00:00:00: JOHANNA: Welcome to Industry Insights, the EFM podcast presented by the European film market of the Berlinale. My name is Johanna Koljonen. I'm a film industry analyst and experience designer based in Malmö, Sweden. Industry Insights, the EFM podcast puts a spotlight on highly topical and trendsetting industry issues, creating a compass for the forthcoming film year. The year-round podcast is produced in cooperation with Goethe-Institut and co-funded by Creative Europe MEDIA. My three guests today all participated in the EFM Startups program in Berlin back in February. 2024 has been a year of AI reckoning for the audiovisual industries and we thought it would be cool to reconnect with these EFM Startups alumni to see what's happening with their companies, but also to talk about how they're using AI technology and dealing with the more complex tensions between our industry and what AI is promising as well as threatening. Welcome to Industry Insights, Zoe, Anna and Fredrik. I'm so happy that you're all here. I thought we'd start with each of you just describing briefly what your startup does and what your role in it is. So if we begin with Black Goblin Audio and Anna Betancourt, your product is called, is it Thol? T-H-O-L?

00:01:22: ANNA: Yes, that's right. Black Goblin started as a consultancy back in 2018. And then about two and a half years ago, we started pivoting slightly to build a product in technology, basically to reduce time-consuming tasks and repetitive tasks in sound design, particularly for film and TV, which is my background. I've been doing this for 14 years now. The whole problem that we're solving essentially is that it takes around 450 hours to basically edit sound for a standard 90-minute film. And 250 of those hours are actually spending in sourcing and sort of cutting sounds and making them fit the image. That's a lot of non-creative work that we do as sound designers.

00:02:09: JOHANNA: So immediately, I had to ask, when you say sourcing sounds, what kind of sounds are we talking about?

00:02:15: ANNA: We specifically focus on sound effects, ambience, basically everything that is not dialogue and music. Everything comes either from recordings or from sound libraries, professional sound libraries. And that kind of bit of, if you want to call it admin, it's not admin, it's obviously also part of the creative editing work. But it is quite boring and quite repetitive. That takes a long time. So that's where we exactly go in and solving that particular issue.

00:02:41: JOHANNA: That's really exciting. And is this a product that exists now, Thol?

00:02:44: ANNA: So, Thol, we actually just conducted a pilot with the BBC that finished at the end of October. And we are running a private pilot with five sound designers, three of which are actively working in Hollywood at the minute. And we're doing a soft launch in March for the wider public.

00:03:00: JOHANNA: That's pretty exciting. How much time savings are they reporting at this point?

00:03:06: ANNA: So at this point, we're going in average with the work that they've been testing from around 250 hours, which is the average that they take with the traditional methods, to about 50. So we're doing pretty much a fifth of the time.

00:03:20: JOHANNA: Well, that's really exciting. That's very cool. All right. Let's continue to Zoe Ramusho and Wrapped. And that's Wrapped with a W, as in it's a wrap.

00:03:36: ZOE: That's exactly it, Johanna. So I'm Zoe Ramushu and I'm the CEO and founder of Wrapped. We are a crewing app and we got started in 2023 with the EFM startups, as well as soon after that, we were nominated for an award in Cannes. And we've just basically been working with a couple of partners similar to Microsoft to obviously evolve our tech. What the app does is it just simplifies the process of crewing up. I have a background in filmmaking and I've been doing this for almost a decade and I just couldn't understand how still, you know, years later, we still use like WhatsApp calls, Facebook groups. And that's the way that crews kind of find each other. So as a producer, when I'm crewing up some of my smaller projects, I hire between 60 to 80 people on bigger projects. You know, it's 100 to 200 people and there's projects that go into the thousands. And all of this is done manually. There's no single database of all crew in one specific place. So what the app does is it helps you to number one discoverability. So it helps producers like me to find talent and crew. But also it really helps with talent and crew being able to be found because the industry is notorious for being a very who you know kind of thing, barriers to entry are high. And the expression, oh, I'm trying to break in comes from our industry. And so what rap does is it's breaking down that barrier where talent can be found by producers, by studios, by companies who want to work with talent. And it's not just limited to your inner circle.

00:05:09: JOHANNA: So it's a database of crew and then it has some other convenient functionalities as well.

00:05:13: ZOE: Exactly. So it is it's an app which has a database of crew and the functionalities are number one, being able to discover crew. So you can search think of Airbnb, you put in the filters of what you're looking for specifically. You can immediately see the date and times that they're available. So that's the same with the crew. You can immediately find out, you can check their resume, which usually takes forever with the back and forth with an agent. I mean, I'll touch on that a little bit later when we talk about ethics and all of that. But we're not replacing agents. We're just optimizing their work, which currently is usually done with a back and forth. You email, hi, I'm looking for a cinematographer and they send you a list of like five different options. They send the CV. You've got to go through it. Now you've got to go back. Are they available for these dates that narrows it down to two? Then, OK, how much are each of them charging? So this conversation is kind of a back and forth, whereas with Wrapped, you can have that entire experience on the app within minutes and be able to select who you're looking for. So that's the primary functionality, being able to actually find people very quickly. And then the second thing that we solve is the paperwork aspect of onboarding, which, again, like I mentioned, I'm currently in production. It's a nightmare getting all of these people to fill out forms to onboard, to get the information, copies of their IDs, copies of bank confirmation letters. All of these things takes so much time to get from crew who are trying to do their work. They want to be on set. They want to be shooting. But now somebody's got to chase up that paperwork. So, again, Wrapped really optimizes that process where just at the click of a button, you can share all your documentation with production and vice versa.

