AI is rapidly transforming the global media and entertainment industry, with a market size projected to reach USD 99.48 billion by 2030¹. As AI continues to revolutionize creative workflows and enhance productivity, companies are racing to implement AI-powered tools to stay ahead of the curve. However, as they embrace this cutting-edge technology, enterprises must navigate the crucial challenges of ethical and safe implementation, as well as robust IP protection. In this episode, our experts explore the transformative impact of AI on the media industry and creatives, all while underscoring the importance of maintaining human creativity. Featured experts Gabe Michael, Vice President, Executive Producer – AI, Edelman Harry Leshner, Consult Partner, Telecommunications, Media and Technology, Kyndryl 1 Source: “Global AI In Media & Entertainment Market Size, Share & Trends Analysis 2023-2030: Market to Soar at 26% CAGR, Reaching $99.48 Billion by 2030” GlobeNewswire, 2024
AI is rapidly transforming the global media and entertainment industry, with a market size projected to reach US$99.48 billion by 2030¹. As AI continues to revolutionize creative workflows and enhance productivity, companies are racing to implement AI-powered tools. However, as they embrace this cutting-edge technology, enterprises must navigate the crucial challenges of ethical and safe implementation, as well as robust IP protection.
In this episode, our experts explore the transformative impact of AI on the media industry and creatives, all while underscoring the importance of maintaining human creativity.
Featured experts
Gabe Michael, Vice President, Executive Producer – AI, Edelman
Harry Leshner, Consult Partner, Telecommunications, Media and Technology, Kyndryl
1 Source: “Global AI In Media & Entertainment Market Size, Share & Trends Analysis 2023-2030: Market to Soar at 26% CAGR, Reaching $99.48 Billion by 2030” GlobeNewswire, 2024
Tom Rourke 00:03
The rapidly evolving pace of AI and generative AI in particular has the possibility of transforming many industries. Media in particular has given rise to both opportunity, the ability to accelerate previous production processes, but also concerns for the future of truly creative roles. What does it mean to be someone in media, in a creative space, when so much work is being conducted by generative AI? But also in terms of the impact on the ultimate product created, do we get to a point of homogenization? And what's the role of genuine creativity in the application of AI in the media industry? I'm delighted to be joined here today by Gabe Michael, who's Vice President and Executive Producer for AI at Edelman, and for my colleague Harry leshner, a Consult partner in telecommunications, media and technology here at Kyndryl. Gabe, Harry, you're both very welcome to The Progress Report.
Gabe Michael 00:57
Thank you. Thanks for having us.
Harry Leshner 00:59
Thanks for having us.
Tom Rourke 01:01
The impacts on AI in the media and entertainment business seems to be one of those places where there's tremendous opportunity, but also very significant challenges. I wonder, maybe, Gabe, if you wouldn't mind give us a little bit of an overview of the landscape of how AI is showing up in the industry.
Gabe Michael 01:17
It changes by the hour, it seems. And actually that's part of my role to just vet and be aware and be open to all the new tech that's coming in. The 3000 foot view is that AI is here. AI is here to stay, and it's going to be in every part of our business. And for me, that touches a lot of creative and creative production as well. And so we are trying to not only vet tools, see what's the best out there, but also actively try to work them into creative workflows.
Tom Rourke 01:51
And I guess, Harry, you work with a broad range of organizations within the industry. What are you seeing in terms of the key trends here?
Harry Leshner 01:58
Really what my background experience is going through the full adoption life cycle with large enterprises. And it kind of follows this three step approach, where it's like step one is you have those early adopters in an organization who are playing with tools and doing a lot of experimentation. Then phase two is where the enterprise really starts to come up with their strategy for how they're going to adopt more holistically and have governance and guardrails in place in that. And then phase three is it really just becomes kind of a normal part of the business. What I tell people about AI, because everybody approaches it like it's a brand new thing, I have the joke that we've actually seen this movie before with other patterns of technology adoption. And it's still kind of early, early days when you think of the large enterprise adoption, we're still kind of more in that first phase where there's a lot of experimentation, and now the conversations that I have with leaders and organizations are really about, "Okay, how do we take it into that next phase and make this something that we can adopt at the enterprise level?"
Tom Rourke 03:04
You talk about how people draw comparisons with other sort of technical waves. And I think there's always people who will put themselves in a position of, there are certain organizations who want to be at the leading edge. There are certain people who will deliberately decide that they want to be fast followers. But my sense is, like AI, feels somehow different. It always feels that there is an extra tension here that says, "Well, actually fast following may not be fast enough." And I'm interested, Gabe, how would you describe the stance in the industry as you see it? Are people leaning into being leading edge, or are people saying largely cautious and letting other people experiment? What's your perception of the landscape there?
