WEBVTT 1 00:00:01.087 --> 00:00:03.949 Hi, and welcome to not really an episode. 2 00:00:04.249 --> 00:00:15.417 You know, every once in a while we do these, and, you know, if I'm lucky enough to get a recording of one of the talks or panels that I do, I tend to just give it a little bit of an intro and give you the opportunity to see it. 3 00:00:15.518 --> 00:00:17.299 This is a great panel I did. 4 00:00:17.419 --> 00:00:25.064 A gentleman by the name of Victor Harwood runs a digital conference called Digital Hollywood, and he has just some great people talking on it. 5 00:00:25.084 --> 00:00:26.986 I was really honored to be able to join him. 6 00:00:27.286 --> 00:00:27.986 It's all digital. 7 00:00:28.046 --> 00:00:29.086 It's all totally free. 8 00:00:29.327 --> 00:00:31.287 So these come up periodically throughout the year. 9 00:00:31.547 --> 00:00:34.028 I'll put it in the show notes where you could sign up for the next one. 10 00:00:34.088 --> 00:00:35.008 Get on the mailing list. 11 00:00:35.368 --> 00:00:36.588 Really, really good discussions. 12 00:00:36.808 --> 00:00:40.889 Ours was a great panel that was run by Annie Hanlon of Playbook PLBK. 13 00:00:41.069 --> 00:00:44.790 And basically it was about where AI should be used in the development cycle. 14 00:00:44.810 --> 00:00:51.072 Where it can be used, where it is used, where people are talking about it used, and more often than not where people aren't talking about where it's being used. 15 00:00:51.112 --> 00:01:00.294 But even just the other day, Matt Bellany on The Town said that, you know, showrunners are actually getting annoyed when writers don't use AI now. 16 00:01:00.894 --> 00:01:06.475 Because not being able to crack something and not being able to crack story is no longer an answer. 17 00:01:06.936 --> 00:01:14.117 Because I think that, you know, one of the things that we talk about on this is AI should not be a solve for writing, but it is definitely a solve for writer's block. 18 00:01:14.277 --> 00:01:17.238 And I think that is welcome news to many writers. 19 00:01:17.698 --> 00:01:21.482 Of course, nobody's denigrating anybody from doing it any way they want. 20 00:01:21.582 --> 00:01:23.044 The purpose is the outcome. 21 00:01:23.524 --> 00:01:25.466 Anyway, but we had some great panelists. 22 00:01:25.546 --> 00:01:27.468 Russell Palmer, co-founder of Saga. 23 00:01:27.689 --> 00:01:29.450 Matt Pfeffer, co-founder of Ritual. 24 00:01:29.931 --> 00:01:32.013 Alex Gocke from Largo AI. 25 00:01:32.353 --> 00:01:35.717 And Albert Thompson, who's a digital strategist at Walton Isaacson. 26 00:01:35.897 --> 00:01:36.917 It was a really great chat. 27 00:01:37.077 --> 00:01:41.659 All of these people are people who are talking about where AI fits in the creative process. 28 00:01:42.139 --> 00:01:49.062 One of the things that we're always doing on this podcast is trying to get people to get out of the two extremes in this conversation. 29 00:01:49.902 --> 00:01:57.024 Either AI is going to solve everything, you're going to press one button and you're going to have a movie, or it's the devil and should never be used in anything. 30 00:01:57.404 --> 00:01:58.245 The reality is... 31 00:01:58.985 --> 00:02:06.673 The world needs more content and entertainment than ever before, more design, more writing, more entertainment, more TV, more everything. 32 00:02:07.093 --> 00:02:10.617 We live in an always-on world, and that is a wonderful opportunity. 33 00:02:11.017 --> 00:02:16.022 People are hungry for creative work, and more people are entering the game than ever before. 34 00:02:16.282 --> 00:02:18.905 But, you know, it does mean major shifts for the current world. 35 00:02:19.285 --> 00:02:24.249 you know, the current entertainment companies, creative companies and design companies, publishing companies, everybody. 36 00:02:24.589 --> 00:02:26.971 So yeah, we try to have this conversation. 37 00:02:27.032 --> 00:02:28.633 We try to make sure that it's nuanced. 38 00:02:28.673 --> 00:02:32.036 We try to make sure that it's respectful of artists and their hard work. 39 00:02:32.156 --> 00:02:33.977 So have a listen to this. 40 00:02:34.358 --> 00:02:35.859 Sign up to Digital Hollywood. 41 00:02:35.959 --> 00:02:36.840 It really is great. 42 00:02:36.880 --> 00:02:38.901 Victor Harwood's a great dude. 43 00:02:39.041 --> 00:02:41.103 It's just worth being a part of these conversations. 44 00:02:41.263 --> 00:02:42.845 I was very lucky to be a part of this one. 45 00:02:43.305 --> 00:02:44.046 I hope you like it. 46 00:02:44.346 --> 00:02:45.587 And yeah, let me know what you think. 47 00:02:54.552 --> 00:02:56.213 Well, thanks everyone for joining us today. 48 00:02:56.233 --> 00:03:04.240 Our session is called the Creative Pipeline, AI Ideation, Visualization, Storyboards, and Plot Structure. 49 00:03:05.061 --> 00:03:11.486 And essentially this session will be diving into the start of the creativity pipeline from the spark of an idea. 50 00:03:12.166 --> 00:03:17.131 So the question is no longer whether AI belongs in the creative process, but how it's being used. 51 00:03:17.191 --> 00:03:17.471 Okay. 52 00:03:18.231 --> 00:03:20.372 And I'm joined by an amazing group of panelists today. 53 00:03:20.692 --> 00:03:25.255 And these are practitioners from the creative technology and cultural trend side. 54 00:03:25.295 --> 00:03:30.337 So we really have a fantastic group to dive into this with us. 55 00:03:30.677 --> 00:03:34.579 And quickly, as always, thank you, Victor Harwood, for inviting us. 56 00:03:34.699 --> 00:03:36.860 It's always an honor to be part of Digital Hollywood. 57 00:03:36.900 --> 00:03:37.641 Yes, definitely. 58 00:03:37.681 --> 00:03:40.022 Round of applause for Victor Harwood. 59 00:03:41.146 --> 00:03:44.989 So I'll start with myself before we do a round of intros. 60 00:03:45.129 --> 00:03:53.474 I'm your moderator, Annie Hanlon, and I am co-founder of Playbook PLBK, which is a strategic advisory and consulting firm. 61 00:03:54.255 --> 00:04:00.699 We help studios, technology companies, brands, and creatives navigate the rapidly evolving future. 62 00:04:00.879 --> 00:04:01.900 Rapidly evolving is... 63 00:04:02.500 --> 00:04:11.671 seems so quaint now, right, of content through AI, virtual production, real-time technologies, and next-gen production workflows. 64 00:04:12.552 --> 00:04:16.616 So let me introduce some of our panelists here. 65 00:04:16.696 --> 00:04:20.120 So I'm going to start, let me get back to my notes here. 66 00:04:20.420 --> 00:04:21.702 Matt, I'm actually going to start with you. 67 00:04:22.137 --> 00:04:22.337 Sure. 68 00:04:23.098 --> 00:04:23.698 Hello, everybody. 69 00:04:24.259 --> 00:04:24.999 I'm Matt Pfeffer. 70 00:04:25.019 --> 00:04:27.301 I'm executive producer and co-founder of Ritual. 71 00:04:28.142 --> 00:04:29.263 Ritual is really two things. 72 00:04:29.383 --> 00:04:31.024 Ritual Film is our heartbeat. 73 00:04:31.044 --> 00:04:35.608 It's a premium production company based in Los Angeles that operates in advertising and entertainment space. 74 00:04:36.329 --> 00:04:39.231 It's built on craftsmanship, emotion and the human pulse in every frame. 75 00:04:40.232 --> 00:04:41.693 We have Ritual Labs as well. 76 00:04:42.494 --> 00:04:50.076 which we call a nervous system, our creative technology venture that takes our storytelling muscle and infrastructure for producing campaigns and fuses that with the power of AI. 77 00:04:51.016 --> 00:04:59.858 So we do a lot in the AI space now with Ritual Labs, working with a lot of directors and AI leads for different campaigns with brands and also dabbling in the entertainment space. 78 00:05:00.038 --> 00:05:00.698 It's great to be here. 79 00:05:01.419 --> 00:05:01.819 Excellent. 80 00:05:01.839 --> 00:05:02.259 Thank you. 81 00:05:02.539 --> 00:05:02.899 Alex? 82 00:05:03.675 --> 00:05:05.496 Yeah, thank you for having me, Annie. 83 00:05:05.596 --> 00:05:09.077 I'm Alex Skokie, Vice President of Sales for Largo. 84 00:05:09.837 --> 00:05:11.938 What Largo does, we've been around for eight years. 85 00:05:11.958 --> 00:05:14.879 We're out at EPFL University, the MIT of Switzerland. 86 00:05:15.279 --> 00:05:20.341 We do content insights and audience testing using a variety of both AI and human applications. 87 00:05:20.881 --> 00:05:23.902 So I'm also a screenwriter and film producer, if you will. 88 00:05:24.122 --> 00:05:25.403 And so, you know, we work with... 89 00:05:25.823 --> 00:05:30.464 variety of Hollywood studios to ad agencies to independent filmmakers. 90 00:05:30.864 --> 00:05:33.585 We do audience testing using simulated focus groups. 91 00:05:34.105 --> 00:05:43.147 It was an AI-based platform that's a highly accurate audience that you could test out creative ideas, early stage treatments, all the way to screenplay, all the way to finish film. 92 00:05:43.207 --> 00:05:46.088 So it's more, what does the audience want? 93 00:05:46.128 --> 00:05:50.449 And that's what the tool set and the workflow and how we support our clients in the community. 94 00:05:50.789 --> 00:05:52.329 So thanks for having me, Annie. 95 00:05:52.629 --> 00:05:52.870 Awesome. 96 00:05:52.890 --> 00:05:53.930 Orlando? 97 00:05:53.950 --> 00:05:54.350 Orlando? 98 00:05:55.663 --> 00:05:58.244 I'm Orlando Wood, I'm the founder of Kubrick and Kubrick Labs. 99 00:05:58.424 --> 00:06:12.330 Kubrick, we were the first company to crack automated coverage back at the back end of 2022, when that was a fairly tricky thing to do, when the input token lengths of the APIs was only 2,500 tokens and our standard screenplay is about 70,000 tokens. 