00:06:50: JOHANNA: This is so exciting. So is it in use? Is the app already launched?

00:06:52: ZOE: Yes. So the app has had its soft launch and we've tried it out on a number of productions. I'm based in Cape Town, South Africa, and I work between here and the US. And so the production companies that have tested it out are based here in South Africa. And we are planning to expand starting in 2025 to Germany, the Netherlands and the UK, our first countries that we've got uptake in terms of pipeline of companies looking to use the product.

00:07:20: JOHANNA: That's fantastic. So then another part like the challenges, of course, the hurdle is to get people to actually sign up to get the crew members on board. So it's not a trivial task, but it's a very appealing, appealing offer. Yeah, that sounds fantastic. Thank you very much. And what about the final company here today is Phont, and that's Font as in typeface, but you spell it P-H-O-N-T. Let's hear from Fredrik Merkel.

00:07:44: FREDERIK: Yeah, the P-H-O-N-T also comes from phonetics, so not just from font. And we are working in a similar field as Anna does. We are also looking into audio, but we have a completely different output. The output is visual, and that might be a weird thing in the beginning. But actually, it makes sense because we are all about subtitles and subtitles are a medium that hasn't been changed in the recent decades. And subtitles only convey the content of what we say, but they don't show how it's being said and what else is happening on the audio. There's so much stuff going on, like the soundbites that Anna basically designs and like everything that happens there in terms of sounds and noises, but also the emotions in the language and linguistic aspects. All of these things get lost on the way. And yeah, we want to take care of that and we want to visualize them. So we basically have different several AIs working in the background analyzing the audio and they get values out and we link those values to design parameters like type and motion design. So if a person is getting angry and is shouting, the letters in the subtitle will get bold and the whole subtitle will change and shift its design. So, yeah, the user and the viewer actually feels the subtitle. Yeah, and that's a solution that is, of course, for the hearing impaired. So it has huge inclusion and accessibility aspect, but also more and more people are actually consuming movies on mute because they are watching it in public or on the way. And also there is more and more for language content being consumed. So that's why there's a lot of subtitle usage also by the hearing. And yeah, so that's that's what we are doing. We are currently building the product and we want to implement it into the platforms because the best vision of a Phont is actually when the user is able to set the appearance and they can make the decisions. We don't want to force the subtitles on anyone. And that's why we want to have it integrated into platforms. But also as a second part of Phont, we are currently developing a web app for the content creators because on Instagram and TikTok, you also see a lot of flashy subtitles. And there's a huge need for that. And we want to have a web app for the content creators so they can upload their stuff on our platform and they can make their changes and customize the subtitles and then download it.

00:10:15: JOHANNA: This is really interesting. And of course, these are two very different use cases, but if we're just thinking about like watching something in my living room on my big screen, I almost said on the TV, I mean, that just being able to set what kind of subtitles is really interesting, because of course, there are different needs. Like if if you're dyslexic, you might not want the font to be squirming around or doing something like that. And if if depending on your needs, what kinds of settings you would use is already interesting. Subtitle innovation also feels like one of those things that we had wouldn't have thought about at all. I can think of only one recent example, which is there was some TV show, I can't remember, maybe Pachinko, one of those that sat in in like a multicultural family. And I remember they were using for the English subtitles, different colors for different languages so that audiences who don't can't hear the difference between whatever it was, Korean and Japanese, maybe can tell when people are speaking different languages at each other. And I thought, well, that's insane. Like, that's very cool. And now you're saying we could use this to visualize everything that's in the audio. That's crazy.

00:11:22: FREDERIK: Thank you. Yeah, that's a basic idea. I mean, there has been a lot of innovation in the past, but it hasn't been so generalistic. For example, the British used to change the colors for the speakers. And that has been a thing for sure. Also, recently some shows like Stranger Things, they experimented with that and wanted to visualize the audio closed captions. And they had some experts like linguists and also emotion designers working on it. But it hasn't been optimized yet. And that's what we are doing for the first time now.

00:11:54: JOHANNA: All of these projects are amazing. I'm so excited to see them come to pass. So what we're talking more about today, of course, is AI. So let's just look at how AI is being used. Maybe still with you, Frederik, since you mentioned at least two different uses. So AI is helping you analyze what the emotional content is in the films, for instance.

