AI in commercial content production

00:00:00: You will make AI expensive and slow again

00:00:03: if you press it into old processes and structures.

00:00:06: This is the commercial content podcast. What marketing decision makers need to know about the future of content. In this first episode, Norbert talks with Max Penk about AI in commercial content production. Max studied philosophy and computer science in Mainz and advertising filmproduction in Ludwigsburg.

00:00:24: He worked as a commercial film producer and talent scout in several production companies for many years before he became agency David and Martin's creative innovation director.

00:00:34: He recently joined the AI studio Tektite.

00:00:37: Ladies and gentlemen, let's dive right in. Here's Max and your host,

00:00:41: Norbert.

00:00:42: So welcome, Max. Good to have you here in my new podcast.

00:00:46: Hello, Norbert. Thank you for the invitation. I'm looking forward to speak with you.

00:00:50: Thank you very much. SoI you're my first guest. So this is my first podcast actually. So I mean I've been guest in some but this is my first as a host. So I had a short introduction in the intro. So people kind of know where you are, who you are, and to get started and to warm things up a little bit, I have a little format called Seven Questions, which I would like to ask you.

00:01:12: So that we can, besides your career indications I gave, get a little impression of who you are. So first one is, which media had the greatest influence on you as a child? Television, games, books, and probably I have to add later on internet or social media. But as you're not my age, but quite, I would ask for television, books or games.

00:01:38: Yeah. So, you were asking as a child, and there's one clear answer that is television. Because I was a television child. In our household, I think there were 6 or 7 TVs back then. So, the television was not just a medium you turned on in the evening.

00:01:55: It was one TV running constantly all the time. It was more like a source of light at home. When you get home, you see the television already running. This has to do with, both of my parents working in television. So, my mother in front of the camera as a as a host, for English lessons, for, for, what was it... Südwestfunk.

00:02:20: SWF was it called back then, not Südwestrundfunk

00:02:24: Before the merger.

00:02:25: And, my father working as a producer behind the camera. So television was a it was a big, big topic. And I think influenced me a lot.

00:02:33: Okay. Good to know. So we had one television, and I think we were always the last to get stuff like VHS recorders or CD players, but we had a television. Okay. So what was your first, conscious encounter with advertising?

00:02:47: My conscious encounter, I think. Yeah, advertising was, everywhere. Also back in the television. And it became more and more when private TV started in Germany. So when, when RTL came up and stuff like this and I think back then the advertising was quite amazing also for children, because it was something which had a repeating pattern, in songs and slogans.

00:03:15: And I think this is something when you see it a lot, you try to imitate it. You try to sing the song in advertising. You try to say the slogan somewhere where you see it. And I think television advertising or television commercials were the ones like, yeah, who was stuck in my mind the first before even out of home advertising or radio advertising or something like this.

00:03:40: And I think that the spots back then there were remarkable. There were some classics which like, undefeated, really good stuff.

00:03:47: Yeah. Good. You say that because then it's not only my impression, because I also remember some campaigns from that days, which kind of stuck a little more in my head than the recent ones. So let's talk about today. Where do you currently prefer to spend your time? Is it the movies, the console, or social?

00:04:05: Yeah, I would say, social, so movies. I try to go to the cinema at least like once a month, but, Yeah, it's it's something I really have to focus on more, because every time I'm there, I'm really happy I'm there because it's like a room where you focus only on this one thing and it's really, really good.

00:04:26: And all the films I've seen, like they were chosen very wisely, that were always great the last times I was in the cinema. So I have to go more to the cinema again. I can, but like if I'm consuming media, it's mostly on my mobile or on my computer. So, it's I would not say only socials.

00:04:45: It depends on how you define socials, but that's all. Yeah. Moving images being it's, YouTube, Twitch, whatever channels, Instagram. Yeah.

00:04:56: Okay. Talking about channels. So if you were on an island and you could only take one app or channel, which one would it be?

00:05:04: One channel. I think I would take one from the public channels because this is something like when you watch it, you always feel like, well, I have to watch this more. This is so great. And but there's a lot of stuff which is, not really obvious or not really advertised as much as in, in the socials, or you don't fall in this hole very easily.

00:05:26: So you if you find the right content, let's say something on the mediathek, I don't know, ARTE, WDR, whatever. Yeah. It's like you always find some gems there. And also the digital offers from, from ZDF, which are really, really good you know. So, you really feel this is a good production where, where people really take care about what they are saying and how they are saying it.

00:05:54: And I think if I would go to an island, probably the channel out of the mediathek.

00:06:02: But yeah, I just yeah. Because actually my, my go to app channel is for the so like kind of my only television program I watch, but. Okay. Good. So how do people outside the industry usually misunderstand your job.

00:06:19: I think they don't understand it at all. It's not even a misunderstanding. It's like, what are you doing? But it's a question. Yeah. I think it's, It's hard to explain actually, for someone not working in the business, how much stuff is happening between technology and creativity at the moment. And this is something like there's no like, fixed job description for this.

00:06:45: You know, I grew up in the business. Yeah. I learned to be a producer at the Filmakademie, so I was more like, doing organizational stuff and putting a team together and working on script and everything. But then, you know, I was focusing more on the creative part and in advertising more working on director's interpretations, choosing the right director, choosing the right DP, putting together the right moods so that the creative visiual vision of films.

00:07:15: And then I moved more into concept and more into script. So I think I had like a little journey, from where I started to where I am now. But it was all in the end, it was all technology which led me to where I am now, like I started out, studying computer science. So I was always interested in technology stuff, and I wanted to go deeper.