100 00:06:13.010 --> 00:06:17.792 And we turned that into a screenplay management system that's in use by over 70 production companies and studios. 101 00:06:18.693 --> 00:06:23.615 But then, you know, I started Kubrick Labs because a lot of those clients were coming to us and saying, we like your product, 102 00:06:24.175 --> 00:06:26.877 But we think that there are other things we could be doing with AI. 103 00:06:26.937 --> 00:06:29.119 And so we've now started Kubrick Labs. 104 00:06:29.159 --> 00:06:31.881 That's now, you know, kind of eclipsed the original company. 105 00:06:31.941 --> 00:06:50.675 And, you know, we build AI tools for development and production and post across the creative pipeline, you know, just to help creative companies use AI on their data and create systems that help them with development and with creative development. 106 00:06:51.556 --> 00:06:51.916 Amazing. 107 00:06:52.797 --> 00:06:53.017 Russell. 108 00:06:53.037 --> 00:06:53.217 Russell. 109 00:06:53.694 --> 00:06:53.974 Great. 110 00:06:54.755 --> 00:06:55.795 Yeah, great to be here. 111 00:06:55.855 --> 00:06:56.716 Such a great panel. 112 00:06:57.816 --> 00:06:58.937 So I'm Russell Palmer. 113 00:06:59.057 --> 00:07:00.738 I'm the co-founder of Saga. 114 00:07:00.838 --> 00:07:05.761 So we started about five years ago and built one of the first AI screenwriting apps. 115 00:07:05.881 --> 00:07:09.142 And then a year later, followed that up with storyboarding. 116 00:07:09.923 --> 00:07:15.206 And now we do sound, music scoring, video, everything. 117 00:07:15.266 --> 00:07:18.187 And we have an editing component, a nonlinear editor in the app. 118 00:07:18.627 --> 00:07:22.970 So, yeah, it's a great sort of end-to-end creative experience from idea to MP4. 119 00:07:24.517 --> 00:07:24.937 Excellent. 120 00:07:25.037 --> 00:07:25.438 Amazing. 121 00:07:25.538 --> 00:07:31.761 Well, as I promised you, audience, we have a very unique and awesome panel today. 122 00:07:32.541 --> 00:07:33.922 So, Russell, I'm going to start with you. 123 00:07:34.682 --> 00:07:43.147 As you said, you've spent years building SAGA specifically for writers to help from ideation, story development, characters, drafts. 124 00:07:43.327 --> 00:07:44.988 We talked a bit about that in our prep call. 125 00:07:45.428 --> 00:07:49.470 Where do you think AI adds the most value before a screenplay even exists? 126 00:07:50.873 --> 00:07:51.353 Yeah. 127 00:07:51.654 --> 00:07:57.479 I mean, there's a couple of places, but I think the most value I've seen in my experience, my brother's my co-founder. 128 00:07:57.499 --> 00:07:58.439 He's a screenwriter. 129 00:07:58.499 --> 00:07:59.780 He works on movie sets. 130 00:08:00.141 --> 00:08:03.924 He's been a PA, a first AD and everything in between. 131 00:08:04.184 --> 00:08:11.891 And, you know, people often say staring at the blank page is hard. 132 00:08:11.911 --> 00:08:13.532 You know, you got that writer's block. 133 00:08:14.513 --> 00:08:14.693 And... 134 00:08:15.130 --> 00:08:17.613 I guess I'd say maybe, I mean, it helps with structuring. 135 00:08:17.633 --> 00:08:20.517 It helps with all the stuff that writers don't want to do. 136 00:08:20.557 --> 00:08:27.706 They don't want to do all the planning and not to generalize everybody, but this has been my experience, but it keeps them going. 137 00:08:28.346 --> 00:08:30.529 And so in an app like Sagar and ChatGPT, 138 00:08:32.255 --> 00:08:33.536 You're writing, you get stuck. 139 00:08:34.116 --> 00:08:35.877 You know, it can just throw out some ideas. 140 00:08:35.957 --> 00:08:40.018 And even what we've seen is sometimes it's not the idea it gives you. 141 00:08:40.198 --> 00:08:42.860 It's what it makes you think of and inspires. 142 00:08:43.520 --> 00:08:47.281 It's almost like it gets all the generic ideas out of the way that you think of. 143 00:08:47.722 --> 00:08:51.763 And then it lets your brain go, okay, I said no to all these. 144 00:08:52.384 --> 00:08:53.304 There's something missing. 145 00:08:54.465 --> 00:08:56.766 And usually it can help you find that missing thing. 146 00:08:57.566 --> 00:08:59.447 And just to close, people say, you know, 147 00:09:00.427 --> 00:09:03.250 It's not, I don't feel like the AI wrote it for me. 148 00:09:03.310 --> 00:09:04.671 I feel like it was my idea. 149 00:09:04.731 --> 00:09:05.892 I just didn't think of it yet. 150 00:09:06.593 --> 00:09:08.215 And so it got me there faster. 151 00:09:08.515 --> 00:09:10.457 It got me there with the quality I want. 152 00:09:11.918 --> 00:09:15.101 And, you know, I'm not losing any writing credits or giving up anything. 153 00:09:15.161 --> 00:09:16.202 It's all my work. 154 00:09:16.322 --> 00:09:17.984 So yeah, I think that's been very helpful. 155 00:09:18.451 --> 00:09:18.651 Yeah. 156 00:09:18.931 --> 00:09:26.498 So you would say to writers who worry that AI is replacing creativity, I'm guessing your response would be it's not. 157 00:09:26.818 --> 00:09:28.039 Do you want to give us a little more? 158 00:09:28.059 --> 00:09:30.621 Yeah, I mean, it's not. 159 00:09:30.681 --> 00:09:32.563 It can make you more creative. 160 00:09:32.803 --> 00:09:33.684 I mean, think about it. 161 00:09:33.744 --> 00:09:34.865 It's a writing partner. 162 00:09:35.706 --> 00:09:40.170 You know, people work in writing rooms or go to coffee shops and, you know, it's a writing partner that's. 163 00:09:41.517 --> 00:09:45.518 For better or for worse, we can maybe talk about the copyright, but it's read every movie script. 164 00:09:45.599 --> 00:09:47.819 It's read the reviews of all those. 165 00:09:47.919 --> 00:09:51.841 It's read books on how to read movie scripts and Wikipedia pages. 166 00:09:52.841 --> 00:09:59.043 And so I always refer back to this great documentary, AlphaGo, by the Google DeepMind team. 167 00:09:59.723 --> 00:10:02.604 And this game's been played for 3,000 years by humans. 168 00:10:03.484 --> 00:10:06.946 And all of a sudden, this AI comes up with this Move 37 that just... 169 00:10:08.306 --> 00:10:09.427 People were laughing at first. 170 00:10:09.467 --> 00:10:10.487 They thought it was the worst move. 171 00:10:10.607 --> 00:10:16.571 And they realized it must be creative to think of a new original move like this. 172 00:10:17.151 --> 00:10:25.575 And the human player, Lisa Dahl, said, I now think I'm a better Go player having learned from this AI. 173 00:10:25.595 --> 00:10:27.837 And so now I'm more creative. 174 00:10:28.357 --> 00:10:30.138 And I think that will happen with writers. 175 00:10:30.158 --> 00:10:30.658 It helps. 176 00:10:31.078 --> 00:10:32.579 You work on a hospital procedural. 177 00:10:32.619 --> 00:10:36.101 It'll help you find some new weird diseases to put in an episode. 178 00:10:36.121 --> 00:10:37.502 So anything, research, anything. 179 00:10:38.303 --> 00:10:47.983 Yeah, it's really amazing too to see just, you know, with ChatGPT and Claude, like how quickly the sort of co-working, you know, the kind of leaning in to all of that. 180 00:10:48.504 --> 00:10:48.845 Yeah. 181 00:10:49.518 --> 00:10:51.059 I think real quick, sorry. 182 00:10:51.079 --> 00:10:51.840 I know you're bad. 183 00:10:51.880 --> 00:10:52.821 I was going to go to you anyway. 184 00:10:52.901 --> 00:10:55.403 I had a thought real quick as Russell was speaking. 185 00:10:56.064 --> 00:10:59.727 You know, creative partner, but I think it's, I think it'd become a sparring partner. 186 00:11:00.027 --> 00:11:00.207 Yeah. 187 00:11:00.247 --> 00:11:00.928 A lot of ways, right? 188 00:11:00.948 --> 00:11:06.052 Where, you know, typical AI defaults to agreements and saying, that's a great idea. 189 00:11:06.072 --> 00:11:07.274 What about this, this, and this? 190 00:11:07.714 --> 00:11:16.702 I think if you can go in certainly with some of the more advanced tools and say, challenge this, challenge this, as we all know, creative field, friction and challenging, but creates a problem. 191 00:11:17.102 --> 00:11:18.383 you know, creative product out of that. 192 00:11:18.423 --> 00:11:18.643 Right. 193 00:11:19.084 --> 00:11:27.471 And so I think, you know, yeah, I think as a sparring partner to really set up those prompts with AI saying challenge X, Y, and Z, this is the end results I want to get to. 194 00:11:27.751 --> 00:11:29.032 Don't agree with me on everything. 195 00:11:29.573 --> 00:11:31.554 Let's get here, generate 20 directions. 196 00:11:31.594 --> 00:11:32.295 Let's remove 18. 197 00:11:32.475 --> 00:11:35.998 And we kind of let that, that last tube be a fight with each other. 198 00:11:36.018 --> 00:11:39.861 Anyways, that, that's how, I think that's such an interesting approach. 199 00:11:39.922 --> 00:11:43.404 I know Russell was getting there with a lot of that was like, yeah, it's a sparring partner in my opinion. 200 00:11:43.545 --> 00:11:43.705 Mm-hmm. 201 00:11:44.145 --> 00:11:52.729 And I think that that's a great point, Matt, because I think that you have the agency to decide how critical it is of you, how it's coming in as a partner. 202 00:11:53.009 --> 00:11:57.010 Just like a showrunner has the decision of, I'm bringing this person in because they challenged me. 203 00:11:57.291 --> 00:11:59.572 I'm bringing this person in because they're better at plot than I am. 204 00:11:59.932 --> 00:12:03.173 You have that agency to decide where you're coming in at. 205 00:12:03.193 --> 00:12:06.875 And I think the best writers I know that are using it are saying... 