00:12:14: FREDERIK: The first part is actually the transcription. If there is no transcript existing, then we do an auto translation if needed. And then we have the audio analysis. And that's like the core part of Phont where we use AI. So we analyze the AI on emotions, on different voice features. So we have different sorts of analysis we are programming right now, for example, for pitch, for the voice pitch or for the emphasis, all of these things. And then, of course, we want the accessibility feature. We want to have the speaker diarization. So we actually know who's speaking right now. So that's a different AI. And then there's also, for example, the soundpad recognition. So what's happening in the background? Is there a firefighter driving by? Stuff like that. And all of these things we can actually visualize. In the future, there's going to be much more use cases, but that's what we have right now.

00:13:12: JOHANNA: And of course, this is work that would be completely unfeasible to have somebody manually do this analysis for, let's say, all of the streaming services catalog or something like that. What about you, Anna? What role is AI playing in Thol?

00:13:25: ANNA: Yeah, so it's got several roles. Basically, if we're talking about the workflow itself in that order, the first thing that we start with is computer vision. So basically video analysis, it's been an ongoing task for us. And basically, we analyze, we take in a video and we have to extract information about general objects and then track them in order to figure out what's happening, what actions are happening, and then use that information to determine context. So, you know, someone is walking in this particular place in this particular form. And then that whole thing is taken with that language, kind of natural language processing so that we can then use it to generate sound effects, basically high quality sound libraries that match the particular timings of the actions that's generative AI for for sound. So there's basically we're using the three big ones, so a large language models, computer vision and generative AI for audio. And that's the way so the whole workflow, it's about automating the parts where you're spending hours basically saying, OK, I'm going to watch this 90 minute film. A person enters here. Let's put a little mark here. A person now jumps, a person now pets a dog. All of those things are automated and already marked in your session. And you have a high quality sound library, which in turn also saves you lots of times basically scrolling through different sound libraries, buying packages. Right now, sound designers are spending an average of four billion dollars every year worldwide on sound libraries. They are not necessarily fit for purpose, that still need a lot of editing to bring down to what the content is. And so that's the particular expenditure that we are tackling.

00:15:08: JOHANNA: Now, I think many people who listen are in the industry somehow, maybe they haven't thought very much about this. So let's say a person like I'm trying to give like a really dumb example. But like you said, people are walking. So if people are walking on film and we can hear their footsteps, typically that isn't actually that sound isn't captured on location, it would be added. So somebody needs to identify every time where a person is walking within hearing range in a film, which can be quite a lot of it. And then find the correct footsteps and add them to each footstep basically in the film. When you start to think about it on this level, it's an insane amount of work. OK, so in addition, you're also like generating the sound instead of trying to match an existing sound.

00:15:53: ANNA: Yes, so at the minute we are doing particularly that, but we also recognize that our core users are sound designers. So we are going at the people working actively now on this rather than around them. And so we do recognize that they do have sometimes terabytes that they've accumulated of sounds that they want to use. So that's another aspect that we're currently looking to fit into our proposition. However, because it is such an ambitious product from a technology perspective, we are basically tackling what's the kind of burning problem at this particular point, which is that aspect of lots of admin tasks that need to be done. And in terms of what you're saying about how sound design works, that's absolutely right. I always kind of think about or bring down the example of, you know, think about Star Wars and what's the first thing specifically think about Darth Vader versus Skywalker. What is the first thing that comes to mind besides I am your father, right?

00:16:45: JOHANNA: I can't even do it with my mouth, but yeah, like it's the lightsaber sound, right?

00:16:50: ANNA: That's it. But, you know, what if I told you that basically Ben Burt, who's the sound designer pioneer for Star Wars, took basically days or weeks to actually come up with that. It was the thought process, of course, the creative aspect, which is the fun thing, but also just layering different sounds. And that's what sound design is. You're not just getting one particular sound and you're done with it. You have to actually first spend a lot of time finding sounds that you think you imagine that might work well together and then experiment. We are saving you the time that you spend thinking, you know, finding sounds that might or might not work together. We're just kind of kickstarting you and jumping into the part where it's fun and you start layering things.

00:17:28: JOHANNA: You get to do the creative work still. That's great. OK, and then what about Wrapped? How are you employing AI there?