00:07:40: And I think now it's the time where I can combine technology and creative stuff, working in media because, yeah, there's a lot of, stuff happening and changing at the moment and industry, and then a lot of, I would say stuff is coming together which was not connected before. And this is

00:08:02: what I'm working on every day and trying, yeah, to connect the dots basically.

00:08:07: Yeah, good you say that because actually, I think that your career or you call it...

00:08:15: Studies, because you're one of the few guys I know who can, like, get the first bird's eye view on stuff because you come from philosophy, tech and advertising and film, and I think there's a lot of stuff coming together at the moment. So, very interesting. So if you were no longer allowed to work in advertising or content, what would you do?

00:08:41: That's a good question. I don't know, I didn't learn anything else, so I, you know, I think in the end, storytelling is, is my is my passion, you know, and it doesn't matter if it's happening on television, on socials or with AI or whatever. It's like, telling stories and exploring stories, talking to people, I'm interested in people and human stories.

00:09:09: And I think I would go more into that route being an, I don't know, journalist, interviewer, writing books or something like this. But, yeah, I always want to know. Yeah, the deeper meaning of stuff. And I think I want to dig deeper, whatever topic that is.

00:09:25: Okay, good. Yeah. So. And what do you do to clear your head?

00:09:32: I have a, I have a big passions since 25 years already. I jump out of airplanes, and this is like one of my big hobbies. And it helps me to clear my mind and to focus solely on one thing. And. Yeah, just flying in the sky with my body and not thinking about anything else. This is something which is really fulfilling and resetting my mind.

00:09:54: Which is kind of amazing, because I knew you would answer that because I tried it once and you connected me to the base, with the pilots at the jumpers. And it was so I did it once and I think that's okay. So, you know, it's like running a marathon, I mean I run a half, but it's like I don't have to.

00:10:12: So that's why I really, I'm amazed by that. You really do it, like, more times on the day. Because after the first time, I was like, okay, good. But that definitely helps to, disconnect.

00:10:25: The community you meet with, people, you do travels, and it's like a lot more than just jumping out of airplane and getting a kick out of it, it's really training stuff, and it's really a big topic.

00:10:37: So thanks. Let's jump in. The topic today is AI in commercial content production. So, we kind of touched it, where you come from and, where we are. So let's go deeper into the status quo. What do you think? Where are we today in media and advertising? What's the what's the situation regarding the history and what's changed in the last like.

00:11:06: Yeah, I you cannot even say years anymore. Have to say months, but. Okay. Where are we now?

00:11:12: I think it's probably the biggest revolution in media and advertising happening at the moment still. And it's an ongoing process. I mean, we said it out about a lot of things happening. We said it about digitalization, when it moved from analog to digital in all the disciplines being it illustration being it filmmaking, being it audiovisual productions.

00:11:40: But I think now, we are at a state where structures and processes are changing, and this means a lot of the stuff which worked before, in a way it was established is not working anymore. And this is where we have to reajust, as an industry, as a business. And I think, the people still have the old processes and structures on their mind, and they are just trying to adapt those structures to new technologies, new toolsets.

00:12:15: But I think it's not enough. So I think at the moment we're still in the old system with some fancy new tools, but that doesn't work anymore.

00:12:24: If I read a headline once somewhere "it feels like 2000 with fancier slide decks".

00:12:35: Yeah, sometimes. Yeah. And I think it's not something you can solve easily, like say yeah, we just restructure and doing it like this. It's a process and we are just in the middle of it. And it's sometimes really hurtful. And I think sometimes it's really fulfilling. But there's a lot of stuff happening for sure.

00:12:56: So I mean, yeah, I see it the same, because, I mean, when you see where we come from, and there was like, especially in TV commercial production, there's like a system which is running and it still is. And I think it's very hard for the players to run this system. I always compare it with like the ZDF when Netflix came or general television when American like drama auteur series came, or Mercedes-Benz working on electric vehicles until Tesla or the Toyota Prius came.

00:13:32: You know, they did somehow, but not really. My impression I'm not from the industry. So what do you think? Because at the moment we see, like, big companies around us struggling. There's some mergers, there's some closings, so, what do you think is the main reason why the industry adapts slow? From my point of view.

00:13:56: It's a big question. And, I think, you know, as I said before, it's not AI, it's just one of the reasons, but it's not the reason why everything is changing at the moment. I think a lot of stuff worked out before in a way, that it was established and everyone was used to it.

00:14:17: Being on client side, on agency side or production side, there were a lot of processes which seemed to work. But now, I think there were some multiple reasons to it why it needed to change. For sure. It's the poly-crisis, as we call it. Like the wars which are going on in the world.

00:14:39: It's the industry as a whole, it's society which is changing. I think there's a lot of stuff coming together right now. And AI is also one part, which is more in the operational section of this change, and it's happening on so many ends that I would say, there are multiple reasons why now everything comes together and it feels like there's a big, big impact now.

00:15:05: And I think this impact is something which is now getting more obvious, but it has been there before. But you could hide it a little bit and now you cannot hide it anymore. It's like it's you have to change now. So everyone is forced to change.

00:15:19: You cannot say some old phrases anymore because people realized that something else happened.

00:15:24: Yeah. Like, you cannot put your old system to use. You have to change your system. And this is difficul.

00:15:32: Yeah.