206 00:12:07.895 --> 00:12:09.977 It's not here to solve writing. 207 00:12:10.197 --> 00:12:14.021 Writing is one of the joys of human life and experience. 208 00:12:14.041 --> 00:12:15.863 It's a great source of pride and self-worth. 209 00:12:16.243 --> 00:12:19.786 But AI shouldn't be solving writing, but it can solve writer's block. 210 00:12:20.227 --> 00:12:24.090 And I think that that's the real benefit of it. 211 00:12:24.110 --> 00:12:28.935 And I think where writers are, you know, I mean, because any writer will tell you, I feel incredibly empowered when I write. 212 00:12:28.955 --> 00:12:33.579 I feel the worst about myself I ever feel when I'm experiencing writer's block. 213 00:12:35.362 --> 00:12:36.884 Yeah, no, I think that's a good point. 214 00:12:36.944 --> 00:12:43.991 And I think, Matt, you know, you started to touch on something because you're sitting between brands and clients and creative and production teams. 215 00:12:44.011 --> 00:12:45.953 Like, you're in the middle of all of that. 216 00:12:46.053 --> 00:12:52.019 So can you give us some examples of how you're using AI as that sparring partner? 217 00:12:52.280 --> 00:12:52.460 Like, 218 00:12:53.070 --> 00:12:57.874 I mean, yeah, I mean, you look like we're using across the lifecycle of content to do. 219 00:12:57.894 --> 00:12:58.094 Right. 220 00:12:58.114 --> 00:13:02.477 You look at, you know, creative development. 221 00:13:02.517 --> 00:13:04.859 I guess let's let's take an example from a brand perspective. 222 00:13:04.879 --> 00:13:10.844 We get an RFP, we get some sort of boards from an agency, whatever that stakeholder is. 223 00:13:11.124 --> 00:13:16.348 And we've got to respond creatively to that, you know, whether it's we're going straight in with the director, right. 224 00:13:16.768 --> 00:13:20.391 or we're just doing it internally, yeah, that sparring partner comes in handy, right? 225 00:13:20.411 --> 00:13:27.435 And I think what we're able to do, I think AI collapses the distance between an idea and something you can react to. 226 00:13:28.736 --> 00:13:37.562 Traditional method is like we have an idea, we put it on paper, we pitch the idea verbally, and then it's up to the imagination to a degree of the stakeholder. 227 00:13:37.602 --> 00:13:40.644 You're pitching it to whether it's a buyer in entertainment or whether it's a... 228 00:13:41.504 --> 00:13:43.946 CD and the brand space or an agency. 229 00:13:44.506 --> 00:13:53.653 And for what you can do now is, you know, at earlier stage, actually like articulate and then actually showcase, you know, what you're trying to pitch in a lot of ways. 230 00:13:53.733 --> 00:14:01.919 And so, you know, before any money is really spent, we can put an idea in front of the client, you know, that then showcases a little bit more and mitigates their risk. 231 00:14:01.979 --> 00:14:04.261 And so, yeah, we're using it across the full life cycle. 232 00:14:04.281 --> 00:14:10.385 I mean, for example, we just, we have a fully 100% Gen AI campaign outright with Lenovo that, 233 00:14:10.645 --> 00:14:12.166 That's that's running in the World Cup right now. 234 00:14:13.368 --> 00:14:15.970 And that was a triple bid situation for us. 235 00:14:16.471 --> 00:14:19.513 And we went with it with our AI lead and director. 236 00:14:20.334 --> 00:14:21.315 They gave us storyboards. 237 00:14:21.335 --> 00:14:22.676 The in-house team gave us storyboards. 238 00:14:22.716 --> 00:14:28.402 We took those storyboards and we brought them to life in the pitch before we even won the job. 239 00:14:29.102 --> 00:14:32.423 And at that point, they could see the fidelity of work that we could produce. 240 00:14:33.143 --> 00:14:36.703 And that honestly was kind of what set us over to win the job. 241 00:14:36.743 --> 00:14:51.266 And so, you know, I think it's, again, and then you take it to the downstream phase of content being distribution, as in market, whether it's performance marketing you're doing, whatever it is, like we can now iterate, you know, as it's in market as well. 242 00:14:51.306 --> 00:14:53.907 So it's really, we're using it across that full life cycle. 243 00:14:53.987 --> 00:14:54.987 And it's an interesting time. 244 00:14:55.707 --> 00:15:10.679 I think the distribution side, it still requires a bit finessing because it revolves a dance, a very delicate dance with media buyers and the producers to see what are the signals we're looking for and how are we reacting to those signals and whatnot. 245 00:15:10.719 --> 00:15:12.300 But it's the capabilities there. 246 00:15:12.320 --> 00:15:14.382 And those are the things that we're testing on right now. 247 00:15:14.949 --> 00:15:15.450 That's awesome. 248 00:15:15.510 --> 00:15:18.032 And I don't know who I stole this from, but I heard it somewhere. 249 00:15:18.172 --> 00:15:21.055 And I love this sentence where they say it's not replacing creativity. 250 00:15:21.095 --> 00:15:28.022 It's compressing the distance between imagination and execution, which I think is kind of an interesting, you know, I should put it on a T-shirt. 251 00:15:28.042 --> 00:15:29.224 So... 252 00:15:31.846 --> 00:15:42.196 Alex, why don't you step in a bit to this as well, because you're coming at it from the analytics side a bit. 253 00:15:42.696 --> 00:15:48.342 Do you want to tell us a bit more about how you're working at that sort of ideation phase? 254 00:15:48.918 --> 00:15:49.619 Yeah, absolutely. 255 00:15:50.239 --> 00:15:50.779 Happy to. 256 00:15:50.819 --> 00:15:52.821 Yeah, I'll speak to the evolution of Largo. 257 00:15:52.841 --> 00:16:00.226 So we started a content insights platform where we train a model, you know, 400,000 plus scripts to recognize content patterns. 258 00:16:00.246 --> 00:16:08.932 So we started with a lot of financial predictions, highly accurate to genre, to casting, to audience emotional appeal just at the script stage. 259 00:16:09.432 --> 00:16:14.596 And then we evolved a few years ago into audience testing using a simulated focus group methodology, right? 260 00:16:14.716 --> 00:16:20.098 we saw what was happening in the film industry as well as in the ad side of things going, Hey, what does the audience want? 261 00:16:20.118 --> 00:16:27.800 At the end of the day, everything we're talking about here is we want to resonate your core message, whether it's a film or ad with your target audience. 262 00:16:28.300 --> 00:16:40.504 And so we have a proprietary system then that we've built where you can test a screenplay to a feature film, meaning an actual video in front of an audience and get 90% accurate feedback. 263 00:16:40.944 --> 00:16:49.454 So the benefit then the whole conversation you're just having around compressing the timeframe Annie is before you ever bring a film to market, you can now test the script. 264 00:16:49.794 --> 00:16:50.935 Hey, what does the audience like? 265 00:16:50.975 --> 00:16:51.756 What are they not like? 266 00:16:52.277 --> 00:16:55.741 And from a creative standpoint, like I'm a screenwriter and I actively use the tools. 267 00:16:56.201 --> 00:16:58.222 Then I could also ask Taylor question. 268 00:16:58.342 --> 00:17:00.422 Well, how do you feel about the father son relationship? 269 00:17:00.923 --> 00:17:02.103 What about the plot structure? 270 00:17:02.163 --> 00:17:02.883 What about this? 271 00:17:03.343 --> 00:17:13.627 And our model based on AB comparisons with billion dollar brands where we've confirmed that it's 90% accurate, you now have that data to then stay firmer on your creative. 272 00:17:14.127 --> 00:17:23.815 So this is an instance where it actually allows the creatives to go, hey, I'm going to use AI to validate that this is going to resonate with my core audience. 273 00:17:24.175 --> 00:17:27.578 So it empowers the creatives to take those bold stances. 274 00:17:27.998 --> 00:17:38.526 And we've tested a variety of things that arguably are outside the genre realm, like Parasite, Everything Everywhere All at Once, films that are outliers from a pure data and story structure standpoint. 275 00:17:38.546 --> 00:17:39.207 And so it's a very good thing. 276 00:17:39.387 --> 00:17:44.632 And our data, our audience confirms that those are outlier successful type projects. 277 00:17:45.073 --> 00:17:47.995 So just to speak to, that's a common question I often hear as well. 278 00:17:48.035 --> 00:17:49.937 Does it just give you what you want to hear? 279 00:17:50.137 --> 00:17:51.539 Is it just market validation? 280 00:17:51.619 --> 00:17:57.144 And in fact, our tools are tailored in a way to recognize those key hits. 281 00:17:57.224 --> 00:17:57.525 Yeah. 282 00:17:59.306 --> 00:17:59.766 Yeah, that's great. 283 00:17:59.786 --> 00:18:06.370 And I think especially you, Alex and Orlando, now that you guys are working so much on the tech side of it, but you come from creative. 284 00:18:06.570 --> 00:18:10.612 And I think that is a big bridge that's necessary. 285 00:18:10.652 --> 00:18:16.415 I think that's when the tools and when the analytics work the best is exactly how you just said you spoke to Alex, because you know, as a screenwriter. 286 00:18:17.135 --> 00:18:20.076 how to talk in story language and in creative language. 287 00:18:20.096 --> 00:18:22.957 And I think that's, that's where it's usually the most successful I see. 288 00:18:22.997 --> 00:18:26.299 I mean, we know when we're talking to somebody, if they've been on that side, right. 