00:17:34: ZOE: So with Wrapped, the major way that we're using AI is in terms of finding the people that you're looking for. So what the app does is it gives you a manual option where you can search by someone's name. You can search by gender, race, whatever it is that you're looking for. You can search by awards or previous films, you know, that kind of thing. And that's kind of the manual option that you can choose. Or you can allow our algorithm to scan and look and suggest specific people for you. So that's the major thing that we're using AI for. And one of the things that I would highlight is that Wrapped is coded for inclusion. So in that, it really is, you know, generating a system that also allows people who currently wouldn't be selected. And it puts people on your radar that you wouldn't, you know, whether it's females, because one of the major things that is an issue in our industry is females not getting specific roles. And I did a couple of years of research when I co-founded this woman in film organization just around why people don't hire women for roles. And the craziest answer, which actually was true, is people are like, oh, I don't know a woman who does that. And it's just that basic thing where people sometimes just don't hire people because they don't know them. If you're looking for a female cinematographer, for example, it's like, oh, off the top of my head, I know two. That can't be true. You know what I mean? There must be more. It's just that ability to find them. So I think that's what's really integral with our being coded for inclusion. It really puts forward people who you may not have been able to find. The more practical thing that we're doing with AI is basically smart documentation. As I mentioned, usually when you join a production, you have to fill out all of the forms, name, ID number, etc., all of your details, emergency contacts and all of that. Once you've set up your profile... Special diets. Exactly. And those things are so specific that need to, for example, dietary requirements. If you've got 60 people on a set, you need to know and send the information to the caterer to say, OK, out of the 60, 30 are vegan, two are pescetarian. You know what I mean? Breaking down all of these numbers, which is usually spent by hours of someone sitting down and creating, if you're like me, an Excel sheet. You know what I mean? Or some people still do it even manually and have to get this information across. However, with the app, once your profile is set up, all your information, once you allow the production, you release your information and that information is integrated onto a greater spreadsheet for the production to just have all of that information. So I think that for me, what was wild when I thought about it is that we actually in the industry, we hire people like so a line producer and a production team. You hire people to help you hire people. You know what I mean? And so now we're basically giving assistance to those people where it would usually take them a full week of pre-prod to just put together a team. They can do that, you know, sitting in one afternoon, you know, and then get to the actual work that they need to do, because set is always wild and there's always a hundred things going on. There's things you can plan for and then there's the things that you can't, you know, and that's what your production team needs to actually be focusing on.

00:20:47: JOHANNA: This addresses a lot of things at the same time. I think it's interesting that all of your services that you offer are just like solving hands on problems or like addressing something in in a very sort of integrated way in your workflows and in your users workflows or in the viewing experience in the case of Phont. And that seems so obvious and great. And that is not the kind of narratives that we hear around AI in film in general. So like if we zoom out a little, we have these two narratives going on around the creative industries where we have huge excitements among investors or at least we've had for a while, pouring money into AI themed things, which seems to be based on very little like actual usefulness. And then on the other side in the creative industries, we have a lot of sort of fear and concern around the use of AI, which is probably also quite valid. But it seems quite absurd when you're looking at like in the context of these three examples. So I don't know, when you're out there speaking to investors, speaking to the film industry, how do people react when you say you talk about AI? Frederick, you're laughing. So what's on your mind?

00:21:57: FREDERIK: It's a completely different language we're talking when it comes to investors or to the industry, because with investors, there's a huge excitement about AI and it's really easy to sell it to them. Because obviously, if you spare people time, what we do in most cases here, or actually in all cases, then it's an easy argument because the time is money and that's what investors want. But the industry is thinking the other way around because time is also their work and they are afraid of actually the work being gone and their job being gone. Yeah, it's different arguments that you make. But of course, you can frame it in a different way. And with most of our products, it actually works that we are not taking jobs away. We also enhance the work and we enable people with AI. And that's also a framing we can use.

00:22:58: JOHANNA: Yeah, for instance, doing this sort of sentiment analysis that might not get done at all without this. But that's, of course, a little bit different for you, Zoe and Anna, because in a sense, this is work that somebody is doing today, even in both of your cases, it's job work that like maybe those people shouldn't be doing necessarily. How do you see this?

00:23:18: ZOE: Yeah, I mean, I'll jump in on that. And to say, like, surprisingly, the industry, well, not surprisingly, but because I am my target audience and client, people are very excited about the app because production is hard. Production is tough. Like, there's a lot of things going on. And especially with budget cuts across the globe, pulling back and making every production tighter. If there is a tool to use that can make your job easier, I've received so much positive feedback. People can't wait to use the product because it's just basically enhancing what you were going to be doing anyway. It helps you to do your job more efficiently and more quickly. And instead of sitting there, I'll just give an example like on this project, we had a DOP who had to check out like I think it was four days before production. And this is like a key, key person. And for us to all put our heads like it was just someone we weren't expecting to not be able to arrive. And so to pivot and find someone, it's literally people sitting and saying, oh, I know whoever, whoever. OK, I need to go check if they're actually available, if they'll be able to come in at our rate. And basically, three or four people have to go out, stop what they're doing for the production and go and phone people physically. Because a lot of people, like I said, with production are on set. You know, they move from one project to another. They don't have time. They're not sitting there checking the emails. And so use this app, for example, would be, you know, you put your date in and you put the rate that you have and it will immediately tell you here are the 10 people that are available. And within the budget price that you guys are looking at, that just cuts the, you know, instead of having three people go out, get on the phones and try and figure this thing out in half a day, you kind of have your answer in a couple of minutes. Another producer was like, I need this. Like when I first was doing my R&D, he was like, can I have this app like today? Because I needed to shoot in another city. Right. And he was like, I just don't know people there. And I don't even know where to start where geographic and geographical inclusion is also a big thing. People don't shoot in places where they don't know people because they're like, I won't even know who to who's on the ground. Is there enough crew there? Are there enough equipment higher places? It just immediately gives and empowers producers and line producers to confidently be able to say, OK, so the shoot that I'm on is an away shoot. And it took me it was like pulling teeth trying to find, OK, is there someone who's experienced enough to do this role? Oh, yeah, there is a guy. I know he worked on another project. Let me get his number. You know what I mean? Now we're waiting for.