Gabe Michael 03:40
I think it depends on what part of the industry is looking at it. I kind of have a split point of view. I'm a creator and filmmaker myself, and then also an executive trying to bring on these tools within the enterprise space. I look at a lot of the tools that are out there, whether they're safe yet or not, to see what is possible, right? I look to open source tools into that space to say, "Okay, this is what's coming on the horizon." But what has remained true is the process within filmmaking, within creative, it's still the same. It's that we're trying to figure out how to take these tools and shape them into our workflows with the ways that we know how to work. But what we're finding is that we still don't know how we can use in the enterprise space legally because of the way it's generated. But we also need to move at the speed of enterprise in order to look at the tools in the right light, make sure that we have the right type of protections in place, and then make sure that we're protected and that our clients are protected as well. So it's a delicate balance of looking at the new amazing tool that's out there, the new amazing feature, versus what we can and what we should be doing on an everyday basis.
Harry Leshner 05:00
Even if you're an organization that is being more conservative when it comes to Gen AI adoption, be realistic, there are people in your organization who are using these tools on a regular basis, whether you know it or like it or not. This technology is really accessible in a way that other technologies weren't. I mean, cloud computing was really accessible because you didn't have to call up your IT department to set up a bunch of hardware for you. So we saw people get their hands dirty with that way before the enterprise caught up. But with Gen AI, you give it a prompt, it can create an image or or a video or a piece of music for you, it's just so tangible and so easy to get your hands around that everybody is just really excited to jump in. Maybe the hype around it is higher than previous waves of tech adoption, and maybe it's going to make things move a little bit faster.
Tom Rourke 05:21
I was going to make an analogy which was absolutely going to date me on this call. I can tell you time wise, but I do remember a period when, if people even remember the term desktop publishing, but there was a wave when people would start to go do things that they previously would have gotten professional services for. But actually, it didn't necessarily mean they did good things. Like essentially there was a swing where people start doing things themselves, and they realize, actually, you need the certain skill to this. Are there analysis processes, you think, in your industry Gabe where, because people will have these tools to hand, they will start trying to do things themselves, rather than kind of going to professional services firms or creative firms to do it. But then the pendulum swings back that essentially, good work is still good work, and that it isn't just about the tools. Or is it just too early to say how those dynamics are playing out in the industry?
Gabe Michael 06:42
I think a lot of the problems with the industry are coming to light. Not everybody is a pro at this, but it is going to give people that didn't have this the time to learn the skill a voice. It'll give them an opportunity to be able to communicate something in a different way, right? So, what we say is it's opportunity to do more, and we're going to be expected to do more. Our clients are going to expect more from us because we have these capabilities and because we are the pros. But what's going to really come back, and I know that a lot of people are saying this, is the taste argument, right? Like, because you can generate these things, and a lot of people have the same ability to look at an art style and say, "I want to replicate that for this campaign, due to XYZ." But when it comes to actually executing it, are we at a point where we can go and then photo edit this particular thing, and is that give the potential of taking away certain creative jobs? It can for some organizations. I think within a more professional one where certain people have certain roles, that those roles are going to change a little bit, maybe consolidate some of those roles into two or three different roles into to one. But can that replicate an entire studio? Not right now. What we're working on is, how do we take this technology and how do we use the people that we have, and how do we create processes and creative workflows that enhance these tools but also use the best of our abilities? And we have to do it in a way that is both professional and the client is used to right now. So will that process change? Potentially. But what we will then be hired for is our creative, right? The creative chops that we have, our creative choices that we make, and the tech is going to be worked around what that creative is, what those those general ideas are.
Harry Leshner 06:42
Yeah, I mean, it's changing the way that work gets done, it's also changing where the work can be done. And then if I'm going to be optimistic, I will hope that there will always be a premium placed on creativity.
Gabe Michael 06:42
And it's interesting, because what I'm seeing, I work with, internally, so many amazing creatives, but I also work with the broader AI community, and I'm a part of a lot of these groups, and I just see extreme excitement and then fear, right? Extreme excitement and then fear of this. And we're still trying, everybody's still trying, to wrap their head around this and where it's going, but we kind of just have to take it and ride with it right now.
Tom Rourke 09:17
Yeah, that anxiety is definitely something we see in the design community, and also, particularly, how are skills going to be formed? How do people develop their skills and develop their art and develop their practice, when so many of the things they would have done to do that are now just being essentially automated out of the whole all of the work?