289 00:18:26.339 --> 00:18:30.380 Or if they've been, you know, only 100% on the, on the tech side. 290 00:18:31.601 --> 00:18:33.902 So Matt, I want to bring it back to you for a minute. 291 00:18:33.922 --> 00:18:36.663 Cause you had mentioned how, how, 292 00:18:37.683 --> 00:18:46.467 brands and clients in general often are able to make probably more informed decisions because you can actually show them what it is that you're going to be creating for them. 293 00:18:47.367 --> 00:18:52.810 Are there particular steps that you take to make them feel more confident about the creative? 294 00:18:52.890 --> 00:19:00.893 Because is there also that piece of like, well, wait, there's an authenticity piece here that I'm going to shy away from in terms of the AI versus the traditional? 295 00:19:01.173 --> 00:19:02.514 How do you bridge that gap? 296 00:19:04.847 --> 00:19:19.177 Yeah, I mean, I think the way to bridge it is, you know, yeah, essentially up front, you preface it with the rules of engagement and how we're leveraging AI, right? 297 00:19:20.178 --> 00:19:31.705 You know, our directors, our creative leads who have a very distinct point of view at the wheel and under the hood is the intelligence, is the AI piece of it, right? 298 00:19:32.885 --> 00:19:41.230 And, you know, as we say, like, you know, we leverage AI as a creative partner, not necessarily as a cheaper, faster shortcut to everything, right? 299 00:19:41.270 --> 00:19:42.111 It's horsepower to us. 300 00:19:43.252 --> 00:19:48.096 And I'll see whoever uses it, you know, that keeps a human governance kind of at the wheel of it. 301 00:19:48.136 --> 00:19:53.020 So for us, yeah, you just really preface it with that, where it's like, this is how we're using it. 302 00:19:53.080 --> 00:19:55.703 This is this is this is how who's governing it. 303 00:19:56.223 --> 00:20:03.450 This is, you know, and then the optionality you get, the creative leverage that you get throughout the process from start to finish. 304 00:20:04.090 --> 00:20:06.212 is expanded way more than it ever has before. 305 00:20:06.652 --> 00:20:09.434 Typical production, all that risk is mitigated. 306 00:20:09.834 --> 00:20:17.220 All that risk, that financial risk from the client, from the buyer, whoever's financing it, is put into those days on set, on a physical set, right? 307 00:20:17.660 --> 00:20:25.706 And across that whole cycle from creative development, pre-production to production to post-production to delivery, it could all fall off the rails. 308 00:20:26.626 --> 00:20:31.170 And so for us, this technology and the way we're incorporating it 309 00:20:31.690 --> 00:20:38.452 really gives the client that creative leverage for the buyer, whoever that stakeholder is that's financing it, more creative leverage than they've had before. 310 00:20:38.552 --> 00:20:38.692 Right. 311 00:20:38.712 --> 00:20:49.135 And so I think when you get to that point of like you're going to make better decisions, fewer decisions and earlier decisions, though, that's that really resonates with with the clients, you know, in a lot of ways. 312 00:20:49.175 --> 00:20:51.516 And so that I think is how we're using it. 313 00:20:51.876 --> 00:20:55.717 And then everything we create always goes through a path of human governance. 314 00:20:56.117 --> 00:21:00.201 It's never let's just prompt something, you know, get a result, video gen, whatever it is. 315 00:21:00.541 --> 00:21:01.142 Let's send it out. 316 00:21:01.162 --> 00:21:02.303 But it's like it's testing. 317 00:21:02.343 --> 00:21:09.050 Like we've been saying from before, like the sparring, the going through the filters of like, no, let's try this again. 318 00:21:09.090 --> 00:21:09.751 Let's try this again. 319 00:21:10.331 --> 00:21:11.492 But it needs that human governance. 320 00:21:11.512 --> 00:21:15.676 There's different touch points of that process to really hone it in. 321 00:21:15.737 --> 00:21:15.937 Right. 322 00:21:15.957 --> 00:21:18.699 And like, as you see, AI slop necessarily isn't just. 323 00:21:20.101 --> 00:21:22.323 the poor fidelity of the generations, right? 324 00:21:22.944 --> 00:21:31.333 Really, in my opinion, the AI slop is generated by a lack of taste in how it's put together from an editorial perspective, from sound design, from all that. 325 00:21:31.733 --> 00:21:38.080 It's just like you need that storyteller who has the knack for telling great stories to still put that together. 326 00:21:38.520 --> 00:21:42.064 And so I think for us, again, it's just that reassurance of like there's a human governance at the wheel. 327 00:21:42.706 --> 00:21:42.866 Yeah. 328 00:21:43.527 --> 00:21:45.408 No, and I think that's interesting because I agree. 329 00:21:45.489 --> 00:21:48.311 I think you can always tell when a good storyteller is using AI. 330 00:21:48.751 --> 00:21:49.392 It's just different. 331 00:21:49.592 --> 00:21:50.253 It doesn't matter. 332 00:21:50.273 --> 00:22:02.043 It could be the cleanest, most beautiful, real-looking piece, but if it's not a good story, it doesn't really matter what technology you use. 333 00:22:02.584 --> 00:22:06.647 Are you seeing too, because I've personally seen, especially over the past six months, 334 00:22:07.468 --> 00:22:11.111 people kind of coming to us wanting to be like, okay, we really want to pull in AI. 335 00:22:11.131 --> 00:22:14.034 Are you seeing clients kind of coming to you getting ahead of that question? 336 00:22:14.094 --> 00:22:20.840 And whereas maybe a year ago, more of an uphill convincing conversation as to why this would be successful? 337 00:22:20.860 --> 00:22:26.845 Yeah, in the advertising space, it is, I mean, I wouldn't say diametrically opposite from the entertainment space, but it certainly is. 338 00:22:28.167 --> 00:22:38.414 they're leaning in and the C-suites, you know, are all pushing pressure down to their marketers, to the media buying teams and saying, how are we using AI, how are we using AI? 339 00:22:38.454 --> 00:22:44.259 And so, see right now, you know, AI has created infinite possibility, but also has created infinite noise. 340 00:22:44.319 --> 00:22:50.804 And so, you know, there's a translation layer that's needed as you talk marketers, because they're feeling that squeeze and that pressure from the top. 341 00:22:51.244 --> 00:22:54.387 And now they're looking for somebody who's got some way to translate. 342 00:22:54.787 --> 00:22:55.508 How can we take, 343 00:22:56.068 --> 00:23:07.414 what we're trying to do from a marketing initiative, from what our CMO and the C-suite have kind of laid on us and execute that, but also do it with trying to, you know, leverage this technology in some way. 344 00:23:07.434 --> 00:23:13.438 And so I think for us, it's like, there's an overcorrection where everyone wants to use AI. 345 00:23:14.158 --> 00:23:15.099 That's not the answer. 346 00:23:15.579 --> 00:23:20.784 It comes down to discernment, circumstantial to what's best served for that creative. 347 00:23:21.144 --> 00:23:22.045 Sometimes it's a hybrid. 348 00:23:22.345 --> 00:23:24.307 Sometimes it's not even anything AI. 349 00:23:24.768 --> 00:23:29.452 And so I think it's just kind of that calmness to the client or the stakeholder of saying, yeah. 350 00:23:29.832 --> 00:23:30.592 We'll figure this through. 351 00:23:30.612 --> 00:23:31.953 Like, let's talk through the creative. 352 00:23:31.993 --> 00:23:34.013 Let's figure out how we can kind of look. 353 00:23:34.113 --> 00:23:40.695 I mean, there's there's ways of checking the box of saying this is used by but with AI without not necessarily creating it fully AI. 354 00:23:40.775 --> 00:23:40.935 Right. 355 00:23:40.955 --> 00:23:45.477 And so but it's we're hearing it this year compared to last year are more. 356 00:23:46.559 --> 00:23:46.779 Yeah. 357 00:23:46.899 --> 00:23:51.003 I mean, it's funny because a few of us on this panel have advertising experience. 358 00:23:51.043 --> 00:23:53.445 And I think that advertising is under very unique pressures. 359 00:23:53.885 --> 00:23:57.929 They need to compress as much as possible the time between ideation and execution. 360 00:23:58.369 --> 00:24:04.235 You know, we were talking with Sergio Perez at Omnicom Production the other day on my podcast. 361 00:24:04.895 --> 00:24:13.082 And he was talking about how, you know, if an idea is ideated 18 months ago and finally it's executed now, 362 00:24:14.023 --> 00:24:16.385 There's a creative dynamic to that as well. 363 00:24:16.445 --> 00:24:19.788 I mean, fundamentally, it was a different world 18 months ago. 364 00:24:19.848 --> 00:24:21.389 Joe Biden was still president. 365 00:24:22.009 --> 00:24:27.153 Think about all that we've learned about who we are as a country in that 18 month period of time. 366 00:24:27.514 --> 00:24:29.735 So compressing that timeline. 367 00:24:29.775 --> 00:24:32.437 Yes, obviously, there's cost optimization and things like that. 368 00:24:32.457 --> 00:24:34.339 But there's also a creative element to it. 369 00:24:34.739 --> 00:24:55.615 And I think that that's why we're seeing advertisers rush into this space a lot faster, because like I said, it's a sort of perfect storm of cost optimization alongside analytics like Largo, you know, and then as well, actual creative benefit, you know, to compressing that timeline from ideation to execution. 370 00:24:56.678 --> 00:24:57.699 No, I think that's a really good point. 