00:15:43: Because, I mean, from a from a bird's view perspective, it's not only, as you said, AI changing the industry. There's a lot of stuff. I also realized which changed before, but I think that AI is kind of the wake up moment for people, because now they say, okay, now we have to act because there was so much stuff before, like, I mean, how long is social around now?

00:16:05: I think 15 or something years or more. So, and I still think that brands are not really doing social right in most cases, you know, still, there's a lot of spots which could air on TV, but they're on YouTube and there are some spots for TikTok feeds which are totally staged and storyboarded and shot by professionals.

00:16:28: You know, this is, I think it has all the reasons. But this goes on for some time already. But now people realize that there is new technologies. And this brings me to the point that there actually are, of course, also new channels, because the system we both grew up in, I mean, you grew up with seven televisions, but there's other channels, which is of course social, but there's also other channels coming up, like, I mean, in the book we touched XR, we touched gaming, you know.

00:17:01: So, how do you see, this whole industry also having to deal with not only technology changes, but also, you know, target audiences behaviors, preferences and other output media.

00:17:16: Yeah, I think yeah, as I said before, the society also changes. And this is something you feel really on social. I think when social started out, everyone was was trying to do their advertising, which they had already to make it work for social and they just made cutdowns, several versions out of it. And I think it worked.

00:17:35: ok-ish back then, because when you started with advertising, you had a like a big attention, but also the way people are using social really changed. And I think working from super easy going snapshots on Instagram to high polished Instagram look, staged settings now into.

00:17:59: The authentic Instagram look.

00:18:02: And now into unhinged storytelling and totally spontaneous. I think there's a like an evolution also going on and it's hard I would say for people doing commercial content to be always on the edge and the tonality, what kind of storytelling works at the moment. And I think it's always lagging a little bit behind, when you're doing advertising.

00:18:25: But I think it's getting better now because special agencies and special creatives can really focus on this. But to use this medium, as how it's intended to be is also changing all the time. And, this is like, this is not a set reality like you have for, TV advertising.

00:18:47: You do like, a standard 32 second commercials. There are some rules which still work. Even 30 years ago. They did. They work nowadays as well. You know, it's like a format which is known and I think about advertising on socials, you know, you have like, you know, something like you can adapt the TV spot to it.

00:19:07: And you said there has to be a hook in the first three seconds, then it needs to be a hook in the first second, then, you know, like, what should you do? Like, you try to measure it and say, okay, this works, this works, and then the algorithm changes again. Then people's preferences changing again. Then you have some trends coming up.

00:19:25: It's a super fast medium. You always, have to be ahead of the wave, as a creative, because otherwise you're just running behind and you will not reach the, the audience you want to reach. I think that is totally different to classical advertising. And, it's much harder to do.

00:19:46: It is probably, especially when your whole process system is set up for TV approvals, because we both know the TV approval chain. If you do like a 30 second commercial and you have like, I don't know, six weeks or something. So there's like certain steps which you take, which is timing was impossible to react in social terms. So how do you see that situation?

00:20:04: Because I think all the big players in the TV commercial segment, are operating on a certain system which is not feasible for social. So there are new players coming up, do you think that the old like, let's not call it old, let's call it the classic processes, can adapt to that, or is it like totally changing and there's totally new companies coming up doing it.

00:20:31: Yeah. They cannot adapt to it. It's something you really see a difference now in the companies, which are now coming up and are new, that they approach the whole process and structure differently. And it's also something we are really pushing hard at a Tektite, at our company, it's like because I was used to the TV approval process, like you said before, for 15 years.

00:20:58: So I knew everything about the timing and the PPMs, you know, and it was like such a strict structure, you have to adhere because it was also learned at the client and in the agency, the production, it was something you could rely on. It was a fixed system. Everyone knows what to do when you're working some years in the industry, it was clear which process needs to be done.

00:21:23: But if you want to have the advantage of new production and AI production, you also have to change the processes and structures. And there's like no ruleset yet how it should work. But this is something we all need to reinvent at the moment, and this is something we are really trying and not trying, but already executing and seeing first results, which are amazing.

00:21:47: I mean, I was used to like, you know, from a project go to a PPM, you probably have something like three weeks, you know, to do a proper pre-production and maybe, maybe it's sometimes four weeks or two weeks. But, there's a big, big process going on after the PBM, everything should be set. You probably had a PrePPM before that as well, you know, with the agency and with the client to get an approval on the cast or location.

00:22:16: But I think these are all processes which take a lot of time and focus and are really expensive. And this is something the client doesn't want. You know, they say, okay, we want a new kind of production, new AI production. We want all the benefits from AI. But then you need to change the processes and structures because otherwise you're trying to press in new technology into old processes and structures, and then it will be expensive and slow again.

00:22:42: You know, you just changing a fancy tool, for a shooting. But that is not how it works. You need to change the whole process. And yeah, we are approaching it totally different at the moment. So there's no PPM anymore in AI productions.

00:22:55: Okay. Yeah. Good. Good to see because I mean, interesting because.

00:23:00: We call it differently. There's something like a PPM

00:23:02: The PGA.

00:23:04: Pre Generation Approval.

00:23:06: Okay.

00:23:07: It has a kind of a similar function, but it's with all the stakeholders. And after the PGA, one week later, you have a finished film.

00:23:17: Oh my God. Okay. Because I remember that, at your times at David and Martin, a nd also now, the latest, not the latest, but the last one we talked about at the Content Club from Tektite, followed the old process. So because, of course, I mean, people are used to it and you cannot, like, totally throw it upside down and say, okay, you have to eat this now.