00:25:53: JOHANNA: No, that makes perfect sense. And again, being able to move around like outside your current network and also just with how financing is working these days. Like that's a huge threshold for a lot of people. What about you, Anna? Because like I think in the short run, everybody wants to not do the boring part of the work. But over time, this could mean that these professionals get paid less because they can work faster. So so what are the people's concerns around that?

00:26:16: ANNA: So there's a lot of unpacking, I think, in this particular topic. I mean, first of all, it's quite interesting. And I sort of have a similar experience to Zoe, specifically in that so far. And I'm not saying it's never going to happen. I'm expecting I'm always expecting to have a conversation with a sound designer or with a post-production studio manager and get a little bit of questioning around what we do. I'm always expecting this. And then I never do. The only times that I've gotten the questions have been from people who don't actually work in sound. It's like, does this actually replace some sense of like, well, no, it should not replace and it's not meant for that. But whenever I talk to a sound designer, they absolutely get where in their workflow that particular product sits. And it's quite exciting for them. I think when we come to like projects and budgets, so there's a particular budget that and, you know, so you're a producer, so that it's kind of assigned a certain percentage that is kind of the threshold and it will vary from project to project, depending on how much you have. But there's a particular budget that is assigned to sound, which is quite small. Right. It's always been historically small. I think in the UK, we are an average five percent or less than that, something like that. And it varies from country to country. But the thing is the work costs that that's the cost of it. Right. So if you have if you are in the independent cinema and you have a budget of twenty five thousand pounds, so in this case in the UK for sound, the fact that you're reducing some of your hours does not mean that your work is less valuable. So studios and freelancers will still and should still be charging the same thing, which is what the rates are basically for the work that is done. And one of the things that we're also looking at changing is how we price things, because particularly sound people have kind of in a wrong way beeg seen as they’re charging per hour or per day. But in reality, the work is it's creative. So how do you actually put a price tag into that? So what we're looking to do is basically allow them to have more time to tailor and use that budgeted, you know, that particular budgeted time or budget to focus on the artistic things. The fact that they might spend a little bit less time doing the boring stuff, that they shouldn't be spending that much time anyway, it will only increase their profits because at the minute the profitability rate is really small. So in average, we've actually gone and found that it's sometimes less than 10 percent when studios are very lucky. So that's not good unless you're in Hollywood. You might have obviously have higher rates and all of these things. But outside the rest of the world is quite hard for an audio post-production studio or for a sound designer to actually go above that kind of threshold of profit. So we need to increase that because the particular budgets don't actually change, but the amount of work that needs to be done that it's not creative, that can change. And that does change the narrative and the sustainability, financial sustainability for sound designers around. So I think that's the focus that we're going with and that's and we're quite transparent about it. I think the other thing that in general with all of our sort of all of our products and all the things that the three of us are doing and everyone in creative industry is the thing that concerns me particularly. It's when basically non creative companies, kind of tech only companies are the ones that are left with the job of developing kind of AI technologies for creatives. That's the thing that concerns to me the most, because obviously the motivations are very different. I mean, I'm not saying that tech companies are the devil by any means. They push the boundaries as well. They bring things that all of us can use in our technologies as well and recycle and sort of bring into kind of push forward. But you need that creative vision. You need that person who's been there and done that like you do in most industries so that you can actually tackle the problems that creatives need solving. And I think that's probably the key. That's the thing that concerns me when there's no when there are no creative companies, when there's no Freddy's or no Zoe's around that, I'm going to be really concerned.

00:30:16: JOHANNA: Yeah. So there's something here also about like changing the value of the time that's put in. Like you can reduce burnouts on set by like taking away stuff like the unnecessary administration. You can get so much more value out of your sound designer if they can actually use the time for creative work and so on. And I guess it's the same for you, Frederik. But I thought this was a really interesting point of concern. And I wanted to ask the others the same question. Is there something that concerns you about the eye in in film, let's say, or more broadly, Frederik?

00:30:49: FREDERIK: Of course, there's a lot of concerns. I'm actually a designer myself. I don't know if I told, but the whole project started from my master thesis. So I studied design before and after that, I was working in a design agency for a year and it was right when AI and MidJourney, etc. came up. So, yeah, there was a lot of concerns. I was afraid of not keeping up with the time and losing my job and losing what I love to do. And yeah, in the beginning, there was like this this huge fear. But then I slowly realized, OK, actually, it's kind of fun to play with these tools and they can help you a lot. And still, there has to be someone sitting behind those tools who is actually capable and has a sense for aesthetics or has a sense for his handcraft, as it is with Anna, for example, also with Zoe. So someone who actually knows the stuff needs to use those tools, otherwise they are not that worthy, I think. So, of course, there's always a little scepticism from my side and a little fear. But actually, I like also the enabling part of AI. So it's just fun to play with the tools and push the boundaries here.

