, that it was recruiting. The first five people were from Google, you know, the person who added AI to improve Google Translate and Google Maps. You know, uh, a couple years ago it got uh, order of magnitude better and that was with AI. And, and he led that team. And then they recruited the best people from Amazon, from Meta, um, you know, Microsoft. And so I just thought this all star team was great. Not only could I learn some new things we were building.

Speaker B: I love how humble you are. You're like, I was surrounded by these all star team. I don't know why I was in the room, but I was. Yeah.

Speaker A: Well, it's funny, they called me the Microsoft guy. It's almost like they're collecting one, one for me.

Speaker B: They're like, I guess we need the Microsoft guy.

Speaker A: Yeah, yeah, well, but, but it's also because it was so new maybe that I was humble because um, you know, Amazon Web Services, we decided to build the bank's platform on top of it. There was a big deal between JP Morgan and aws, you know, one of these billion dollar, ten year deals. And um, ah, you know, it was a chance to learn, okay, how do we build things from scratch on us, like literally set up a data bucket and make sure, you know, JP Morgan's big thing was our data has been for 200 years on these Cobalt servers because we just don't want to change them. And no one's hacked them yet. So they work, but they realize it's time to move into the cloud. But they're like, oh God, you know, if there was a recent Amazon hack where a disgruntled employee let out some private customer data into the world. And so our job was to make sure that would never happen. You know, single tenant, top encryption. But yeah, they ended up uploading a lot of their data to, to Amazon and then using services like SageMaker, um, um, and Spark to, to train on it, do data analysis. And so I worked. We were sort of a hub and spoke team. Like we were in the middle with the five um, business line of business divisions around us. And so they'd come and say you know, um, we want to be more accurate in our loan approvals, for example, um, help, help us build something that would, you know, allow our, our, our employees to say, you know, all these are 95% sure these people are going to pay it back. Here's the hundred you need to be more focused on. And so, yeah, again, it's not about, you know, people hated when they heard that, um, that the healthcare company recently had turned on AI to just turn down people's, um, um, you know, funding, uh, reimbursements. And obviously there's a lot of anger about that. So they wanted to have something where just make our employees more effective and more faster and more accurate. Um, and so there were a few other projects, um, that were kind of secret, so I can't name. But yeah, you can imagine. You know, they tried to build an AI that would predict the stock market. Turns out it doesn't really work, um, because just things happen, things change and that previous data just can't always account for. But um, yeah, they had a ton of projects, uh, Omni AI it was called. And um, it won an Award from C100 Magazine as the most innovative tech project of the year. But Covid hit and um, so I was working from home and you know, everybody's saying, oh, it's the sort of the great resignation where people are saying, is this the thing I want to do for the rest of my life? Or maybe now it's time for a change. And I just realized after a couple years, um, finance is not my thing. And so to get back to your original question, ah, filmmaking is a huge passion of mine. During the pandemic I just had a lot of downtime. Stanford, uh, continuing studies, put all these courses online from their best professors, you know, AI product management, you know, how to build human centered AI projects. So I took a few courses, um, from a professor, ah, Dr. Ranjan Nag, who's sort of a famous entrepreneur and AI, um, investor. And um, in, in the class, you know, first of all we, we saw an early demo of GPT3, which was the original light bulb going off moment. In terms, you're like, wow, it can do this now it can speak English, robots can speak English. And at the time it wasn't a chatbot, so you'd have to type a sentence and finish it. And so when I got my chance to try it, I wrote, you know, it was the best of times, it was the worst of times. And I was thinking, is it going to copy D or is it going to invent Its own. And it did. It invented in seconds, this beautiful, poetic short story. Um, but the more I tested it, the hallucinations were there. You know, you try to write a story, it forgets the main character, it invents a new one. So I was talking to my brother. Uh, we grew up together as kids making movies. And, um, he now works. You know, he went to computer engineering, but same thing. He's like, I don't want to work a cubicle nine to five. He went back to film school and majored in creative writing. And so he works. Uh, a lot of movies are filmed in Toronto and Vancouver. So he worked on, like, the show Suits, the show, the boys, Superman and Lois, some. Some DC movies. And so he's telling me, like, well, here, you need to put structure around this thing. Like, what I learned in film school is, you know, here's how character arcs work. Here's how actors. What should be an act. One, two, three. Um. And so, you know, the professor, uh, said, write a paper on AI disrupting an industry. So I wrote it with my brother on Hollywood. And we tried to say, here's all the jobs that, you know, may transition to new things or be at risk. Like if you're. If you make puppets or models or costumes, um, or if you're a writer, you know, maybe. Maybe, um, it can scan in. That's what I did. I. The API came out, and I built a prototype for my brother. I scanned in all the novels and scripts he's ever written, and I tried to make him a writing assistant. Um, and he would take it on set and be working on it on his lunch break. And again, this was years before the strike or chat GPT people weren't afraid of AI, so they would come over and say, oh, I've had this idea in my head for years. You know, I'm a stunt coordinator, but I. I took classes during the pandemic on creative writing. I want to be a screenwriter. That's my dream. Type in my idea in the log line and see what characters and people were just blown away. And so we said, okay, maybe this.