371 00:24:57.719 --> 00:25:05.724 And I feel we saw this a bit, even with VR, I feel because I'm one of the folks on this panel who comes from advertising and commercials as well. 372 00:25:06.224 --> 00:25:12.968 And it is that piece of, you know, VR was funded by marketing departments at movie studios and by brands. 373 00:25:13.028 --> 00:25:19.733 And, you know, it wasn't always going to be a perfect execution, but it sort of felt like a little bit of a safer space to play with technology. 374 00:25:19.773 --> 00:25:21.974 And it really helped push things forward. 375 00:25:21.994 --> 00:25:23.655 So I feel like there's always a bit of that. 376 00:25:23.755 --> 00:25:26.017 They're on a slightly different adoption curve 377 00:25:26.907 --> 00:25:28.447 than other folks are. 378 00:25:28.548 --> 00:25:39.691 And it is really interesting to see creatively, you know, especially at this sort of spark of an idea space that we're talking about today, how it has given them so much agency to be able to do that and to play with things. 379 00:25:39.731 --> 00:25:40.451 And it's really great. 380 00:25:41.131 --> 00:25:47.313 I think that's a win with all of this, whether the end product is exactly perfect or not, almost doesn't matter. 381 00:25:47.333 --> 00:25:48.594 It's just the fact that they're stepping in. 382 00:25:49.514 --> 00:25:50.555 Correct. 383 00:25:50.575 --> 00:25:51.095 Because it is. 384 00:25:51.155 --> 00:25:52.196 I mean, look at all the updates. 385 00:25:52.216 --> 00:26:01.342 There's constantly a tech update or something that's a new version of something that used to be, you know, with Epic, wasn't it like every, maybe every year? 386 00:26:01.502 --> 00:26:04.644 And now it's literally like every few weeks we're seeing an update of something. 387 00:26:04.664 --> 00:26:07.426 So things look quote unquote old really quickly. 388 00:26:07.446 --> 00:26:11.969 To your point, Orlando, of like that 18 months is, you know, kind of a non-starter. 389 00:26:11.989 --> 00:26:12.090 Yeah. 390 00:26:12.990 --> 00:26:13.751 in a lot of ways. 391 00:26:15.733 --> 00:26:21.219 So Russell, with you, I mean, you know, we're talking about how does AI help to write scripts, right? 392 00:26:21.259 --> 00:26:24.322 But what is it actually good at? 393 00:26:24.342 --> 00:26:26.384 Because you use it for several different things. 394 00:26:26.504 --> 00:26:27.765 Is it ideas? 395 00:26:27.845 --> 00:26:30.408 Is it more the story structure, arcs? 396 00:26:30.448 --> 00:26:33.451 What are some of the things you're seeing where it's really adding value? 397 00:26:33.471 --> 00:26:33.571 Yeah. 398 00:26:35.470 --> 00:26:41.112 Well, you know, 18 months ago, I'd say it wasn't good so much at idea. 399 00:26:41.152 --> 00:26:44.274 I was at AI on the lot and Paul Schrader gave the keynote. 400 00:26:44.554 --> 00:26:44.774 Yeah. 401 00:26:45.174 --> 00:26:51.957 And he asked, for those who haven't seen it, you know, ChatGPT gave me an idea for a Paul Schrader type movie and plot. 402 00:26:51.997 --> 00:26:54.798 And he read it and it was good, unquestionably good. 403 00:26:54.858 --> 00:26:58.820 And he said, he might not like it, but this is good. 404 00:26:58.860 --> 00:26:59.580 I'm Paul Schrader. 405 00:26:59.681 --> 00:27:01.361 I'm the one who... 406 00:27:01.381 --> 00:27:02.842 So it is getting... 407 00:27:03.949 --> 00:27:05.210 better at ideas. 408 00:27:05.390 --> 00:27:10.514 It's really good at, you know, we have like what we call our series matrix. 409 00:27:10.554 --> 00:27:11.594 You can do a TV show. 410 00:27:12.155 --> 00:27:15.537 So you can plan all your seasons first and then episodes and seasons. 411 00:27:16.338 --> 00:27:25.264 And it will keep track of all the plot lines, all the storylines, all the different character arcs, and make sure that they all close. 412 00:27:26.204 --> 00:27:27.645 You can ask for script coverage. 413 00:27:27.665 --> 00:27:32.268 So, you know, it'll tell you like, there's some sayings in cinema like, yeah, 414 00:27:33.229 --> 00:27:34.890 Every scene should advance the plot. 415 00:27:34.930 --> 00:27:36.851 Every line of dialogue should. 416 00:27:37.331 --> 00:27:41.373 And so it can tell you what's sort of extraneous in there that maybe you don't need. 417 00:27:42.354 --> 00:27:45.556 Yeah, I love the sparring partner analogy. 418 00:27:45.716 --> 00:27:46.556 I think that's great. 419 00:27:46.596 --> 00:27:56.382 When I see we do like usability studies and people will, you know, they'll ask for ideas and they'll say, oh, this isn't, I don't like this, I don't like this. 420 00:27:56.422 --> 00:27:58.883 But when they do a script coverage report, 421 00:28:00.523 --> 00:28:05.148 It's up to them to say, well, I don't agree with you here, here, and here, but you're right. 422 00:28:05.528 --> 00:28:06.809 You know, you're right about this one. 423 00:28:07.289 --> 00:28:08.811 And now I can go change that. 424 00:28:09.872 --> 00:28:12.814 Again, it's like it feels like I'm making that decision. 425 00:28:13.035 --> 00:28:15.637 I'm, you know, the one who's fixing it. 426 00:28:16.478 --> 00:28:19.180 And so they still feel like it's their work. 427 00:28:19.300 --> 00:28:20.281 They're very involved. 428 00:28:20.782 --> 00:28:24.044 But yeah, the stuff that humans, you know, our memory is not perfect. 429 00:28:24.265 --> 00:28:26.206 AI's memory is getting much better now. 430 00:28:27.527 --> 00:28:32.329 And so, you know, ask, has this ever been done before? 431 00:28:32.489 --> 00:28:33.649 Is this an original idea? 432 00:28:33.709 --> 00:28:36.990 Oh, no, it's been done on The Simpsons in South Park or that sort of thing. 433 00:28:37.030 --> 00:28:37.190 Right. 434 00:28:37.590 --> 00:28:38.370 Or research. 435 00:28:38.390 --> 00:28:43.612 I was talking earlier, like people who work on police shows or hospital procedurals. 436 00:28:44.743 --> 00:28:48.745 They need like a new weird crime or a weird disease every week. 437 00:28:48.785 --> 00:28:50.106 And it helps with the research. 438 00:28:50.566 --> 00:28:57.310 So you don't have to be, you know, an expert in a certain field to be able to tell, you know, part of your story about it. 439 00:28:57.430 --> 00:28:59.452 So, yeah, I mean, it's just getting better. 440 00:28:59.472 --> 00:29:08.497 The, the video is, I would say, photorealistic now, like pretty much outside the uncanny Valley, the voice performances are getting better. 441 00:29:08.517 --> 00:29:09.498 Um, yeah, 442 00:29:10.458 --> 00:29:13.701 So Google's Lyria 3 in our app. 443 00:29:13.741 --> 00:29:18.185 We can say, like, look at the script, look at the storyboard, you know, score this scene. 444 00:29:18.205 --> 00:29:21.148 I kind of like Interstellar, but I don't like the, you know, negative prompt. 445 00:29:21.168 --> 00:29:22.229 I don't like the organ. 446 00:29:22.269 --> 00:29:25.692 And you can tell it things you like, and it will help you. 447 00:29:25.752 --> 00:29:28.214 So it's helping with just about everything. 448 00:29:28.275 --> 00:29:31.197 And then self-distribution is sort of the final frontier. 449 00:29:31.217 --> 00:29:31.317 Yeah. 450 00:29:32.250 --> 00:29:37.652 People can just put their stuff on YouTube, cut a trailer, a sizzle reel, put it on TikTok. 451 00:29:38.393 --> 00:29:40.734 And so you're solving discovery. 452 00:29:40.754 --> 00:29:43.975 I mean, that's not AI, but technology is really helping people. 453 00:29:45.116 --> 00:29:47.437 Scorsese says everyone has a story to tell. 454 00:29:48.077 --> 00:29:51.459 Soon everyone will be able to make a movie out of it. 455 00:29:51.639 --> 00:29:54.040 So I think AI is really going to help from that front. 456 00:29:54.738 --> 00:29:58.199 But I like how you talk about like the human piece of it. 457 00:29:58.259 --> 00:30:04.441 So can you, can you sort of reiterate a few of those, like where you think the human should stay in control? 458 00:30:04.461 --> 00:30:06.062 You just mentioned a few, but just to. 459 00:30:06.102 --> 00:30:06.322 Yeah. 460 00:30:06.342 --> 00:30:09.203 I mean, in my experience, they always come with an idea. 461 00:30:09.403 --> 00:30:16.005 In fact, every writer I talk to has, they say I have seven or eight ideas tumbling around in my head for the last, you know, 10 years. 462 00:30:16.525 --> 00:30:17.425 I don't need ideas. 463 00:30:17.965 --> 00:30:19.446 I need you to help me land them. 464 00:30:20.226 --> 00:30:25.287 I need an act three twist, or I need to close this storyline, or I need to make this character more interesting. 465 00:30:26.827 --> 00:30:41.470 So no one is doing what I think the fears of 2023 and the strikes were, which is just some executive who only got there because they're good at climbing the corporate ladder and they have powerful friends or whatever. 466 00:30:41.490 --> 00:30:49.272 They'll type, you know, come up with a movie that makes me rich, put money in bank account, you know, press this button. 467 00:30:50.200 --> 00:30:51.481 They have a story too. 468 00:30:51.882 --> 00:30:52.102 Yeah. 469 00:30:52.122 --> 00:30:53.883 Yeah. 470 00:30:53.923 --> 00:30:54.864 You've got a story, right? 471 00:30:55.525 --> 00:30:56.366 But yeah. 472 00:30:56.426 --> 00:30:59.368 And so it's kind of like, um, it's not writing the script for them. 473 00:30:59.428 --> 00:31:00.729 It's not doing the work. 474 00:31:01.870 --> 00:31:06.214 Um, um, the human is still driving it, you know, they're the pilot. 