Harry Leshner 09:35
Well, yeah, I mean, there's certainly a lot of anxiety in the industry about jobs going away altogether. The comment that I made about where the work can get performed, you know, if you're suddenly able to not have to be on location somewhere, right? You know that changes things. And then just to the comment about creativity, there's something inherent about using generative AI, where the output is derivative, right? It's trained based off of what you fed into it. And so I'll be careful what I say here. You know, there's arguments that can be made about how original certain content is made today, but if you're using generative AI more than humans, then you know it's fundamentally derivative, and so there's a concern about overall quality with that.
Gabe Michael 10:26
I 100% agree. If you're looking at it on a one-to-one basis and if you are saying, "I'm going to sit here and prompt, I want to Studio Ghibli type shot, and it's going to look like this." It's going to look like something that has already existed, right? But there are now workflows where it's like, we're using Gen AI especially for video, as a generative engine. It's like the backbone behind all of it, right? So we can come in with custom-created backgrounds, and we can really control every step of the process. If you look at it that way, if you are looking at it as it's a magic button, or I'm gonna type in a sentence and get the final output and never touch it again, then it's 100% derivative, right? But if you're using the tech as that engine, then it becomes specifically for video. It becomes animation, just in a new form. You know, you're taking away a period of learning for people, and that is the thing that I don't know how it's going to shake out, or how people are going to adapt. But, for us, we get those types of questions daily, right? So sometimes, it's like, "Well, why can't I just build this deck myself using a certain generative tool?" It's like, you can totally do that. But is that going to replace the creative team that has been doing this for 20 years? No, we have the experience.
Tom Rourke 11:43
The other thing I'm quite interested in is around, does it lead to a sort of an even greater kind of fragmentation of essentially the end product? Does it lead to, then a point where, actually the market fragments into we have a greater tolerance for derivative content, but then there's a greater premium on things that really stand out at a different level?
Harry Leshner 12:01
I don't know if we know the answer to that. I mean, you're already seeing today, there's a massive increase in the content that's being developed, and it's not just, you know, at the studio level, right? We all are subscribed to four or five different streaming platforms. But then think about all of the content that's being put out on YouTube and Tiktok, because people have the tools now. You know, an individual sitting in their studio apartment now has the tools where they can create stuff that looks pretty good. I would just expect that trend to continue where there's more and more content being created. I do think there's going to be a premium placed for stuff that's original and creative.
Tom Rourke 12:46
So what are the other pitfalls, and how maybe can organizations protect themselves? Because, again, it's actually been generally a positive conversation we've been having about exploration and experimentation and growth, right? But I am conscious there are dangers, but equally, is it that people are having to just accept living with a higher level of risk around this stuff until the path becomes clearer? Or are there things people can be doing today to take a safer path forward?
Gabe Michael 13:10
There's definitely things that you can do today, and we're doing that at Edelman. We're looking at the terms and services of every model, every new tool that comes off, and we're really looking into what type of protections and identification that people offer. Every model is getting better, right? So we have to find a solution to bring on tools, and we have to start at the most basic, safe level. Once you kind of have and you figure it out within your organization what your baseline comfort is with the tools, I think you need to have conversations with legal teams, with business affairs teams, and seewhat is your "cans" and "can'ts", right? There is some risk here, if we just have somebody type in, "I'm making a Steven Spielberg film and making it look like 'E.T'." Putting in these unethical prompts, as opposed to somebody coming with a designed data set of like building your own custom model, right? I think a lot of the red flags are companies that don't offer any sort of identification. They don't stand behind their models at all, where they say it's the risk is completely on you. So you have to say, "Okay, well, if the risk is putting on us, do we either find a solution to mitigate that risk and make sure that we're doing things in a more ethical way? Or do we tell our client, 'You really want us to do this thing? This is the only solution that we have to do it, and there are these risks associated with it. Do you want to take on that risk? Do you want us to take on that risk?''And you have to have those conversations.
Harry Leshner 14:49
This is really kind of a key way in how Kyndryl helps our customers adopt Generative AI safely. It's developing all of the policies and onboarding the models. Also, setting rules around ethical prompts. Can it use profanity in what it gives back? Yes or no? Is this safe going to be safe for children? Yes or no? But then also keeping up to date on that and tracking that is a huge undertaking and we leverage technology to do that. But as the policies on the provider get updated actually having a mechanism to track that and update what you're using. There's also a cost aspect of it. These models can get quite expensive if you run them a lot, and so we have mechanisms to track the cost and manage budgets. The questions that I kick around with our clients all the time is, what's the ROI? Because, just because something is fun and it produces a cool output, there's an assumption that there's a number of man hours being saved in using generative AI. But, I mean, it sounds terrible, but that doesn't necessarily yield the kind of business results given the amount of investment that it may take, which I think is a reason why, at least at the large company level, a lot of the focus in making progress on the generative AI strategy to date has been a lot more of the operational workflows that are really just kind of direct, taking cost out.