Speaker B: Maybe there's something here.

Speaker A: Yeah, yeah, exactly. So to just, um. The final step was, um, another, you know, I try to look for the silver lining and clouds. Another benefit coming out for us of the pandemic was Y Combinator. You know, the famous startup accelerator, put their batch program online and called it startup school. It's really hard to get in, but this is free and anyone could do it. So I was just, you know, before quitting my job. We just tested it, uh, and they make you interview, uh, 10 potential customers a week. Um, and so we would, um, go on set and interview 10 people, and then they would, oh, my friends got to see this. Ten more. And, uh, you know, we must have interviewed dozens. Dozens. But by now a hundred people, you

Speaker B: were doing customer discovery and quickly getting feedback from potential users. And overall, what was the feedback you were getting?

Speaker A: It was great. It was great. People loved it. Um, the big question was, is this my idea? Can I keep it? And that's what led to. We can maybe talk about this later in terms of ethics and legality, but this is what led to our decision. I built this for my brother, not to steal his idea. So our policy is, unlike some of our competitors, you know, if, if you make it, um, it's yours, you own it. 100. We only use APIs and models that grant the same, um, and don't retrain on your work. And, um, so that was great. And, you know, a couple years later, the strikes happened, and that's what they fought for and won. And we were happy to say. I was talking to someone from the wga, the Writers Guild's, um, AI Committee, and we said, great, we love that you chose this. This was our policy all along. And he's like, yeah, let's, let's, let's

Speaker B: talk about that a little. Yeah. So you create this incredible product, then a strike occurs. And one of the major issues of the strike was just AI in general. Here, here you guys are in center stage.

Speaker A: Yeah.

Speaker B: Of, um, the actual. What the strike is all about. Right. And kind of tell me a little bit about that. And, and what. Where was your positioning and how was it received overall? And, and how were you guys able to show in general that you guys are here to. I guess we talked about this a little bit earlier to, uh, be more of a writing assistant and more give writers, uh, the created creative edge or helping them create better products. Right. If I'm understanding that correctly.