475 00:31:06.274 --> 00:31:14.161 They're in charge of the safety of everyone on board, the co-pilot or the autopilot can sort of take over for a little bit, but, um, 476 00:31:15.382 --> 00:31:20.025 We believe in a world where humans will always be behind the stories, even acting. 477 00:31:20.085 --> 00:31:24.347 Like I just don't see Tilly Norwood, you know, she's coming out in a feature film soon. 478 00:31:24.367 --> 00:31:26.949 I don't see people enjoying it. 479 00:31:27.089 --> 00:31:30.671 Just like, I don't think people would watch robots play a soccer world cup. 480 00:31:30.771 --> 00:31:40.476 I mean, maybe, I don't know, maybe in Japan, they were like that, but like, you like seeing humans and hearing their infallibility and their, you know, they're not perfect and they've overcome things. 481 00:31:40.636 --> 00:31:41.157 And, um, 482 00:31:42.078 --> 00:31:45.001 Yeah, if you heard an AI, I mean, I'll close on this. 483 00:31:45.081 --> 00:31:47.303 Like, there was an AI rapper. 484 00:31:47.944 --> 00:31:51.408 You know, they were trying to have a musician, FM Mecca or something like that. 485 00:31:51.468 --> 00:31:55.393 And, you know, just I love hip hop, but hearing the lyrics like... 486 00:31:56.423 --> 00:31:59.825 You know, I'm the toughest guy in my neighborhood. 487 00:31:59.885 --> 00:32:02.126 And, you know, I went to prison and no one messed with me. 488 00:32:02.366 --> 00:32:03.067 It's like, no, you didn't. 489 00:32:03.087 --> 00:32:03.767 You're a robot. 490 00:32:04.467 --> 00:32:07.289 So that stuff, I don't know if it will connect. 491 00:32:08.189 --> 00:32:17.174 So to your sort of follow-up question, I feel like, yeah, the human still needs to be the one behind it and driving it and making the final decisions. 492 00:32:17.234 --> 00:32:18.475 But the AI can help. 493 00:32:19.015 --> 00:32:22.778 when you get stuck and it can maybe give you some new ideas. 494 00:32:24.140 --> 00:32:27.483 And but it's still up to you to sort of put it all together. 495 00:32:27.503 --> 00:32:31.847 And that's why we think people will get copyright because arrangement of what, 496 00:32:32.708 --> 00:32:43.354 okay, AI helped with these storyboards, and maybe it did this score over here, and then, you know, it wrote a bit of the script, but I had to rewrite this scene. 497 00:32:43.794 --> 00:32:51.378 You know, by arranging things and editing them and putting it all together in one movie, you will get copyright for it, we believe. 498 00:32:51.758 --> 00:32:55.320 And, yeah, because it's not going to be push a button, get a movie. 499 00:32:55.340 --> 00:32:57.201 There will be that human voice behind it. 500 00:32:57.561 --> 00:32:58.201 Yeah. 501 00:32:58.401 --> 00:33:00.022 I think one of the things that... Oh, sorry, go ahead. 502 00:33:00.202 --> 00:33:00.703 No, no, please. 503 00:33:00.723 --> 00:33:00.783 No. 504 00:33:01.543 --> 00:33:12.992 Well, I think that one of the things that Russell said is, you know, in his robots playing soccer analogy is that, you know, at the end of the day, Garry Kasparov was beaten by Deep Blue in 1996 or 1998. 505 00:33:13.372 --> 00:33:18.076 It didn't usher in a whole era of people watching robots play each other in chess. 506 00:33:18.516 --> 00:33:19.497 And chess has never been bigger. 507 00:33:19.537 --> 00:33:22.419 I mean, ever since Queen's Gambit, I mean, they're playing it on Twitch. 508 00:33:22.479 --> 00:33:25.221 It's absolutely, you know, Magnus Carlsen is a big deal. 509 00:33:25.722 --> 00:33:28.064 You know, it's chess has never been bigger. 510 00:33:28.284 --> 00:33:29.044 And, you know... 511 00:33:29.865 --> 00:33:33.011 we all know that all of these people could be beat by a computer. 512 00:33:33.172 --> 00:33:39.263 It doesn't change the fact that we want to watch humans struggle and try and, and engage with one another. 513 00:33:39.283 --> 00:33:40.005 Mm-hmm. 514 00:33:40.909 --> 00:33:42.650 No, I think that's 100% true. 515 00:33:43.050 --> 00:33:44.110 I completely agree with that. 516 00:33:44.811 --> 00:33:46.711 And Alex, I mean, how does that work? 517 00:33:47.031 --> 00:33:51.333 Tell us a bit more about where the human's in the loop in the work. 518 00:33:51.433 --> 00:33:53.774 Yeah, I love this conversation. 519 00:33:53.794 --> 00:33:57.335 I mean, I would just speak to, you know, learning curve, right? 520 00:33:57.635 --> 00:34:01.817 Once you get to a certain stage where you're consciously competent in your craft, you're 521 00:34:02.017 --> 00:34:06.160 then you have a level of self-awareness around, hey, what is lackluster, right? 522 00:34:06.700 --> 00:34:15.166 And so like how I've personally used our tools and applications, we have a simulated focus group, which means an audience panel that's 90% accurate with real people. 523 00:34:15.826 --> 00:34:20.269 So I've used it like 35 times as I've evolved a coming-of-age fantasy kid script, right? 524 00:34:20.810 --> 00:34:24.272 And so I always encourage clients, attack your weaknesses. 525 00:34:24.672 --> 00:34:26.295 You know what scenes are lackluster. 526 00:34:26.315 --> 00:34:27.496 They're sort of dragging. 527 00:34:27.657 --> 00:34:31.522 And then you could ask, hey, how do you feel about this ex-husband, ex-wife relationship? 528 00:34:32.063 --> 00:34:33.645 Oh, it's not really grounded in reality. 529 00:34:33.665 --> 00:34:35.047 It's the 60% of an audience. 530 00:34:35.468 --> 00:34:38.533 Well, another 40% don't like this one dynamic. 531 00:34:38.893 --> 00:34:39.554 Oh, well, let me. 532 00:34:39.674 --> 00:34:59.252 revisit that let me do a rewrite and then retest it and so it becomes an iterative tool to validate so that's one two like i recently did a table read on it and i looked at the people around very grateful for them to be there but there's largely older people who are reading it whereas like poor audience is more the pixar 6 to 11 range right yeah 533 00:34:59.452 --> 00:35:03.054 Now I could test within a six to 11 range an actual audience. 534 00:35:03.134 --> 00:35:12.137 And so I could do different scenario plan, different core audiences within the tool set, get guidance and then figure out, well, what is going to resonate with different types of audiences? 535 00:35:12.657 --> 00:35:14.398 And then it's all in a closed loop environment. 536 00:35:14.418 --> 00:35:19.340 So I could go back to them over and over and I could even interact with the simulated audiences. 537 00:35:19.360 --> 00:35:23.562 We have a chatbot feature to go deep dive on specific aspects. 538 00:35:23.982 --> 00:35:25.603 So I find it helpful from a... 539 00:35:26.203 --> 00:35:28.787 creative standpoint, from a market validation standpoint. 540 00:35:28.807 --> 00:35:42.024 So if I do take this data and then I bring it out to investors or partners or financiers, et cetera, then all of a sudden it goes, hey, it's on par with these other projects because there's a benchmarking against other, you know, the onwards and inside out of the world. 541 00:35:42.044 --> 00:35:42.385 Yeah. 542 00:35:42.925 --> 00:35:51.086 So also helps our clients and helps me, frankly, both to try to position their projects for funding that much better. 543 00:35:51.547 --> 00:35:54.547 So it provides like proof points across the lifecycle. 544 00:35:54.967 --> 00:35:58.488 And then also fast forward 12 months, I have a rough cut film. 545 00:35:58.568 --> 00:36:03.609 I can now test it in an audience in Germany and go, hey, this is how it might do there. 546 00:36:04.049 --> 00:36:05.109 Or I could test a trailer. 547 00:36:05.289 --> 00:36:11.650 So literally across the entire production lifecycle, the tool sets are helpful from that standpoint. 548 00:36:12.410 --> 00:36:18.971 But it all goes back to the human is going to lead the charge in terms of their creative vision. 549 00:36:19.852 --> 00:36:24.013 And where I find it is then I don't get stuck, as Orlando mentioned earlier. 550 00:36:24.213 --> 00:36:25.653 I don't get stuck as easily. 551 00:36:25.673 --> 00:36:30.194 And I don't even hit writer's block because I constantly have rolling notes of I need to do this. 552 00:36:30.354 --> 00:36:31.134 Let me test this. 553 00:36:31.334 --> 00:36:32.174 Let me test this. 554 00:36:32.574 --> 00:36:35.235 And I just do piecemeal iterative improvements. 555 00:36:35.255 --> 00:36:35.635 Yeah. 556 00:36:35.815 --> 00:36:38.576 to improving my overall work and then I validate it. 557 00:36:38.696 --> 00:36:42.959 So it becomes like this process where overall it's improved trend line over time. 558 00:36:43.699 --> 00:36:43.939 So. 559 00:36:45.040 --> 00:36:59.347 Well, it's interesting because it sounds like you're also saying that having that data is sort of opening opportunities as opposed to what I think everyone was worried about where everything was going to become very homogenized and it was only going to be like this one thing that would sort of make it through, you know, that validation. 560 00:36:59.957 --> 00:37:00.378 Correct. 561 00:37:00.458 --> 00:37:05.443 Well, and our tools are tailored so they're benchmarked to real people that have filled in behavioral surveys. 562 00:37:05.523 --> 00:37:09.428 And what that allows for is it will tell you you are not the next Tarantino. 563 00:37:09.588 --> 00:37:11.530 It will say it's a 3.2 out of 10. 564 00:37:11.811 --> 00:37:13.232 Like it will not blow smoke. 565 00:37:13.532 --> 00:37:18.999 And so then that way it grounds you for better or for worse because we've put through purely AI generated scripts. 566 00:37:19.019 --> 00:37:19.599 We've put through... 