00:23:37: Otherwise it doesn't work because people have to adapt to it, which takes some time. But I realized that you use the old classic process, like with the casting and styling and the PPM to guide the people through the process and then on point, say where it's changing now. So it's good to hear that it's already evolving into a PGA, which, is interesting.

00:23:59: So how does the clients react to this? Because we all know when you just said that. Well, that's what I was laughing when you said in the PPM everything is fixed. And we all know that there's always like something, some loose screws which, are still, moving. So, especially when people come in, which is the case in many big companies who haven't been in the PPM or, stuff changed, people misunderstood, which is usually not possible in the PPM because you really talk details.

00:24:29: But if people come in who haven't been there or higher decision makers who have different opinions, you always have to change stuff afterwards. And from all I hear from AI productions, because if you said after the PGA, like one week the film is finished, how do people react, or is there less opportunity and wiggle room to react?

00:24:50: People have to take then is it like 100%? Because in real production there's always something you change in the online.

00:24:58: Okay, it's wrong to say the film is finished, but you have like a finished film which works already. So like after one week, we have something where we can talk about moving images already into what kind of order, how does it feel? How long is each scene and how is the expression of the cast and everything.

00:25:19: So really talking about stuff which is happening inside the film and not so much about settings, lightings, details about set design, because we thought it's good to have a first draft really, really early because it's much easier to speak about the film and the feeling it gives. Then to talk about details where you lose yourself sometimes in details which do not play any role in the final film.

00:25:47: So you look on to like a still of an image and you say in the background, this screw or this table, I think it needs to be different and you don't know why, but when you see it in the finished film, it doesn't feel wrong. So I guess it's good to have like a first draft.

00:26:05: And from there do some kind of lean development on it, more like reacting to what do you feel? What do you think? We can change everything. The thing is, on the first draft, with an AI production, everything is still possible. You can basically change the whole setting, the whole cast again. And it's really easy to do.

00:26:26: So that means because I remember the biggest problem in the beginning was like character consistency and like reprompting takes anyway, which looked like kind of the same like the takes you had before. So it's still a problem or is it like in the process you can really change details, because I remember that was the biggest thing like one year ago.

00:26:47: Yeah, I mean that there are some limits to it because sometimes you really have to have to prompt something which you need to rely a little bit on the machine's cooperation with everything you want to achieve. And sometimes you really hit it and think, okay, that's perfect. But, if Ix want to change this or that during the acting, it's probably hard to regenerate.

00:27:13: But if it's something subtle, you know, like, a hand or also a little bit more smile, you can do that. I mean, you can you can edit it. You know, it's like it's easy to do, but it's it's hard to do, like a whole camera movement, actor movement, which was right on point with the scene to do it exactly the same but just different.

00:27:32: It's really hard. But speaking about character consistency and it has been solved.

00:27:39: Okay.

00:27:40: Our actors look the same the whole film.

00:27:42: Yeah, I saw it. Actually one of the projects you did, I showed it to people and even producers from the advertising industry, they said, okay, what did they do with AI? And I was like, everything.

00:27:56: So and I mean, that's for sure some classic process behind it, like editing, you know, it's like editing something which has to do with rhythm, speed and timing and the feel for the film. You don't let the AI do the editing. This is something you need someone who understands which rhythm works and you need like a human understanding of it.

00:28:16: But good to do with AI film stuff, but you put it into a classic edit editing timeline.

00:28:25: And that brings us maybe to the point, because some of our listeners, you know, they're still we're not in Midjourney prompting and, take finished, we cut it together states anymore I think so as far as I heard. So can you maybe explain, the process? Because there's a lot of tools involved. There's professional pipelines. And just for the listeners to, explain a little bit how to, you know, how can we imagine this if you do a film?

00:28:59: Because there's of course, like there's reference modes, there's products, there's, you know, there's you used to work with directors, which we probably still do. Maybe they're called different, you can tell and DoPs and a whole lot of crew and now you have different people. So can you tell a little bit about the process?

00:29:16: Yes. So I think basically when it starts out, with a script development and everything, that's it's still the same, you know, you have to have an idea, which either sells the product or makes the brand look great or has some kind of image to the branding. And it's a classical development. For sure. I think the, the work you're doing for like strategy and research and development regarding the brands, this is probably heavily AI infused nowadays.

00:29:44: I think there's no strategy department in any kind of agency or client who do who doesn't use AI to research. I think it just makes things faster and easier and more precise to some point. And after that, you go into the creatives or you try to develop the first ideas. And this is also something where you need a creative gut feeling, which is based on the data you researched, how you come up with a script which works.

00:30:13: And this is also something you would not give totally to an AI to say, like come up with scripts based on this strategy. But I think, every creative likes to have some kind of sparing sometimes with AI just to, to check an idea, to see it from a different angle, different perspective. If you don't have a creative partner you can talk to, you can talk to the AI.

00:30:36: But the creative work regarding the first concept, you have to do yourself as well, for sure. So this is like more like a classical classical process. And then also when you go from the concept to a script, this is also something you need filmic knowledge. I mean, you can try to let the AI write a script and scenes and, and camera angles and movements.

00:31:00: It would probably do an okay-ish first draft. But this is also something if you would take this and do a finished film out of it, this would feel wrong or like would feel generic, basically. So it's hard to put in handwriting to it if it's made from a machine, you know? So you you need some director's knowledge for this.