Speaker A: Yeah, it's, it's absolutely meant to help them. So, again, my brother working in Hollywood for 15 years, he founded his own production company where he, you know, he writes novels and, and publishes them on Amazon and he writes scripts. It's so hard, especially if you're not in la, to break in. So, you know, he was the guy, after every movie raps, he'd go to the executive producer. Hey, I, uh, did some good work for you. How about you? Could you read my screenplay, give me some feedback? You know, no one ever helps Especially if it's your first one. And, um, he had a lot of trouble converting his novel to a screenplay, which a lot of people struggle with. He ended up hiring what most people do, which is, you know, a ghostwriter, a co writer, a script doctor. Depending on the project, it costs tens of thousands of dollars. You know, our app is $20 a month, so. So the first thought was, okay, that's one thing we can say, like, people in Hollywood do this already. You're just replacing it, uh, the human you pay with an AI. Um, but, yeah, it was interesting. We launched the month before the strike started. And so, um, uh, you know, we got some feedback. Maybe just keep your head down for a couple months and build some new features. And we did. We built our script editor and our. And our storyboarding, um, app. But, yeah, I mean, leading up to the strike, my brother is in the guilds in Canada. The wgc, the Writers Guild of Canada, Directors Guild, Producers Guild. So we were tracking it, and I can honestly say in the months leading up to the strike, AI Was last on the list, and number one was streaming residuals. So the. To summarize that, you know, Seinfeld, they got paid a million dollars an episode by the end. And then also the show's been running for 20 years. They continue to get paid residuals for those reruns. When Netflix started and streaming, and no one thought it was going to be as big as it was. You know, Blockbuster turned them down for an acquisition. Um, yeah, it was just really interesting, um, to see that once the strikes started, the studios who were pushing back started releasing numbers. You know, we're trying to compete with Netflix now, and we're losing a billion dollars a year on Disney plus, and we're having to lay off people. And so I think the story was just bad. Like, give us more money from your division that is losing money and you're laying off staff. Like, it didn't. Probably didn't sound good on the picket line. And their storytellers, um, chatgpt just came out. And so I think one of the. Maybe it was, uh, David Zaslav of Warner Brothers or someone made a threat saying, oh, ChatGPT's here. If you don't like the offer, I'll just fire you and replace you with a robot. And I think that's what turned up the heat on AI and made it. It's a great story. We're creator, creators, and we're gonna save the arts from, uh, these robots. Um, so we kept our head down once the strike ended. And ruled in our favor. Uh, in fall we had, um, rebranded to Saga and relaunched and things took off. And one last funny story. Well, maybe not so funny. It was a little scary. Um, I had pitched because we launched. I had pitched, um, Jason Calacanis, uh, he hosts the all in podcast. He wrote the book on angel investing. He was the first angel in Uber. Uh, you know, and he invited me on and I did a pitch competition live on his podcast, front of millions of people. Um, and he loved it. And you know, oh, great product, great everything. You know, I'm going to give you the same offer I gave Uber, but I hate the name Cyber film AI. You gotta rebrand. He said, I'm the guy who told Uber to change it from Uber Cab, like the famous social network line. Just Uber. It's cleaner, right? So, uh, so we were looking for, you know, a four letter, sort of two syllable name that he, uh, recommended. You know, looks good on a T shirt, looks good, people remember it and, and it sounds cool and it represents what you're doing. Um, the bad part of it was I, I, uh, the next day he went on the all in podcast, which is listened to by millions, maybe tens of millions of people, and mentioned to, you know, the besties Chamath and David Sachs. Hey, I met this. He didn't name us because I hadn't signed yet, but he's. I met these two brothers, you know, one's a Hollywood filmmaker, one's Silicon Valley AI, product manager. This is great. And David Sachs mentioned, you know, well, there's a, there's a strike about to happen. They should not give, uh, studios a reason to learn, look for alternatives. And this just, I guess, went viral in the Hollywood community because the next day I got two things. One was nice, which was all my old friends, cousins, colleagues, you know, high school buddies who all listen. Oh, was that you on the podcast? Oh, they shouted you out. Great. On the other hand, uh, our Twitter just started immediately getting death threats, um, to be.

Speaker B: Oh, wow.

Speaker A: Honest. Yep. Um, a lot of hate, a lot of anger, you know, calling, calling us a scab. And um, so it was tough, but, um, you know, I would. A lot of people reached out, you know, some famous A list producers and would say, um, you know, I heard about this. I want to talk to you and see a demo. But I just want to let you know, I hate this. I hate everything about you and, uh, I hope your demo sucks. And I just want to confirm that, um, AI is not good. I tried ChatGPT, it's you know, it's cliche, it's bad at writing, it hallucinates. And, uh, one of my proudest moments as a founder was, you know, flipping that person and people like him to say, okay, I see. I see the vision, I see what you're building. It's a tool everybody uses. Script doctors, script writers, co writers. Um, it's just with an AI. And in fact, I like it because sometimes those people sue me. They want, you know, more of a writing credit, more of a percentage. This AI is not going to take any of my ideas. And I can, you know, it never gets tired, it never runs out of ideas. I just keep clicking, you know, give me a better name for my movie and to infinity, until you see one you like. And they always find one they like. So, uh, ups and downs, but we're in a better place. Better place.

Speaker B: So how many users are using Saga now? How many users do you guys have on the platform?