567 00:37:20.160 --> 00:37:34.108 crap intentionally just to like just to just to validate hey where does it land and it'll give you the honest truth right and so that that speaks volumes then to when you do have something that's quality on your hands that go hey this is where it stands so 568 00:37:34.725 --> 00:37:36.466 Alex, I'm going to send you my brother-in-law. 569 00:37:36.486 --> 00:37:37.447 He keeps sending me scripts. 570 00:37:37.547 --> 00:37:38.727 You can give him the real truth, man. 571 00:37:38.747 --> 00:37:40.448 I got you. 572 00:37:40.488 --> 00:37:42.390 I'm getting a script a week from this guy. 573 00:37:42.410 --> 00:37:43.090 I got an idea. 574 00:37:43.150 --> 00:37:45.211 I'm like, all right, I'll read. 575 00:37:45.251 --> 00:37:46.872 He's like a plumber in Boston. 576 00:37:47.112 --> 00:37:48.193 He's a plumber in Boston. 577 00:37:48.213 --> 00:37:49.394 He's got these great script ideas. 578 00:37:49.654 --> 00:37:50.994 So I'm sending him to you, Alex. 579 00:37:51.035 --> 00:37:51.315 Okay. 580 00:37:51.375 --> 00:37:51.835 Thanks, man. 581 00:37:53.353 --> 00:37:54.154 That's awesome. 582 00:37:54.214 --> 00:37:55.495 That's awesome. 583 00:37:55.695 --> 00:38:02.762 Matt, so your brand produces professional projects, like you were saying, like with Lenovo and for FIFA World Cup. 584 00:38:03.422 --> 00:38:10.969 So as we're talking about sort of this authenticity piece, brands spend years building up that, you know, authentic connection with their audience. 585 00:38:11.009 --> 00:38:18.696 So when using AI, like how do you keep that sort of human point of view and connection so that you're not... 586 00:38:19.593 --> 00:38:24.235 you know, that you're keeping that audience engaged as opposed to turning that audience off. 587 00:38:25.135 --> 00:38:25.495 Are there? 588 00:38:25.515 --> 00:38:30.237 I think there's a principle, you know, across whether it's advertising or entertainment, right? 589 00:38:31.238 --> 00:38:41.382 What you're doing when you produce something that's taken from a script, somebody's brain, you put it onto a screen, you're trying to suspend the notion of disbelief, you know, with that person who's watching it, right? 590 00:38:41.422 --> 00:38:41.702 So, yeah. 591 00:38:42.562 --> 00:38:58.382 I think to a degree, AI, whether it's live action, animation, whatever medium you're producing it through, it's just a similar medium or tool than you would apply for a live action production or an animation production in some ways, right? 592 00:38:58.402 --> 00:39:00.485 Where it requires just, again, that human production. 593 00:39:01.666 --> 00:39:02.106 point of view. 594 00:39:02.927 --> 00:39:06.070 To me, audiences don't detect tools, they detect intention. 595 00:39:06.090 --> 00:39:12.056 And for us, one of the maxims, anything we do from a creative perspective is we want to make you feel something. 596 00:39:12.517 --> 00:39:15.019 Happiness, reflection, motivation, whatever that is. 597 00:39:15.099 --> 00:39:15.339 Right. 598 00:39:15.600 --> 00:39:20.144 To us, that's what we want to evoke when you watch a piece of something that we produce. 599 00:39:21.025 --> 00:39:25.629 And really, the only way to do that is to come from a point of view. 600 00:39:26.330 --> 00:39:29.191 and a skill set to be able to produce something. 601 00:39:29.211 --> 00:39:30.512 A point of view is one thing, right? 602 00:39:30.532 --> 00:39:38.776 But then to be able to actually tactfully put something together editorially with sound design, the ebbs and flows of emotion, you know, and actually move somebody with sound. 603 00:39:39.016 --> 00:39:41.818 Like it is a work of art in a lot of ways, right? 604 00:39:41.858 --> 00:39:42.118 So... 605 00:39:42.898 --> 00:39:44.938 for us, it really comes down to that. 606 00:39:44.998 --> 00:39:49.539 Lenovo for us, it was a soccer spot with technology, right? 607 00:39:49.559 --> 00:40:08.303 And so, you know, the feeling that we wanted to evoke in that was energy and belief that, like, there's, like, this Lenovo collaboration with FIFA and that they're powering all facets of the World Cup from Vietnam, where some go better than others, as we know in America, to other facets of the World Cup. 608 00:40:08.323 --> 00:40:11.524 And so for us, it was more about energy and, like, excitement, right? 609 00:40:12.844 --> 00:40:17.466 Again, you need that not to be a broken record. 610 00:40:17.546 --> 00:40:19.106 You just need that human point of view. 611 00:40:19.586 --> 00:40:20.586 Our bar at Lounge is simple. 612 00:40:20.767 --> 00:40:21.947 Zero slop. 613 00:40:22.067 --> 00:40:24.988 If the human point of view isn't in the frame, the tool doesn't matter. 614 00:40:25.508 --> 00:40:25.908 So that's it. 615 00:40:26.708 --> 00:40:27.849 Yeah. 616 00:40:28.629 --> 00:40:29.629 Albert, how are you doing? 617 00:40:30.169 --> 00:40:33.050 I'm going to pull you right into this conversation. 618 00:40:33.070 --> 00:40:34.491 Hi, Albert. 619 00:40:34.971 --> 00:40:35.351 Thanks, guys. 620 00:40:35.922 --> 00:40:40.407 So as we've been talking about, like creative work has traditionally started with the blank page. 621 00:40:41.428 --> 00:40:54.323 So given the expertise that you have in like cultural trends, can you tell, give us your thoughts, I guess, on, you know, has AI fundamentally changed that moment or has it just changed how we approach that blank page moment? 622 00:40:54.857 --> 00:40:58.798 You know, look, I think it's a fantastic conversation that will be debated probably to the end of time. 623 00:40:58.818 --> 00:41:00.959 I mean, look, some of this comes down to engagement. 624 00:41:01.359 --> 00:41:01.859 Does it attract? 625 00:41:01.899 --> 00:41:02.359 Did it work? 626 00:41:02.379 --> 00:41:03.020 Was it good enough? 627 00:41:03.060 --> 00:41:03.800 Do people watch it? 628 00:41:04.280 --> 00:41:05.160 So what? 629 00:41:05.300 --> 00:41:06.241 Is it monetizable? 630 00:41:06.281 --> 00:41:06.581 Yes. 631 00:41:06.781 --> 00:41:07.761 And do brands like it? 632 00:41:07.801 --> 00:41:10.742 But does it create an ignition point for them to sell their stuff? 633 00:41:11.042 --> 00:41:11.802 Fantastic. 634 00:41:12.143 --> 00:41:14.083 I think right now everything is very much experimental. 635 00:41:14.103 --> 00:41:16.504 I mean, look, we have three different marketers on the planet right now. 636 00:41:16.824 --> 00:41:20.267 You know, when you look at it from the agency side, from the brand side, you have Gen Z, Gen X. 637 00:41:20.968 --> 00:41:29.154 And then obviously you have the, you know, millennial, you know, Gen X very much built the frame of marketing, was very much about the big idea, landed, rinse, repeat it. 638 00:41:29.194 --> 00:41:30.195 That's what Hollywood has done. 639 00:41:30.676 --> 00:41:33.178 That's why we're rebooting and rebooting and rebooting and rebooting. 640 00:41:33.218 --> 00:41:34.219 Star Wars will never die. 641 00:41:34.739 --> 00:41:40.481 You know, when you start to look at, you know, the millennial type, they're very much about, well, what's the storyline? 642 00:41:40.841 --> 00:41:42.142 We like origin stories. 643 00:41:42.202 --> 00:41:42.942 Give a good story. 644 00:41:43.302 --> 00:41:44.983 That's a very different lens and very different filter. 645 00:41:45.383 --> 00:41:47.183 And then Gen Z is like, where's the community? 646 00:41:47.624 --> 00:41:50.285 If the community is good, we're part of the community, we'll buy whatever you build. 647 00:41:50.665 --> 00:41:56.027 So when you start to look at the advent of AI into the production lens, it looks, you have to look at those three facets. 648 00:41:56.127 --> 00:41:59.308 What you probably need to do is manage to the center. 649 00:42:00.188 --> 00:42:08.153 have a good community behind the idea of AI assisted and generated content for people who like Japanese anime, so they don't even need to see a real actor anyway. 650 00:42:08.513 --> 00:42:10.294 That's a very different audience. 651 00:42:10.354 --> 00:42:12.376 To people like, look, just have good stories. 652 00:42:12.416 --> 00:42:21.041 If it's a good story, I can follow because let's just be clear, you know, some of the action scenes and chips, the old California Highway Patrol weren't all that good and AI is better than that stuff we saw. 653 00:42:21.081 --> 00:42:22.622 Look, it's better than some CGI stuff. 654 00:42:23.042 --> 00:42:29.565 And then the idea, the big ideas that land to continue to rent and repeat and monetize until the frame shifts again. 655 00:42:29.805 --> 00:42:30.865 I think that's the biggest thing. 656 00:42:30.985 --> 00:42:33.186 The other thing that AI produces is a ton of metadata. 657 00:42:33.686 --> 00:42:36.267 People talk about data is really the little data, which is metadata. 658 00:42:36.287 --> 00:42:42.390 So when you look at Grace Snow to think analytics, these are the people that have data behind the smart TV consoles from your LG Vizios. 659 00:42:42.410 --> 00:42:43.030 They know who they are. 660 00:42:43.050 --> 00:42:43.450 We don't. 661 00:42:43.910 --> 00:42:44.611 So we know Nielsen. 662 00:42:44.631 --> 00:42:45.411 We don't know Grace Snow. 663 00:42:45.751 --> 00:42:50.453 The metadata is the insights is going to help provide what you should be making. 664 00:42:50.913 --> 00:42:57.138 So I think part of the stick, when you start looking at AI, we see a lot of front end magic or shiny keys is what people say. 665 00:42:57.518 --> 00:43:02.362 It's the metadata that really should be steering a lot of decisioning on the back end to produce what audiences want. 666 00:43:02.762 --> 00:43:06.966 And as I say, as the head of NVIDIA said, AI is going to dismiss your excuses. 