00:31:23: Because they know what kind of angle, what kind of camera, what kind of lens, works, just out of a storytelling perspective and also what works best for the brand, you know, because there's much more than only telling the story. It's also what kind of look does it have? What kind of, warmth? What kind of feeling, does it provoke the viewer's, audience?

00:31:49: And I think it's this is something where we still need humans, to work out this first part and also then when you work it out and then you the final script and you want to sell it to the client, I think it's great to do some pre-production with AI to show some visuals, some first moods, some first film clips.

00:32:10: This is something where we feel like the process was getting much, much better and much more precise with AI to show the film you have in your mind to someone who doesn't know this film. So this is like really a step forward. And then, when you get an approval, okay, we want to do this film, you already have probably some prompts out of the testing

00:32:33: you did before, a lot of the stuff you did, just to show the idea or several roots of an idea and, Yeah. Then you have to do some kind of approval meeting, which is, which is the PGA, the pre generation approval.

00:32:46: I like the word.

00:32:47: Yeah. And the pre generation is not only the pre generation which happened before this PGA, but it's also pre generation because we are generating after this meeting. So the pre is meaning before and also after this meeting. But approval means we need to focus and we need to approve that, we are doing this film now with this kind of look, this kind of moment, this kind of storyboard and this kind of cast location.

00:33:12: So basically it's a little PPM, but it's much further already. And it's also, from this point on, much faster in the first stage of the film. So normally the client would see the first part of a moving image when they are sitting at the set and looking at the preview monitor. And this is a day where everything needs to happen, you know, or two days of shooting.

00:33:39: It depends on how long you're shooting because you need to watch everything on the preview monitor. You need to approve it. As you know, you walk to the to the client stand and say, do you like it? And they have to approve it. And after this, it's hard to do another take of the same cast in the same location because it's shot and it's done.

00:33:58: And for us, we try to have this preview monitor, like after one week already and they see it, but we open for feedback. So basically the production is already wrapped in a classical way, but they can still say, I want another take of this. And if I need more smile here, and I think we need a different camera angle here and then we can still do this.

00:34:21: So it's basically it's an endless shoot if you, if you want, if you want like this because you can still edit it and there we have some more kind of feedback channels so the client really can control and can decide what they really want. And then we go into some kind of agile development of this film.

00:34:41: You know, it's like, we working on it and send another versions and another version of this take, send a different version of this and work on the final film till we have some kind of offline approval if we want to use this word, we don't use this word, it's just the version which we all like, probably after 1 or 2 feedback rounds, we have a version we all like and then for sure, at the moment there's a lot of classical post-production workflow happening after this.

00:35:11: Because AI still has its flaws. Not everything is perfect. And this is something you see in this first draft of a film, it's hard to do every effect. It's hard to do very precise, pixel perfect compositions of overlays of stuff happening. And this is where we combine classical workflows. Then we take over with classical workflows, like retouching, compositing, after effects, etc. and bring the film to this level of detail, which everyone is used to.

00:35:45: So from there on, we have more like a post-production, but which is also heavily AI infused because the post-production process is sometimes also part of the AI workflow. But, yeah, it all depends on what you want to tell with the film and how, much of animation stuff is inside, how much of a surreal world is inside, how much of a real world and what you really want to tell?

00:36:12: And, and yeah, there are several direction which it kind of.

00:36:15: Is there actually, because I was thinking about, the picture lock, is there a new word for the picture lock? Is there a picture look, and is there a new word for it as well? Do you want to coin a new new word for the picture lock?

00:36:26: Yeah. It's I think we call it still picture lock.

00:36:32: But there is one, there is an end to it. Because when you just said we can because that's the main, the most unknown, the main thing, the most thing, the first thing which came into my mind when you said, we're not going on a shoot, we can change everything forever. I was like, oh my God. Because usually the people want to do this.

00:36:46: So this it has to be an end to this. No.

00:36:48: Yeah. I mean, you have a fixed timeline usually.

00:36:51: The spots needs to air. And so, but, yeah, there's, there's some kind of, of a picture lock when, when you want to finish it and say, okay, it needs to run next Monday, then we need to do the master copies and everything. But to our impression is if we tell the client, if you want to change something, we can we can still do this after the spot already runs.

00:37:11: This opens up a lot of new possibilities which were not there before. And this is something where clients can really react to market changes. So we did a big campaign, for in fact for the brand, and we did the finished films and they were running on media and they were running successful, hey did the market research also and the market research found out that like in two scenes, people were not reacting the way the client really wanted it.

00:37:42: So they saw some emotions going, not on the same scale like they expected it. And they said, okay, we are really happy with the spot, but, if it's possible, we want to change the expression of the actor and the scene and we want to change to a different scene at this point of the spot. Is this possible?

00:38:05: So usually if you're doing a production you would say, yeah, we have to do one more shooting day. We have to get the same cast, go to the same location and do another shooting. It probably cost about the same like the spot, because, it's really expensive to do it well and on the same level. And we said, yeah, we can do that.

00:38:23: We can send a new film tomorrow, but we use the same existing scene, and we changed the details and we exchanged those scenes, and then it's done, and you get a new version.

00:38:34: And the new version are actually maybe one of the jokers in AI filmmaking, because, I mean, this is because at the moment I realized that people are still, you know, some people say I you know, it's not it's not as good as filming. And I don't think that's true anymore. First thing. But the second thing is that people, I think, underestimate, that you can literally change everything afterwards.