Speaker A: Um, we're not really releasing exact numbers at the time, but, you know, it's in the multiple Thousands, um, and $20 a month, so pretty affordable. But, um, yeah, not quite. You know, we're. It's been a couple years. We're still pretty early stage and, um, we want to do better. So, yeah, our growth is, you know, our target is, um, We've calculated there's. There's about 450,000 people, you know, if you look at all the people who are storytellers and work in media, so not only professional writers, but, um, novelists, uh, screen, you know, screenwriters, all types of writers. Game developers love it. So they like working on the game mechanics, but not the character names, the dialogue and the cutscenes and the plot. Um, and even commercial advertisers are, you know, one of our biggest requests is can you build a new template so I can just pump out like a, ah, 30 second ad? Even, um, a 30 second ad has three acts. They're just 10 seconds each. You know, it has a character that alludes to a backstory. You know, Jake from State Farm, we kind of know about him and his personality and, you know, there's dialogue. And eventually with tools like Sora now and Google's VO2, you know, you can generate 60 seconds of photorealistic, lifelike quality. So we're hoping to integrate those soon. And, um, so I think people are a lot more receptive. And especially now that we see, you know, Ridley Scott used AI for the rhino in Gladiator 2. You know, no one's gonna blacklist him or give him a Death threat. Um, Paul Schrader who wrote Scorsese's Taxi Driver and Raging Bull, um, just posted I, I tried AI and it's, it gives me better script notes than any executive in my 30 year career. You know, um, Tom, uh, Hanks, Robin Wright used the metaphysic AI to de themselves in the movie. Here m are starting to use it. Everyone I talk to says even if they're hating on it in public, they're, they're using it in the background. And Hollywood is very competitive people. It's a war, it's a war out there and people will use whatever tools and weapons they can to succeed because most people do not succeed. And it's not always a meritocracy. Sometimes it's just you don't have the. Even if you move to la, you don't have the right connections. Handing out your screenplay in coffee shops doesn't work. And so um, yeah, we're hoping this is not going to replace jobs, but just really help people. We put their name on it. They don't even have to say they use our app. We just want to see people like my brother, you know, who spent 15 years trying to, trying to break into a director. He's an assistant director.

Speaker B: But so can we go a little bit deeper into the product? So what I'm understanding so far is it's a tool that can help streamline and help you either come up with, help you uh, with the idea creation part, help you with the note taking part, help you refine your script. It's not writing it for you per se, it's more guiding you and making sure you're following all of the best.

Speaker A: It's helping. It's helping. Yeah. And in fact I had a lot of uh, students from USC School of Cinematic Arts, you know, taught top school in the world, possibly Toronto Film School where we're from best in Canada. Um, and it's. Some professors and students say it's like a teaching school. Like I have not yet read the incredibly long Hero with a Thousand Faces by Joseph Campbell. You know, maybe I read Save the Cat, but I haven't got to Robert McKee story or Syd Field screenplay. You have to read a lot. All the books my brother made me read when we started. There's a lot to get through and a lot to remember and especially if it's your first movie, maybe you can read about something but it's better to learn by doing. And so our app, you know the, and people can see it on um, uh, write on saga.com, you know, slash demos if you want to see a video of it in action or some screenshots. Writing is in, you know, W, R, I, T, E. Um, and my brother said, you know, when I did film school in creative writing, there would be a week on plots and a week on characters and then acts and how to structure a beat sheet. You need to plan this all before you even start writing your script. People do it today. They write on little cue cards. They put them on a cork board and then they'll have to like snap a picture to send it to people. And it's as you know, final draft was invented in 1990. It hasn't, uh, really changed since then. We kind of think this stuff needs to move to the web, to the cloud and take advantage of this tech. But yeah, so our app, the pages from left to right are plot, start here, then characters, then acts, then beats, then scripts, then storyboards, then video. You can go in any order you want because everybody has a different, um, process. But yeah, the idea is most of our writers, we do track this data and do sort of interviews and user surveys. Most people end up filling in like 60, 70, 80% of because it's not an idea generator. When we did the Y Combinator interviews, people said, I don't want an idea generator. I don't want something to make a hundred page script for me based on whatever it thinks is good. I have six or seven ideas. I've been tumbling them around in my head for years. People say sometimes years they've been trying to get a. So we say, okay, just start. Type in the log line. You don't know the title yet. That's fine. You know, the main character and their backstory. Okay. Type in their, you know, their ghost that haunts them from the past that makes them believe a lie. And in Act 1, there's something they want, but really by Act 3, if they have a positive character arc, they will, you know, uh, learn something, uh, and get what they need. And that's a happy ending. So, you know, some students will say, oh, I didn't know my.