667 00:43:07.226 --> 00:43:08.147 There will be no excuse for it. 668 00:43:08.167 --> 00:43:09.047 Well, it didn't work well. 669 00:43:09.248 --> 00:43:12.050 You should have tested with AI against synthetic audiences first. 670 00:43:12.530 --> 00:43:13.992 or audiences that mirror real audiences. 671 00:43:14.032 --> 00:43:14.812 Why didn't you do that? 672 00:43:15.173 --> 00:43:16.414 Forget the traditional testing. 673 00:43:16.434 --> 00:43:25.282 So I think AI is going to mitigate and de-risk a lot of projects because you can prototype, you can do a proxy, you can see if it lands, put it in the community, see if they hate it. 674 00:43:25.302 --> 00:43:26.643 If they hate it, you don't make it, they love it. 675 00:43:27.044 --> 00:43:28.665 You consider ushering it to the front. 676 00:43:28.705 --> 00:43:35.592 So I think those are the things, it's the signals that AI is going to create around audiences to really determine, are you doing what you're doing? 677 00:43:35.672 --> 00:43:37.494 Hollywood had the benefit of always dictating. 678 00:43:37.994 --> 00:43:38.554 This will work. 679 00:43:38.654 --> 00:43:39.334 We're going to make it. 680 00:43:39.514 --> 00:43:42.175 And since we own the theaters, we're going to make you show up there and watch it. 681 00:43:42.795 --> 00:43:49.297 Now they have to be a lot more intentional about what could be a real flock because certain generations won't even go to the theater to see it whatsoever. 682 00:43:49.657 --> 00:43:52.758 And I think that's a great moment in time that has never existed before. 683 00:43:53.138 --> 00:43:54.538 The question is, are people going to harness it? 684 00:43:54.558 --> 00:43:59.940 So a studio has not to be like, hey, we use AI for particular scene captures. 685 00:43:59.960 --> 00:44:01.180 We didn't have to permit that. 686 00:44:01.220 --> 00:44:02.000 We just rendered it. 687 00:44:02.020 --> 00:44:05.181 But what they're going to say is, look, we're looking at metadata that's telling us. 688 00:44:05.976 --> 00:44:13.202 where engagement is, what we can monetize, what we can tie shopping and make it shoppable content to. 689 00:44:13.582 --> 00:44:15.924 So the brand says, okay, you bulletproof this thing. 690 00:44:15.984 --> 00:44:17.605 That's what you're supposed to do. 691 00:44:18.046 --> 00:44:20.668 Don't just sell me on what you love and think I love it too. 692 00:44:21.188 --> 00:44:24.171 Sell me on what mathematically should not fail. 693 00:44:24.600 --> 00:44:25.861 Yeah, no, I think that's interesting. 694 00:44:25.881 --> 00:44:29.866 And I think it is that view of it's not just the one product. 695 00:44:30.006 --> 00:44:32.128 It's the it's the long tail of that product. 696 00:44:32.169 --> 00:44:35.552 Like you said, it's it's merchandising, it's shopping, it's all of these other things. 697 00:44:35.672 --> 00:44:36.994 It's digital derivatives. 698 00:44:37.034 --> 00:44:40.518 I'll say derivatives, not in a bad way, but in just a another opportunity. 699 00:44:40.738 --> 00:44:40.958 Right. 700 00:44:42.400 --> 00:44:50.551 So, Orlando, kind of coming off that, you know, we talk a lot about image generation, but really what you guys are doing is building production workflows. 701 00:44:51.192 --> 00:44:56.218 So what's that impact on like the traditional phases like that? 702 00:44:56.278 --> 00:44:58.101 You know, how are you seeing that? 703 00:44:59.392 --> 00:45:08.239 Yeah, I mean, I think that like everybody realizes that, you know, development is the longest and costliest part of production because it's the riskiest. 704 00:45:08.599 --> 00:45:09.700 You don't know what's going to hit. 705 00:45:09.720 --> 00:45:11.301 You don't know what's going to move into production. 706 00:45:11.681 --> 00:45:16.385 You know, in history, only 1% of all the screenplays that have been written have ever been made. 707 00:45:17.385 --> 00:45:24.050 You know, that means there's huge amounts of already written wonderful human creativity sitting around desperate to be made. 708 00:45:24.110 --> 00:45:29.193 Like, you know, there are scripts that Lord and Miller made that nobody's doing. 709 00:45:29.233 --> 00:45:32.455 And like, you know, they're great at this, evidently. 710 00:45:33.315 --> 00:45:43.682 You know, so I think that that's, you know, understanding that these tools can help with that process to make it cheaper and accelerate the path to production. 711 00:45:44.083 --> 00:45:58.472 But that's not the problem that we're seeing at the moment is every studio seems to be taking the same attitude, which is we can use these tools to speed up the development process, speed up the work process, you know, back office stuff. 712 00:45:58.933 --> 00:45:59.293 None of them. 713 00:45:59.313 --> 00:46:01.955 I mean, we see a lot of people, you know, 714 00:46:02.395 --> 00:46:16.225 up in arms about AI use, but none of the studios really output any content that is, that is AI, you know, that is, is full of AI or what we call slop now, you know, and so public perception is the thing really holding everything back. 715 00:46:16.265 --> 00:46:22.829 You know, I mean, we saw the, the A24 deal with Google and people are freaking out about it without even realizing the deal. 716 00:46:22.869 --> 00:46:23.690 Like we work with A24, they 717 00:46:26.612 --> 00:46:28.874 their data incredibly seriously. 718 00:46:29.295 --> 00:46:34.460 And, you know, they have said they're not, you know, using their data to train on the Google models. 719 00:46:34.540 --> 00:46:36.122 And I take them at their word. 720 00:46:36.522 --> 00:46:37.723 Nobody's even reading that. 721 00:46:37.763 --> 00:46:39.445 They're so up in arms about it. 722 00:46:39.465 --> 00:46:47.092 And so it's this sort of storm in a teacup that's not really related to anything physical, you know, but, you know, and, and, you know, 723 00:46:47.833 --> 00:46:54.378 Studios are kind of saying things like, everybody says they don't wanna use AI, but everybody's coming back from the bathroom with really good ideas. 724 00:46:54.938 --> 00:47:02.023 And I think that's kind of the big problem at the moment is that film companies know they need to make things faster and cheaper to keep up. 725 00:47:02.423 --> 00:47:08.207 It's not that they're doing it because AI suddenly makes everything cheaper, they're doing it because the current model is stressed. 726 00:47:08.607 --> 00:47:14.751 And social media has been eating everybody's lunch because it's changing the dynamics of attention. 727 00:47:15.211 --> 00:47:24.657 And so I think that the thing really holding back further development right now for at least the big corporations and studios is the public perception. 728 00:47:25.118 --> 00:47:31.302 On the development side, it feels like there's a lot of really rich engagement and they're building workflow tools. 729 00:47:31.322 --> 00:47:32.802 They're building optimization tools. 730 00:47:33.223 --> 00:47:38.026 And that is, you know, making parts of it cheaper and I think will result in more... 731 00:47:38.406 --> 00:47:39.547 ideas being made. 732 00:47:40.507 --> 00:47:45.871 I would jump into that because look, when you look at it, every new technology created a new lens, YouTube created a new lens. 733 00:47:45.891 --> 00:47:47.652 Like, oh my God, this Android content looks like crap. 734 00:47:47.672 --> 00:47:48.272 It's all blurry. 735 00:47:48.573 --> 00:47:49.253 And then we love it. 736 00:47:49.273 --> 00:47:50.694 The world's like, it's just authentic. 737 00:47:50.774 --> 00:47:51.254 It's just raw. 738 00:47:51.594 --> 00:47:56.177 And then TikTok comes and it creates a new filter in terms of let's act like a fool, but let's be funny, entertaining. 739 00:47:56.478 --> 00:47:58.559 But now we get people to transact because it makes us do it. 740 00:47:58.999 --> 00:48:01.180 Every new technology enters and creates a new lens. 741 00:48:01.300 --> 00:48:02.121 AI is creating a new one. 742 00:48:02.141 --> 00:48:05.282 The first thing we do is hate on it until we're all effectively doing it. 743 00:48:05.322 --> 00:48:07.563 We'll keep hating it until we've all effectively started doing it. 744 00:48:07.603 --> 00:48:08.103 Like, oh, you know what? 745 00:48:08.123 --> 00:48:08.644 It's really good. 746 00:48:08.964 --> 00:48:11.025 So, I mean, this is no different than anything else we've seen. 747 00:48:11.365 --> 00:48:17.348 I think what in the back end we have to do is power through that the consumer is the only one guaranteed to win. 748 00:48:17.688 --> 00:48:18.668 They're still winning. 749 00:48:18.688 --> 00:48:19.829 They're still going to win. 750 00:48:20.289 --> 00:48:22.250 And AI is just going to be part of the winning formula. 751 00:48:22.915 --> 00:48:26.586 Albert, you just had the last word because we are out of time. 752 00:48:27.107 --> 00:48:29.314 Unfortunately, thank you all so much. 753 00:48:29.354 --> 00:48:34.328 I have so many more questions I didn't even get to, but I really appreciate this great conversation. 754 00:48:34.348 --> 00:48:35.331 Awesome. 755 00:48:35.853 --> 00:48:36.474 Thanks, everyone. 756 00:48:37.016 --> 00:48:37.537 Thanks so much. 757 00:48:37.558 --> 00:48:38.199 Thank you. 758 00:48:38.621 --> 00:48:38.701 Bye. 759 00:48:38.721 --> 00:48:39.022 Thank you.