00:38:59: So of course, I mean, there's always like big VFX in productions and you can, you know, retouch things and but you can not reshoot. So but I mean, talk a little bit about the possibilities of versioning and iteration maybe.

00:39:10: Yeah. I think it's one of the big advantages and it's still something we as a creative also need to find out a way how to use this wisely because, exactly the example I was given like change an existing spot into a different version. This could also mean a different market. This could also mean a different target group because we can change the cast, we can change the acting, we can change the wording.

00:39:35: I mean, we can change everything basically, even if we have an already finished spot. And this opens up the possibilities. We did not have before when we were talking about versioning, because versioning was more or less, yeah, we have a different length, we have a different aspect ratio. We probably sometimes has to have a different language which needs to be dubbed, but that's it.

00:39:59: You know, this is the stuff you could do with, with the material. And probably sometimes you have like a different ending. You do a different packshot. Yeah. We have new products.

00:40:07: Oh, I remember, I remember that, yeah.

00:40:08: Yeah. Like pre-roll. I wanted to have a new poster or whatever, you know, this is something like usual ask from the client, we needed a new packshot.

00:40:21: That's actually a good a good Stichwort for, for, I want to talk about hybrid production. So, actually, you know, I was thinking about this before, like a year ago or something, when you're sitting in a PPM and then you talk about takes you have to film, and then there's like a lot of packshots and you need like a packshot for a different country.

00:40:41: And then you have to, you know, you have to shoot plates and empty whatevers. And you cardboards and then you have to go to post and fill titles and stuff. So how about hybrid production? Is it also possible to I mean, are you into that or is it generally possible to make a normal shoot and then just do the iteration, the versioning in AI?

00:41:05: you know, for example, different pictures, language versions or whatever?

00:41:09: Yeah. We're not into hybrid production. We don't like this word. I mean something like in the past we were doing real shoots and had like a CGI component to it, you know, so we replace the actor, replace the car. We do like a little comic character with it. So you had like a real shoot, which was enhanced or edited by CG .

00:41:32: It is a hybrid production. I think this is just a word someone came up with where they try to.

00:41:37: I think it sounds very good. You know, we have a hybrid unit now. We're doing hybrid as well. No.

00:41:42: Sorry, I don't like this word because it feels like old production companies trying to find a way to tell that they are now also doing AI.

00:41:51: Yeah, okay.

00:41:52: It's a CG production basically. I mean for sure I think at the moment they use it just to put out the word that they want to adapt to AI. But, I think it's something we should not use. You're doing a real shoot with CG elements that would make it because CG is computer generated, AI is computer generated.

00:42:16: End of discussion.

00:42:18: Okay, good. Yeah. But talking about, you know, because you're from production. I'm from production. So, if you watch the industry right now, I mean, what do you think is the biggest challenge for the production companies who are classic and now go hybrid or also work with AI? Because I think there is some clear advantages for AI productions.

00:42:45: There's still some reasons to, like, shoot a real film, you know, whatever, like tier one brand campaigns, whatever. How do you see the landscape changing, like first production companies and then generally the whole process chain of who commissions, who, and, you know, where is this all going? Because, AI so fast. Yeah, that this is going to be soon maybe.

00:43:07: Yeah, absolutely. So first of all, I love real shoots. I love still real film sets and everything. I mean, this is where I spend most of my work time, on, on real film shoots. And, it has a magic. It's hard to do, on the computer, actually, you know, when, when most creative people come together and generate something on set, it's still really valid for a lot of projects and a lot of stuff you want to do, and I hope it will persist.

00:43:39: You know, it's like, I almost sometimes feel bad because we're doing AI only production, but I think there's still a place for real shoots. And if we get sometimes a script where we read it and it feels like, this is more like a real shoot, then we say to the client, it's something we won't do because you need to shoot it for real.

00:43:59: And like our company, Tektite is focused on full AI production. So either we are doing it fully AI or we don't do it at all. I mean, for sure, sometimes we're doing like reference shots or if we want to integrate something into the film, you need some reference images for it when you want to generate it.

00:44:17: But I would not call this hybrid. This is more like reference material for the generation. But yeah, I think where the classic productions are struggling is processes and structure. And this is something they need to reevaluate how it works. I mean, because if you have like a classic structure, classic process, you cannot put AI inside because you will lose all the advantages of AI.

00:44:45: You will make AI expensive and slow again if you press it into old processes and structures. So AI is a new category of filmmaking, which is basically a new genre which needs to be treated as this. And this means different understanding, which we all need to learn together actually is the process we need to do together how it's the best way to use the advantages and the speeds and, yeah, everything which is possible with AI and not lose it during the process, which is old school.

00:45:19: Yeah. So that applies to production companies, of course. In a way it also applies to agencies because, I just made a note here which says in-house because, how do you see the development? Because when we look back, I mean, when I started, everything was shot on film. So it was kind of impossible for anybody else to do it because it was like a secret passed on from Druid to Druid.

00:45:42: You know, it's like only a handful of people knew how to do this. And then there was we all remember the video cameras, where you could put film opticals. There were some dudes in Munich, P&S, that you could put a 35 millimeter optic on the video cam. I have a photo from Cape Town where we shooting with a digital Betacam and, like, a long lens.

00:46:03: So, that opened up. Not that. But then later on, they came the DSLR of course, and now you can shoot everything with an iPhone if you want. So, how do you see this? Also, business model wise, taken away the classic production actually from production companies because that what happened in the in the advertising agencies on a level in the last years.