Speaker B: You know, it sounds like the structure and the workflow is the name a similar process you would be going through anyways.

Speaker A: Yes.

Speaker B: And where your, uh, saga really steps in is it, it's, it is allowing you to no longer worry about the blank screen or getting stuck. And it just kind of triggers some new suggestions based, based off of Best in Class. And then from there it's not the best answer per se, but it gives you, okay, yeah, no I don't mean this, but I mean that. And, and then the, the collaboration begins. Right, so it's more of a writing collaborator or, or a co writer, per se. Is that a better way of describing it?

Speaker A: Yeah, yeah. And, and that's. So when you start our app, it's not, type a sentence and we'll make a movie for you. Like our, one of our competitors, um, they came out soon after us. Their app looks similar, but their whole thing is type a sentence and we'll give you a 10 minute movie. No one wants that. Ours doesn't prompt you to. It encourages you to start writing and filling in as much as you can. And then if you say, okay, I know this character, just take an example. Everyone knows his name is Bruce Wayne, um, but I don't really know his backstory yet, but I know he's this type of person and this is what he's, he wants. And, and so you can click, okay, what is the ghost? And it might say, you know, a criminal murdered his parents when he was growing up. And then maybe you go to the villain page and you say, okay, come up with a name and it could pop out something like Joker and say, oh, Joker was the one that killed him. And here, here's Joker's deal, uh, and backstory, uh, which we saw in the wonderful new movie Joker. And so the idea is if you know the character but not the love interest or villain or mentor, like a Gandalf, um, we, uh, can help you think of things to complement. So whenever you click Generate, even if it's like just what's this character's name? We fill in. Because we use GPT4 Turbo, it has like, I think 165K context window, which is equivalent to, you know, you can send it a prompt with like 500 novels of text worth of, you know, worth of text. So we, we send everything. So if you've written 10 pages of script, half the characters, two of the acts, some of this, some of that, and then you say, okay, what happens in Beat 40, the epilogue? We take everything, um, and it's Turbo now. So it's so fast. So in seconds it will say, here's five ideas for this thing you're asking for, um, based on your movie. And so to your point, you can click on it and say, okay, you almost nailed it, but, you know, I'm going to edit it a little bit, um, and then continue on. So, yeah, it's definitely. It doesn't write it for you. It's not this, um, you know, people on the picket lines.

Speaker B: They were worried that David, is your brother happy? How many successful, uh, stories have you been able to get out of? Get out?

Speaker A: That's quite a few. Quite a few. So we did actually, um, I just wrote a paper for IEEE Computer magazine. It's coming out in April, and we had to do a case study. So we used my brothers and we sort of did the average of all the 100 page screenplays he's done previous to AI and the average was it would take him two years because, you know, he works a day job, he's not paid as a writer yet, so he's trying to do it evenings, weekends. That's why it's so slow. One to one took seven years, which was wild. Um, but then he asked someone, I think at his, uh, studio or one of his colleagues or his agent or someone like, just give me a premise. And so he typed it, the premise in the log line, click generate, expanded it out, professional log line. He worked through it again part time and got all the plot, characters, acts, everything, the script and even some of the storyboards completely finished in ten days. Um, and sentence.

Speaker B: So from a couple years to 10

Speaker A: days, two to three years. Yeah, to 10 days. And when he sent it to people, because we wanted to do like a blind test to, you know, his friends, family, everyone he usually shares his work with with, a lot of them said, this is by far your best work. So it's not just about speed, it's also about quality. Um, and in fact, in the interviews, writers say, you know, because we had to make decisions, ah, do we use GPT Turbo? It's faster, more expensive. We use this model. You know, what, what does speed matter to people? And everyone said the same thing. Like, I, this idea, if it's in a movie, could last forever. You know, it's in my portfolio forever. People may still be watching it in 20 years. I want the best idea. I don't care if it takes two seconds longer to generate it. Um, but yeah, so the case study was great. You know, it wasn't a scientific study. Maybe someone else will do such a thing. Um, but yeah, we're having a lot of people, um, now from Hollywood reach out. They've heard about us. Um, you know, we started our blog in 2021 to, you know, to call back to our earlier point, here's all the things we think will go well and go wrong. You know, deep fake politicians.

