RealmIQ: SESSIONS with ANDY BEACH

In this podcast episode of RealmIQ: SESSIONS we feature Andy Beach, a technology strategist, advisor, and author, specializing in AI, media, and digital transformation. As the former CTO of Media & Entertainment at Microsoft, Beach shares insights into AI's impact across industries, including media, sports, entertainment, and music. The discussion covers AI-driven workflows, copyright concerns, ethical AI, and how AI is reshaping production, content distribution, and consumer interaction. 

Topics Discussed

• AI’s impact on media, entertainment, and sports
• AI’s evolution in content production, distribution, and monetization
• Hollywood’s slow adoption of AI and governance concerns
• Ethical AI, copyright challenges, and content provenance
• The AI "bubble" and rapid industry disruption
• Branding and advertising in an AI-driven landscape
• Deepfakes, digital twins, and AI-generated avatars
• AI in music, including estate management and virtual voices
• China’s role in AI development and global competition
• AI’s influence on streaming and content discovery
• The role of AI governance in media organizations

Quotes 

"People were highly unaware of where they were already using machine learning. AI has been in Netflix recommendations, closed captions, and voice assistants for years."
– Andy Beach

"Hollywood has been slow to adopt AI, but the thaw is happening at the studio level."
– Andy Beach

"The AI adoption bubble isn’t bursting anytime soon. People will continue integrating AI into their workflows, businesses, and creativity."
– Andy Beach

"AI will democratize creativity, but the people who push its boundaries will stand out."
– Andy Beach

"The debate over AI’s impact on copyright and provenance is just beginning. The EU is ahead of the U.S. on AI policy."
– Andy Beach

"We are not done with AI’s innovation bubble. The technology is evolving so fast that what’s done today will look quaint tomorrow."
– Andy Beach

"More content isn’t the issue—content discovery is. AI will have to fix that broken experience."
– Andy Beach

"The studios and music industry are moving cautiously with AI, but independent creators are advancing faster."
– Andy Beach

"AI could personalize our streaming experience, but right now, it’s just chaos. Someone needs to fix it." – Curt Doty

"We’ve entered the era of digital twins. AI can bring back artists, but do we really want Elvis selling us hamburgers?" – Curt Doty

"What’s ironic is that billions are being poured into AI, but somehow, they can’t spare funds for licensing fees? If this can be solved with money, there’s no shortage of money." – Curt Doty

"Social media agencies and creative firms are in panic mode. There’s fear, but also opportunity. If you don’t adopt AI, you’ll be replaced by someone who does." – Curt Doty

Transcript

Who doesn't want to save money on their movie production budget? Both on the producer side as well as the studio side. But what are, what are some of the signs you're seeing in, in discussions with the Hollywood players? Yeah, so I. I think there, there's always a little bit of misconception because, and, and again, I'll, I'll go back to sort of generative AI's launch being sort of this trigger point.

There was tons of AI use that's been going on for years, a decade or more in, in, in Hollywood, around, around using AI in, in areas of both production and, and how they manage archives when you're dealing with that level of content. At scale, you're, you're going to need some level of computer assistance in, in the work that you're doing.

But they, they of course, got worried and there were lots of   early problems around things like provenance and, and, and IP issues.  So, they rightly sort of slowed, slowed adoption down and, and held things in place. What I do think is, is telling is that we're starting to see. Media companies and, and studios in particular have effectively like an AI governance board, whether, whether they call it that or not.

Some, some literally call it that. Others have different names for it, whether it's their office of innovation or, or other areas or maybe just part of like a CTO's office. And what they're really doing is looking at the, the overall implications, PE people sort of from all the different walks of the.

Of the business I mean HR, legal production R&D and all being able to look at, what are the areas where they really want to advance AI and use it.  And then looking like a group at what those implications are, are there threats within that? Are there safe ways to, to go implement it? And then the super important one that I feel like gets left out a lot is how do you go educate.

The, the workforce. How do you show 'em how to use this and sort of maybe most importantly, how not to use it. Like what are the, what are the ways that are going to steer us into the wrong and, and the bad troubles? And what are the, what are the, the golden path? What is the way to use it? That is helps everybody be efficient, but still.

  checks the boxes that business as a studio needs around protecting their IP and, and all the other pieces that they go into it. So, I feel like that has started to happen. And of course, that takes a while to set up and get running. But there are studios that are starting to do that and they're running pilots to test out ideas and then beginning to, to shift those into production.