00:32:05: JOHANNA: And Zoe, what are your concerns?

00:32:07: ZOE: I think, like to Anna's point, what you mentioned is really important that there are people who are actually from these industries involved in creating the tech for the industries, because then there's just like a sensitivity and an understanding of the nuances of the industry and how it works. So I think that's a major concern, it being out of the control of the industry and it being just brutal capitalism. And I think that's one of the things that, for me, I take into consideration where this needs to be a building tool. And as I said, I'm part of my target audience. This tool needs to be able to make it more efficient for us to make films as opposed to create less jobs. And I think on that, particularly on job creation, like film is one of the... You hire a ridiculous number of people and those are families that are fed through the work that we do. And so one of the major areas that for me I was particular about was agents, because currently a lot of people get hired through their agents and it was like, oh, is this app going to now replace agents? And of course, that's definitely not what this app does. It makes an agent's job more efficient. There's still elements that need the human touch, you know. There's still elements where you need to pick up the phone and actually have a conversation with a producer before you send out your crew to certain places. When there's now an issue on set, an app is not going to resolve the problem. It's going to be a conversation between your line producer and the agent. So absolutely not replacing human jobs in that way where there's just certain things that need to happen between human engagement. But it also is giving people who didn't have the ability, you know, agents are also selective and only take people who are now at a certain level or a certain level experience. And they do that because they don't have capacity to take on more and do well for their clients, right? So first thing that it is actually doing is expanding the number of clients that an agent can have. So now you can actually manage a lot more people because the admin is less. But it also assists those who don't have an agent, can't get an agent and still need to get work, because that's the chicken and egg situation. It's like, OK, you need work, so get an agent. Then the agent's like, oh, but you haven't worked, so I can't take you on. So it's really resolving that aspect in terms of for me, job creation is a big thing. And so I said, like what Anna said, it's really important that it's people who I know, these people who are who are using the app and what's important to them.

00:34:31: JOHANNA: Yeah. And I think it's an interesting aspect here is that, like, when we look at just AI more broadly in society, any kind of application where it's used to categorize people or job applications or something like it, it tends to be terrible right now. Like the outcomes are terrible when you're using AI to sort people because it's unintelligently applied, right? Like it's based on it becomes with a lot of built in built in bias and so on. This is kind of the opposite, because this is people opting in to offer data points that are relevant in this context and connected with human recommendations. So that you know what somebody what it actually means that somebody's worked with someone or you know, who is this reference that you can contact and so on. So it's kind of the opposite use case of that, which I think is a really good reminder that like we can't be categorical about, like, for instance, AI and job search is bad. Like, no, like depends on how you use it.

00:35:23: ZOE: The other concern for me is not only people who are creators, but also people of color and diverse people, women in tech. It's really important that the aspects that like, for example, I'm thinking of certain things because of who I am. And you can't expect literally like a white cisgender male to think of these things. So it's when these apps are being developed, when these products are being created, all of those nuances need there needs to be representation in the room when that's actually being created so that I also then we teach it how to think. We teach it. We code it with specifics in mind. And so I think that's one of the things that is a concern for me, that I just don't see enough people who look like me in the spaces at all, which means a lot of the things are not like basically a specific demographic is just not thought of or considered. And that becomes really dangerous. Yeah.

00:36:15: ANNA: Yeah, I absolutely agree with this. And I was I was thinking so many things when you were talking Zoe as well. But yeah, I mean, kind of very selfishly, my motivation with building a tool was basically thinking about all the people like me that were struggling, like you said, to break into the industry that maybe where you were talking about, like particularly women. You know, I've been in so many productions, I did production sound as well for many years. And that was always a conversation. It was very hard to get a job unless you started sort of saying, hey, recommend me for stuff like I can do it, too. So it's quite hard to do it. But I think the human kind of aspect within all of these things, like you very well said, we're all solving quite specific problems. But it's quite interesting that in pretty much the three of our kind of sectors or the areas of focus for each one of us, there's not been a lot of development or embracement of technology prior to this. So, for example, like Zoe was saying, you know, people using spreadsheets or like sometimes doing it by hand. And Freddy was also saying, you know, it's not been a lot of innovation in a long time and subtitles. And in sound, there's been innovation, but it's always very careful, very kind of stifled, very much like, OK, it should look like the consoles from the 70s, guys. Remember that it should look like that. And so why is that? Is that coming out of fear of us basically losing our jobs? And probably the answer I know for myself, I cannot speak for your particular sectors, but I know that it probably does. So how do we use all of these innovations, what we're doing and what, you know, hundreds of other creators are developing, the apps that they're creating to sort of kind of subvert the industry a little bit and say, hey, no, actually, again, going back to my point and the point that you were having Zoe as well of the work should be valued for what the actual benefit that it brings to a production or to the post production stage or to the kind of audiences like in your case, Freddy, that their experience of films, of TV, that's where the value is. So the work shouldn't be just, you know, how many hours you put in it. And so that's one of the things where I think if we had just enough products, just a good amount of products that solve very specific issues like this, then that in itself organically can start sort of changing and revolutionizing our industries as a whole.