Speaker B: Let's talk a little bit about that. Like ethics. Right? So it's a hot conversation here. And I I think the copyright office just recently ruled on this as well too.

Speaker A: Yep.

Speaker B: So what. What do you foresee as AI becomes and it continues at the. At the speed that it's getting integrated into everything, not just Hollywood, from every product, from. From, um, marketing to, you know, filmmaking to. You kind of name it. And I don't think there's an industry that hasn't been affected by it at this point.

Speaker A: No.

Speaker B: What do you see?

Speaker A: Some of the law, uh, medicine, law,

Speaker B: medicine, you name it. So what's your take on the whole ethical challenges here? And where do you. What's your viewpoint on this as well too?

Speaker A: Yeah, I have a slightly different take. And I want to just start by saying, like, these companies, I think OpenAI, um, and stability AI companies like this, they have built the best products I've ever seen in my life. They have the best teams. I look up to them. You know, I wanted to work for them. At one point I was considering applying. Um, but yeah, I mean, no company's perfect. Everything moves fast and there have been missteps. So maybe I'll mention a few. Um, but I do believe that, you know, the. Too long didn't read it. TLDR answer is generative. AI is ethical when used as a tool in the hands of human creators to make more of their best work faster. I don't want that scenario, the strikes mentioned, which is, well, Zaslav's just gonna get his own version of AP of chat, GBT and Sora trained on all of Warner Brothers videos. And, you know, he probably thinks he's super smart and has ideas, so he'll just fire everyone. And just a one person, now billionaire, you know, movie maker. That's not what we want. It was built for people. You know, my brother's my co founder because we wanted a Hollywood filmmaker in our DNA to make the right decisions. We give our users 100 ownership of everything they make in our app. That was our decision in 2021. You know, and the outcome of the writer strike that they fought hard for was writers can use AI, can't be forced to, but if they do, only they own the credit, not OpenAI or what, you know, name whatever app they use. Um, so I think we got ahead of that and that was important. As you mentioned, the copyright office just this week ruled that, um, you know, basically there was always a clause in copyright which said if you even arrange something in a new way, um, you can copyright it. And so the big case study precedent, uh, ruling was, you know, someone said, okay, I wrote a very Good prompt in mid journey. Can I copyright that? No. And that's still true. Can I copy the one image it made for me? No, not really. You didn't really. You didn't put enough work in. But, uh, and someone did this. I believe it was called Daria of the Night. It was this graphic novel. Someone spent, you know, a thousand hours and trying 20 different tools generating tens of thousands of images, possibly to find all the perfect ones to range in a comic book. They wrote all the narrative that was always able to be copywritten. And now it's, um, you know, official in the law that, you know, putting in enough human work, even if you use an AI tool, you can copyright. So I think ethically that makes sense. On the other hand, and I think this is where we're different than other companies. Um, and why I mentioned OpenAI and Stability AI, I very much respect them. They built two of the first image generators. You know, the Internet, the way it works, the World Wide Web is you go to a website and the server serves you the images. So literally you are downloading images and that is fair use. And if they inspire you to make something, um, you can do that. But there was an example where stability AI, they use this lay on 5B data set. It had all, every picture on the Internet, including from Getty Images website Getty Copyright means you have the right to copy if I allow you, and in this case if you pay me. And here's a watermark to make sure you can't until you do. And so they use these watermarked images anyways. And Stability AI, stable diffusion started pumping out images with pseudo Getty watermarks in the side because it thought that's what makes for good images. And we didn't like that. So we said, we're not going to integrate Stability AI's image generator until they change their ways. OpenAI for Dall E3 took a different approach. They were lucky to get $10 billion investment for Microsoft. So they said, we'll just buy every image Getty has, you know, and all those artists, that's a new income stream for them. They made a lot of money. Um, and so we thought that was better. But where I think things can still change even with this copyright ruling is I just believe if you make anything and put it on the Internet and someone will scan it and use it to give other people similar ideas. You know, the strike wanted three 3C's consent. You got to ask me first so not opt out. I don't have to put a no robots tag on all my, um, you know, I want to opt in if I choose to, and I want the second C control. If I decide I don't like your company anymore, I want to be able to take that out of the database. And the third is compensation. You know, you have $10 billion you're making. You know, ChatGPT is the fastest growing, most successful, hugely profitable product in the world. And you can't pay some people for gimp for their images. You already paid Getty, why not us? Um, and so, you know, it made me think when we built our app, I did not want. I made a clause that says you can enter your. Because we were talking to NBC at the time and they were like, if I type in an idea for the next season of Law and Order or whatever it is, are you going to retrain on it? And before I can even film it, someone else is collecting this. So we said we won't train on our, uh, customers work. But it occurred to me that's a bit hypocritical because while we're giving this option to our users, OpenAI and, um, all these other AI companies did not give them the option and we're using them. So the ethical decision is tough. And I write a lot about it on my blog. People can check it out on medium or link from our website on the blog page. And so we tried to propose ideas. So I'll just stand on this. Um, you know, Goldman McKinsey, all these analyst firms are saying, uh, generative AI alone will add. You know, the range was 10 to 20 trillion dollars over the next 10 years to the global economy and GDP. Can we maybe reroute and pipe some of that back to the people whose work was used to enable these things to be built? And people, well, it's impossible. How are you going to pay everyone in the world? And so we're thinking, hey, maybe we stable, uh, diffusion is open source. Uh, it's just their weights. But I think there's some code for the image generation portion. Maybe we can swap out this troublesome data set. And we have a new product. And it's a data set where people can say, you know, I, I've been submitting, I've got thousands of pictures, tens of thousands. I submit them to Getty. You know, I get paid 10 bucks a month. Like, I need a new income stream. I will opt in and say, okay, these are my images. I can prove them. You know, I've submitted copyright registration. Maybe we use a blockchain. And so it's, it's public. And so we can track and everyone can know whose picture's who. And, you know, you can always look up a picture. Oh, no, that's mine. There's a dispute. Okay, interesting.