00:46:27: So I think that this is also a possibility that people will try to do that in-house. Of course. How do you see that?

00:46:36: Yes. I think from a technical point of view, it's totally doable if they invest, a lot of time in skilling up all the art directors and creatives, that they can use, technically, I think it's possible everyone can come up with a film and probably execute it, and you will have a film and it will be enough, of quality to run it on socials, TV or wherever.

00:47:00: I think it's possible. But I think what's really hard to learn is filmmaking itself. So I think it's easy to learn a filmmaker, how to prompt. But it's really hard to learn a prompter how it's going... okay, let me rephrase it. It's really easy, to learn for a filmmaker how to prompt, but it's really hard for prompter, how to make a film.

00:47:29: And I think that's the main problem. Also, when you talk about what is possible, with all the stuff agencies were doing before, like visuals out of homes, motives, for sure. A client can do it themselves. They can prompt something. They can go to Midjourney, they can go to Nano Banana and do a visual. But will it be on the same quality of creative level?

00:47:52: Then when someone is doing and who has an experience doing that, I doubt it. So it's hard, to do this well, you know. The technical limitation is gone for sure, but I think the creative limitation is still there.

00:48:07: But there's a big danger of a lot of slop probably that you I mean, because people will try to do it themselves, which is okay-ish, you know, and then, if it's also money question, we have to come to that as well. And it might result in people doing stuff because they can because, you know, when I started, nobody would shoot a film because nobody could.

00:48:29: I know they can. And so is it an opportunity or is it the doors opening that a lot of bad content will be on the internet?

00:48:41: I don't know, it's, I think for some people it's some kind of awakening that they can try it out to have some good, output already if they use a tool. And. Yeah, it's nice, but it's hard to do it on an exceptional level, a level which, differentiates from others. You know, this is, this is something where they first feel, okay, now I'm enabled, but then the limitation is already there.

00:49:03: Because when you give like someone, like an access to a new video AI tool like VEO or something like that because some companies were using enterprise licenses and giving out 20 licenses to all their in-house teams, and they were prompting and I think I spoke to one client and, and and she said, yeah, like they were trying out a lot of things, but everyone was doing like guerrilla V-logs, because they have seen it on social that it works, you know, like Gorilla telling something about the brand filming himself and stuff like this, because it was this trend of V-logging back then, and everyone was doing the same because they saw it on socials.

00:49:41: And I think this is something where, like a filmmaker would approach it differently, they will question the storytelling itself, like what can we really do? And they would like feed their, their creative toolbox out of all the stuff they've shot and all of stuff they seen, they can reference in their minds which kind of stylistics they have what kind of lenses, settings, etcetera.

00:50:04: Directors. And I think that's something you need the experience for.

00:50:09: So I mean, there will be I mean, of course, I think because that's the money question we're coming to now, because I think there's going to be a lot of content which will be done in-house or automized to a certain level, because it's also about, which quality you need actually for the end clients, for the customers to understand the product loved it, buy it, or maybe the love is coming from a different campaign, and maybe it's just about the buying.

00:50:37: So there's going to be a lot of content, which can be then, I mean, I'm not talking about full optimization, but there's a lot of tools for people. If they have shot a tear one brand campaign, to adopt like socials in all formats and stuff, maybe in-house, because then you don't need the whole process chain with all the players you have now.

00:51:00: And that's actually also a money question, you know.

00:51:02: Yeah, it is. And also when you talk about this, we need to talk about performance advertising, which I think is a different discipline than the creative advertising we're talking about. So I think performance advertising can be done in-house completely. And because the platforms are offering now all the services you need to get all the insights, from the reporting, from the data you need, what kind of target group you want to address, what kind of format it needs to be played, you have an A/B/C testing, you have everything, from the toolset, from the platforms.

00:51:33: And they are really expanding this knowledge. And I think this is a battle which is lost.

00:51:40: I'm just I'm people on the video see me smiling because I was thinking about the TV quota, like in Germany, you have this, I don't know, 5000 households which have like the thing to actually measure the quota ho's watching which program, which is kind of the opposite of this. But this is just a little side note. So yeah.

00:51:57: Interesting. And so as we're talking about the budgets, you know, when you said that you sometimes get script where you said, okay, this is probably for reshoot, why do people address you in the first place? Is it because they think, this is I know we are faster, cheaper. We can do everything whatever or what's the reasons?

00:52:18: Yeah, I think people also want to try it out and want to experience themselves what is possible, what not. It's sometimes really hard, to say when you read a script, okay, this can be done by AI or not, because this is changing all the time, you know? And what kind of quality can you expect. And sometimes you get like a briefing sometimes from agencies or clients, and they have some kind of AI disclaimer at the end of the script where it says, yeah, working with an AI is always hard.

00:52:48: We know that the outcome could be totally random, and you cannot control. And like sometimes there's stuff which is not right anymore, but which was correct, like 2 or 3 months before. But now we can solve it. And it's really hard to say this is possible and not because it's only, as a status quo, which you're referring to, and it changes all the time.

00:53:11: So I guess when you think about film scripts, it's not easy to say this can be done for real. This can be done with AI because it's only valid for one week. The statement.

00:53:24: Yeah, yeah, I remember you're telling me when you booked for keynotes, you only do it like one week in advance because otherwise you can rewrite it again.