Tom Rourke 16:26
We always like to kind of get to the end about thinking about, well, what does progress look like? And we all have different views, that is. Maybe, if I start with you, Harry, if you look forward, you're immersed in this, what within the media industry would you view as like progress when you think about the industry and the application of AI over the next number of years?
Harry Leshner 16:44
I think when I talk to people who are new to this, there's an idea that you're going to give chat GPT a prompt, and it's just going to create a movie or feature length film for you. And, you know, I think in the conversation, that's really not the intent here. I don't know if that's ever going to be even something that we want. So I think about it as things that are being done well today and are producing business results in terms of like ideation and story development. I think that chat bots can be really, really great. You could give it a prompt that says, especially in a series, "Tell me everything about this character and kind of summarize it for me." That's really useful in story development. I think that works pretty well today. Anything around transcribing videos, you know, you used to have to hire people, you know, how much did you pay them per hour to watch a video or listen to a recording and type it out so that you could use it for story development or maybe editing later on? Gen AI does a pretty good job of that now. Quality control, right? Watching a video, looking for dead pixels, or things like that. AI can do a pretty good job of that. That's kind of the first tranche. I think the second tranche is really things that are pretty good today, probably could use some work. I think, you know, in two to three years, they'll be kind of where the first set is, and then maybe AI assisted VFX, B roll generation, set development and extensions, AI assisted script writing, stuff that you can do, but I don't know how great it is yet, and I think it'll be there in two to three years. And then the last tranche is really five or more years out. Don't know if we're ever really gonna see those like synthetic actors, right? Or you have no actors, no sets, and you're producing things, maybe with a series of prompts, or, you know, creating a whole scene from scratch. I think that stuff is more kind of science fiction at this point.
Gabe Michael 18:54
I think, you know, everybody's first assumption is automation, like it's gonna be a completely automated process, and we're doing some really great research and some great work on our AI product team with that at Edelman, specifically for our clients, and then also internally, but like in a creative and production setting, until we get to the point where it's like a movie that's being generated on the fly, right? Which is its own thing, right? Like, that's going to be its own type of content. You have to look at things in buckets. And I think you were saying exactly this. I look at it as AI with productivity, using it as a producer to help me find locations, help me vet these things that would just be, you know, a team of two or three other people that were doing it before. And then we have our bucket of AI added on to things that have already existed in pipelines that already exist, right? And then the third bucket is what I'm I'm calling AI animation, not necessarily things that look like AI, but AI as the driver for it. So it's like a lot of the things that I've made, like a short film made completely out of the majority of AI tools. But, you know, we can't put all of those three in the same bucket. There's productivity, there's visual effects and that tool set. And then there's complete workflows. It's not a one prop thing. It's creating a movie shot by shot, just like I would be if I was animating it or if I was editing it. What goes away there is the first step of physical production, but that's not going to even go away. There are things and opportunities that come up across the industry. Actors have the opportunity where they do one or two photo shoots to create a data set for a model, and then they have a licensed digital file that they pay for the usage rights of, so that could go into still photography, which then if anything goes into still generation, it can become video. And I think that there's like a massive opportunity there on the commercial side and on the social side for voice actors as well. And we have to then prepare ourselves for the next evolution of this, where people have these already. So I'm going to be hiring from a marketplace, I like this person, this person, this person. I'm going to pay them XYZ, and I'm going to download their digital file in a safe environment. We've seen what's happening with Nvidia, with their Omniverse, right? We have the opportunity in a 3d environment to have a pixel perfect object, and we're going to be able to just take that and a couple of, you know, scenes that have been created, and then turn it over to the creative team. So all this is possible. The thing that is going to be the most important will still be that creative idea, right? The execution will be expected, but the creative idea will still be the king.
Tom Rourke 21:41
Gabe, Harry, thank you so much for a fascinating discussion to the afternoon we could easily have spent twice the time we've allocated. And I think you've highlighted some of the real concerns here about maintaining that tension between rapid adoption, as we want to try and embrace the productivity benefits, but also the sort of ethical considerations and the implications for the formation of skills and roles in the future, and in particular, how do we maintain the human in the process, genuine human creativity leading to a clearly higher quality product? I hope you found today's discussion as enthralling as I did. To stay in touch with future discussions of The Progress Report, please do, like, share and subscribe. Thank you for listening to The Progress Report.