Speaker B: So you think we're going to get into a world where ownership of, uh, digital assets is going to become much more clearly defined. Maybe, maybe the utility of the blockchain and then also what you created versus what AI created, there might be a. There might be some way of differentiation. Am I hearing that correctly or, or this was AI assisted. This was 100 human.

Speaker A: The idea would be, yeah, we don't want synthetic data because, um, that's my biggest fear is humans will go, okay, copyright office ruled. I'm giving up on photography and I'm never going to do art again. And the AI companies now, what they do is synthetic data. You know, I don't want to drive a car for a, um, million hours to teach it to drive. So we create synthetic, you know, game engine world maps. So if all the human artists in the world gave up and it was only synthetic data training these sets, I think we'd hit sort of a local maximum, to use a machine learning term, and where the arts will go. So we want people to keep creating work. The copyright, the founding fathers said in the Constitution, this is to incentivize people to continue creating. And so if they continue creating images, in this case and photography and paintings, whatever, putting it in our thing, they're allowed to take it out if they want. And then let's say we put, you know, the open source image generator on top and we say, okay, this is now a fair trade, uh, image generator. It's like Starbucks coffee beans. They say, you know, these cost $2 more per bag. But here's the story of the farmer, and they're not mistreated, and you know it. So it's like maybe Hollywood would use the power of the guilds to say, okay, okay, team, we're going to boycott these ones and we're going to use this one. Because everyone who pays to use this, the company gives 20% of their revenue to pay out dividends to every single person on that blockchain contract. And that's possible through something called Solana, which is basically a digital contract. So you don't have to write 10 trillion contracts. It's just done automatically. So this is one example of a product where we are not, no, it's fair use to do AI Copyright ruled in our favor, but we still think it's in the best interest to build this product. Um, Big Tech's not going to do it because it's not going to make them rich. So why not an artist friendly AI company like us?

Speaker B: I love it. So Russell, honestly, it's been incredible hearing your journey and uh, from where you were to the inception of cyber film, AI announced saga and shaping the future of storytelling. As we wrap up, I'd love for you to share one, uh, final piece of advice for creators or filmmakers navigating the rapidly evolving AI driven world. What's the mindset or key takeaway they should embrace as, as they thrive, as, as this technology reshapes the industry. What do you think would enable them to be successful?