00:53:31: It's outdated. Yeah. And then it's still like sometimes we look at old workflows we were doing like three months ago, and now we say, well, this is funny. Like we did it this way was super complicated, super expensive, like from GPU time and everything. And now it's done totally different. And I think this evolution of how we are working with AI will be going on for a lot more time now.

00:53:59: It's like it will be not done end of the year. We can agree on a workflow how we do it. No, it will change again.

00:54:05: That brings me actually to, the kind of summing it up. So what do you think was the biggest what made the biggest difference in the last? Because in the last 12 months, because we're talking about a long time already. I mean, not a long time, like in terms of, I don't know, three years, like, we need to talk about this.

00:54:23: So, what made the biggest difference, which actually also made you, like, join the AI studio right now, you say, okay, this is the moment. Now it's going.

00:54:34: It's not like one certain event, I would say, but I think the quality was rising all the time. I think there was a big step, you could say when Google Nano Banana was coming up, was ready to use because it changed a lot of the workflow because we were training a lot LoRA's before and training models, on high GPU times and have to do a lot of reruns.

00:54:59: And I think nano banana changed the workflow a lot, but this was not the reason I changed, from agency to AI production. But it's something, I felt immediately this could open up a lot of new doors, which we can step through for sure. Yeah.

00:55:16: And what what do you think is the biggest potential in the next three years? For films?

00:55:22: Yeah. Yeah. I think automation workflows are super interesting. And I think the, the biggest thing which we need to change is how we are telling stories. And this touches on so many topics because as I said before, when you're doing like a commercial, the 30 second commercial, usually on TV, this was then finished. You know, you have the approval.

00:55:48: Spot is finished. But I think now it's the start. It's the start to evaluate it and it's a start to test it and to make it better. With no going from finishing a product into a lean or agile kind of working with creative stuff. And this is more like a software development where you work in sprints, you know, it's like not exactly the same, but it has some similarities to it because I was working in software development, it's great to have iterations of stuff you are doing because you can always improve on it.

00:56:21: And this is something where we need to rethink how we are telling the brand. Which kind of assets does the brand have, which kind of legacy, which kind of origin can we can we use everything out of it? And also, can we adapt to what comes back from the client or from the audience or from the buyers? So what is the feedback

00:56:44: we are getting on socials? What is the feedback we are getting with a product being sold, you know, when you're advertising like for three different products like, and one is getting bought more than the other, like, can we change something and like you can react to, to the market much more. And I think this is a big revolution where in the past you would always need to come up with a new campaign, different campaign, different approach, and now you can iterate.

00:57:09: And I think we need to learn this.

00:57:12: Great. That actually comes to the next point already, because if you could give three short advices for marketing decision makers, how? Because it's like a we have a big bucket of fancy things now, so we could. It's always the problem with the idea that you have like so many stuff and you can do like kind of everything.

00:57:32: What is the, what can people do now? So what's the what's the reaction on this?

00:57:38: I think, it's really good to get an early and early first draft of what you want to do. And see for yourself if it feels right. Versus a big, big development, which takes months and a lot of money to be done. So I would say, try it out quicker, test it and improve it from there.

00:58:06: And I think this needs different thinking than before. I still believe in big ideas which need to be worked out. But I also believe in the first draft being done very quickly. And I think this is just one advice. The other advice is question processes and structures. Question why it does take so long, why we need ten people in the meeting?

00:58:31: Why do we need another approval chain? Why do we need this? So re-question everything. It's good, for the whole industry, because then they need to adapt to what you really want. And also, free yourself. And this is the hardest thing to do from stuff which was impossible before, storytelling wise. Also production wise, because this is not there anymore.

00:58:58: You don't need to think in locations and costs and situations anymore. And to lose this, is really hard.

00:59:06: But it's a freedom in the end.

00:59:07: It's a freedom, yeah, but it's also hard, you know, to start from a blank page. It's much, harder than to say, okay, we can afford one shooting day with four people and, in three locations that it's very easy to do. You can write a script with this, you know, you can start to develop. But when you say you can do whatever you want, it's super hard.

00:59:30: Great. So, Amazing. So I'm actually. Yeah. Thank you for all the insights because, there was, a lot of stuff in there. And, so it's like a lot of new possibilities, which you have to have a look at and, find for yourself what to do with it and, not jump on every new train.

00:59:52: But, you know, I think now it's unavoidable to actually take a look at it and, change old patterns, maybe, and embrace new possibilities, because I think what I always say is, I mean, it's there you know, so people will use it, talking about AI, so it makes sense to maybe not be the last one to, to adapt it.

01:00:17: And then there's, interesting times ahead.

01:00:21: Certainly. And, I think it's it's a great time if you if you're struggling, if you think, okay, the whole industry is going down now, it's the perfect time to step inside and to do something you always wanted to do. It's a great time.

01:00:36: Great. That's an amazing finishing statement. So, if you want to catch up more from Max, of course in the show notes, you can find the LinkedIn profiles and TikTok website. I will also post maybe a little glossary. I was thinking about maybe explaining some terms we used, like LoRA, and V-log and stuff.

01:00:56: So I will do that. And of course, there's also the link to the Commercial Content Campus, where you can catch up cases and interviews, and of course, the next episode. Thank you very much, Max, for being my first guest on this new show.

01:01:11: Thank you. It wasn't an honor, Norbert

01:01:12: Thank you very much. And also to the Publix studio. No, sorry to the and also to the Publix studio here in Berlin Neukölln from where we're broadcasting. Shout outs to Publix. Thank you very much. Thank you. Bye bye.