00:38:23: JOHANNA: Is there anything specific as you're following like the what's happening in AI in general? Is there something you're excited about right now? Some breakthrough or product or development that makes you feel like this is this is a cool field to be in?

00:38:36: ANNA: I'm going to be super and I want to say what Freddy and Zoe are doing because I mean, there's definitely lots of really exciting things, but I'm going to be cheeky and say that because I've actually experienced both things. So, I mean, I have a soft spot for Freddie's product because it basically analyzes sound and I'm like, yeah, I like that. You know, and it sort of brings all of these elements which are often very kind of in the back of our mind and it sort of makes them graphic and makes them kind of in a language that we all understand. And so it makes you conscious about it and experience them in different ways. I love that. But then I've also been a production sound mixer, so I also have a soft spot for Wrapped because I'm thinking I really wish that this was around when I was actively doing production sound. So I'm going to be cheeky and say that.

00:39:23: ZOE: I'm going to ride on what Anna said again and say like literally as she was speaking about it, I'm just thinking right now, I'm like, I need Black Goblin, like because, you know, we're working on specific projects. And obviously, as Anna said, budgets are always really tight and you can get a freelancer who says, listen, I want to do the job, but it's actually going to just take like I can't beat the amount of time it's going to take me. So I can, you know, when you're negotiating, it's like, oh, I can reduce my rate, but it's actually just going to take me three weeks. And there's nothing I can do about that. I'm going to have something to say. Have you tried Black Goblin? So definitely like that's one of the things that excites me where it's like there is this barrier where it's just a time consuming thing and people would be willing to make more like a lot of film, a lot more films could get made, you know, if there wasn't these barriers to specific things. So it's like, OK, now I can't get like really good sound design because it's just going to cost too much because of the amount of time that can just be compressed. And then you can actually get someone who works for you and rather than someone who works for the budget. And with Phont, we've already been having discussions on our films, you know, and which ones that it would work for, which I think, yeah, that's also another really cool product that I'm definitely going to use as a producer. There is something that I've recently seen and I've been following. It's I think it's I want to say Sora, that is doing visualization from script. I think it's text to screen, basically visualization. I'm quite excited about. Yes, that is definitely an anxiety point. You know, like, OK, so are we not just going to be watching like non-real people and all of that? But you know what I mean? Like I know that there's like especially as a director, there's an emotion that is carried, that's on set when an actor is performing. And it's an essence that leaves them when they perform something, especially like a very intense scene, which I don't want to say, but I feel like I don't think that can be captured unless we're going to have like some spiritual capture. But I still believe that there's elements that have to have to have that human character, because ultimately we are viewers. And what do we respond to? Other humans. That's why we experience emotion when you sit and you look at someone in the eyes and they have tears is you are a human and you're connecting with another human. That's what you laugh or your yawn. You know what I mean? Because we are we're reacting to ourselves.

00:42:00: JOHANNA: So I think so anxious and excited about this.

00:00:03: ZOE: OK, yes, so I'm saying that part, but the part that I am excited about is there's certain things that just people wouldn't be able to do. There was certain genres that people would like a lot of companies would just immediately turn you down. We're not going to give you a budget of that to do a sci-fi film. It just costs too much. You need to have done at least three or four more feature films then come back. You know, even if the idea is brilliant and now that's changed because the budget amounts aren't as high, the risk isn't as high because a lot of the stuff can be done by AI. So it's going to allow filmmakers who previously would have been given a no or not now. It gives them the now option. And I have a project that I thought of like almost seven years ago. It's got aspects of like being shot from a blind, a partially blind person's POV. And I always, you know, I kept playing around with it on how to visualize how to get this on screen, how to get that experience on screen. And like now with AI, and I've met a couple of AI filmmakers and we're having conversations. It's just completely different. Like what I can do. It's now about what how I want it to look and I can create it as I want it. And so that's what's exciting. That's the exciting part.

00:43:14: JOHANNA: That's great. What about you, Frederik? What's exciting for you right now?