Speaker A: Yeah, I think honestly Big Tech and I work in Silicon Valley, it sort of lost the world's trust in recent years. Scandals like FTX and Theranos and you know, the promise of crypto. Some people made money, a lot of people lost money and it didn't really live up to the hype. And now, you know, this is the next one. And when AI, uh, generative AI came out, people were saying, oh, this is just the next scam. And a lot of people refuse to even try it and, or listen to people or, you know, if anything's posted on Reddit or Twitter, you get brigaded by just hate and death threats. I would say take a step back, have a little faith in America, uh, or I'm from Canada originally, dual citizen now, but you know, have a faith in um, maybe at least the western countries. There's some adversarial countries I wouldn't call out by name, but you know, they're going to do whatever it takes. This stuff is not going anywhere. Stop screaming about it and letting it live in your head rent free and just give it a try. Um, and have faith that while some rulings may not go in your favor and someone can make a movie with the help of AI, it will compete with your movie. But art is a market. The best movie should win. No matter what tools are made. You're a human, so step it up, learn from these tools, get better. And like, you know, the least. Atoll was the go champion. He played and lost to Alphago. He said it helped him become a better player. He thinks about the game differently. Um, Hollywood movies are stuck in a rut. Everything I watch on Netflix, uh, I predict lines. I know how it's gonna end. Obviously, you know, the superhero is gonna win the love interest and defeat the villain. And you know, it's the same 15 story archetypes since, since the ancient Greeks, like maybe AI can be the thing that gives all us a kick in the butt and says, stop remaking, you know, the Dances with Wolves into Avatar and, you know, Romeo and Juliet in different formats. And imagine movies like Memento that are just weird and trippy and, you know, Pulp Fiction, that it goes out of order. No one's done that before. Like, help it think, help you break out of the box. Uh, you don't have to tell people you used AI. So, um, until it's, you know, more accepted. Yeah, keep your head down, don't make yourself a target, but try it, see if it helps you. And, um, have faith that the bad stuff, like deep fakes, you can't deep fake your CEO's voice and call the CTO and say, you know, transfer all the money to my bank. It's, it's illegal to robocall during an election with the voice of Joe Biden. Like, um, people are catching up on this stuff. Laws will pass to protect you. Um, so, so, um, I just wouldn't ignore it. And especially, you know, if you have kids and they want to get into filmmaking, you know, restricting them from learning AI would be a big detriment to them. Every college student we talk to says, I use chat GPT 10 times a day. It helps me with my, my resume, my cover letter outline, my essay. You know, don't let it do the work for you. You still need to learn. That's why you're paying the $80,000 a year to go to school. You know, still go to film school, be an actor. Um, but um, use it as a tool where you can. And if you're against AI, you can be proud and content. You don't have to use AI, keep using it. But like, you just maybe don't feel the need to like boycott and attack movies and send death threats because, um, you know, people are just creating movies and doing the best they can. And I think it's going to help more people break into Hollywood instead of people in the past they had to move to la, never made it, never met the right people, weren't good and good looking enough, end up as a waiter, waitress, give up on their dream. I really think this is going to grow the pie, grow jobs, give us all more great movies to watch, and the next generation is going to do it anyway. So, you know, don't be the, um, the puppeteer or model builders, you know, who helped on Star wars. And then when, um, when James Cameron and Spielberg were using CGI for the Abyss and Terminator 2 and Jurassic park, there's a big outrage. And like the Luddites, they were like, you know, we're going to try to ruin you and, and destroy you. And you. Just one you. Even the guilds won't be able to make it happen because people will just go make movies in Canada and the uk, huge film markets. Nigeria, Nollywood, huge film market. Bollywood. China wants to get into it and they are building incredible AI models. Um, so, yeah, just try, try to keep optimistic, have some encouragement and, um, keep making great movies because we believe humans will always have a place in filmmaking.

Speaker B: Perfect. I love that. If people, if the listening audience wants to get a hold of you, what's the best way?

Speaker A: Yeah, go to. If you go to our website, uh, right on saga.com there's links to our, uh, please follow our Twitter. We have a lot of demo videos, um, on YouTube. Uh, all our handles are at, you know, at. Right on Twitter. Or, uh, you can email me at Russell Pyberfilm AI And Russell has two S's and two L's in it.

Speaker B: Perfect. Well, thanks for tuning in to Today's episode of 8 Growth Ventures podcast. If you enjoyed the conversation and want to hear more from visionary leaders like Russell Palmer, be sure to subscribe to the podcast at your favorite platform. Don't forget to leave us a review and share this episode with fellow growth leaders and creators. Stay tuned for more insights and with the instruction of growth, innovation and creativity. See you next time.