00:43:18: FREDERIK: I would also be cheeky and join the other Startups and talk about how cool they are. I really love both of the products. And yeah, apart from that, if I look at the industry, it's actually crazy evolution we've had in the last two years. You can you can't really feel it or touch it. It's just so quickly evolving and it's really hard to handle all of these tools and to look at it. And at the same time, with this rapid development, the ethics, it's really hard to grasp for me. So there's a lot of cool tools, obviously, I like. And it's especially in the everyday day workflow as a startup with ChatGPT and all of the AI note takers and calls. Notion as an organization tool with all of the implementations you get there. It's really like the workflow is much, much easier. That's a special part for me. Apart from that, I don't use too much AI for the creative process, actually. But yeah, I mean, the ethics are a huge part we have to take care about. Also at our startup. I mean, the good part is so far, we don't really take jobs away. It's more an enabling tool and people will be able to start a new creative process on top. And that's another level. But what we really have to take care about is biases, for example, if we do analysis of audio and there's different cultures, most of the algorithms are trained on Western cultures and languages. So that's where we have to take care that it's bias free. Also, for example, a common bias with audio analysis with emotions is that women, if they speak up, if they are a little louder, they usually interpreted by AI as more aggressive than they are. And these are like stuff that we really need to focus on, that we get this away and we have a bias free and ethical algorithm.

00:45:24: JOHANNA: I got to say, as we're running out of time, that you have all touched on something here at the end that I think is super important and exciting to talk about. And I just want to spell that out, which is that often in the industry, we talk about what can we do in the established kinds of filmmaking. But you've all touched on this, what it does for people who aren't established yet or for production from small countries like Finland, where I'm from, where certain genres haven't been available to us or you come from developing economies or there are countries certainly in Europe, for instance, which have very small public funds and just the markets are too small to carry certain kinds of filmmaking. And the fact that if you're a content creator and you can use font to add new layers of storytelling to your content and make it more accessible or make it more vivid, or you're a sound designer early in your career or you're working with filmmakers with very small budgets and you don't have to shell out fortunes for certain softwares and sound libraries, it's possible to get to a higher level of work just earlier. And the same thing with making production more efficient, because that's work that needs doing anyway. So if people aren't getting paid for that work, they're just going to do it as free overtime. And that's not sustainable. And making it more possible to work more sustainably early in your career would be a huge game changer for this industry. So thank you all so much for your projects. I do have a final question for you all, which is this the application deadline for the EFM Startups program is coming up. Would you have any advice for people who are applying for it or who will end up participating in the program?

00:47:05: ZOE: I would say apply, like do it. There was a moment for me where I felt like, oh, maybe I wasn't ready because I was a very early startup. But I made the application and literally it's propelled my project so much, like so incredibly the amount of networks and connections I've made. From being at EFM itself, but also the cohort has been, you know, being a founder, especially an early founder and someone who's not from a tech background, just having other people to ask like, guys, is this normal? You know what I mean? Or, you know, how does this work? So I would say apply. That's my only advice. Make sure you do apply. Don't let anything convince you not to do it.

00:47:47: FREDERIK: Totally. Yeah, I'm totally with Zoe here. It was a great time. Just apply. The only thing I wouldn't do a second time is when you're going alone, maybe don't buy the Berlinale ticket because there's so much to do at the EFM itself. There's so many meetings, so many conversations I had that maybe I think I attended only two films. I mean, still worth it. I saw Wim Wenders and yeah, it was great to see him live. And so that was worth it itself. But you're not going to see many movies, to be honest, during the program.

00:48:22: JOHANNA: And what about you, Anna? What's your advice?

00:48:25: ANNA: I mean, all of these things, I think for the application itself, yeah, definitely don't let anything get in your head and say don't apply. I think if you're solving a problem, it doesn't matter how big or small or specific it looks, if it's a problem that that you know exists. So basically just make sure that you have a very kind of well defined issue that you've at least tried a little bit. Doesn't matter how early, like I was quite early stage, even though my company's been around, we were a different thing before. So it doesn't matter as long as you've identified a proper problem than it's solving in our industry, then that makes your application stronger than you probably think. That was definitely the case with me as well. I thought, oh, no, this is just, you know, it's not going to happen. But then it did. So it was and it was really, really good, I think. Yeah. And use the team a lot if you're lucky enough to get into next year's cohort, then use the team a lot because they're really, really good. They know their stuff. They know they have so many connections and a big network. So if you kind of open with them and as you know, I'm looking, for example, I'm looking for some designers, I'm looking for people in the city or they'll probably help you arrange meeting. And I think even if you apply and fingers crossed you get it. But if you don't get it and just go to the Berlinale is really cool. There's people and even just interacting with, you know, other startups, people on the on the EFM. It's it's quite useful. You get to meet a lot of people who are really relevant and know their stuff and that in itself is it's an amazing experience.

00:49:52: JOHANNA: That's a wonderful thing to hear. Thank you so much for your time. Frederik Merkel, Zoe Ramushu and Anna Betancourt, thank you so much for being here. That was all for today. Industry Insights is produced in cooperation with the Goethe Institut and co-funded by Creative Europe MEDIA. The deadline for applying to the EFM Startups program is January 3rd. So be sure to send in your application by then. And please keep tuning in to Industry Insights. Find us wherever you get your podcasts and on the website of the European Film Market www.efm-berlinale.de. Thanks for tuning in. Goodbye.