It's very different, obviously, in, in an independent world where you have less of that oversight and overhead and, and so you see movement quicker there. And there are even other industries where I feel like studios can pay attention and look at the, the work they're doing and learn mistakes that they might make along the way.

Advertising in particular, I feel like has been pretty quick to, to AI adoption because they have so many less concerns around things like union. With it. I think because they're working directly with brands that, that have IP, that that's where the concerns might, might be. But I think they're being more bullish about rolling it out into production and testing with it.

So, it gives you a couple of different places if you're in a studio environment that you can go learn and look and explore. And, and that's been what they've done for the last couple years. And, and hopefully the ones that are, are now moving into those. Larger tests around how they're going to, to use this we'll have some success and start rolling it out.

I think the biggest part though, for them to remember as they do that is that, again, the, a like I said earlier, AI is going to be changing so rapidly that those implementations are probably going to change just as rapidly. And what, you can do with AI today in a, in a production environment. We'll, we'll either change in six months to a year, or it'll be radically cheaper because the, the compute got so much more efficient in, in the way that it goes.

So, so it's a continual iterative process for the next several years I.  Yeah. What's done today? Will, will look quaint tomorrow. It'll be adorable. Absolutely. Oh, that was so cute. But if you think about it, the, and again, like if you, if you just think of the timescale now being so much more compressed, but if, if we think about like the, the innovations that went around things like cameras and nonlinear editors in the like nineties and two. 

I mean, when I was in school for film and video in, in the early nineties, like the, the, the lowest quality camera you would ever think about shooting something in was like a 250 or $300,000 camera that had probably another 150 to $300,000 worth of lenses on the, on the front of it. And even that wasn't going to be a film quality.

Yeah. Thing that was, that was just your, your, your broadcast productions that were going on. And you jump ahead 10 years, and you've now got nonlinear editing tools that you can run on consumer grade equipment. And you've got on your phone even, well, the, the smartphone was probably still another half a decade away at that point.

But, but well, I mean, there you can, but you had the red cameras by then and the, they were like a five or, I think at that point it was like a $15,000 device.  So, I mean, it was order of magnitudes cheaper, which meant it a lot more people were getting into that space and doing it in a way that they couldn't before.

Or you saw a lot of people who, had to go either work for a big studio or a, a big agency that could afford the, the hardware and the equipment, and they could go off and now and do it on their own.  Right. We're going to see the same kinds of things with ai, where, where it, it gets prohibitively cheaper or it gets much less expensive as, as we move forward with it.

And it, it, it ultimately democratizes the ability to go out and do these works. And again, that it's dependent on us making sure that we do it with ethical models and, and in a way that makes sure that everybody feels protected and, and safe in the way that we're using it.  Yeah. Back to the Madison Avenue aspect of industries that are benefiting and moving forward at a rapid pace.

I do believe because of the state of AI and video, which has mainly been short clips, kind of.  Linear non sequitur type of fantastic product promotion, right? That doesn't bear the burden of a Hollywood narrative, right? Where you have an actor and seeing consistency and continuity. It's like all those issues kind of disappear with Madison Avenue because they're just like, Hey, in 30 seconds we just want to titillate you with a

New shoe from Nike or a juicy hamburger from Wendy's or whatever. All, all these things are possible. With AI and I and I, it is causing disruption and kind of reorganization and rethinking, and you're seeing kind of the big conglomerates kind of shed some of their creative agencies either. Out of anticipation of that market collapsing, or they don't want to be in the, that creative agency business anymore.

And there's a lot of growing concern. As   from the, the visual effects side. Companies are going under technicolor, jellyfish among others, and   if those are going under, where does that leave? Creative agencies.   I, I came up through the ranks of production in a branding agency world wasn't exactly advertising.

Mm-hmm. And we were kind of the lark in that, in that world, but we were doing really cool shit by integrating typography with live action videos back in the early nineties when, kind of broadcast design was just starting, but.    it was the birth of motion design, right?

Absolutely. And, and, and so there's a lot of motion designers, a lot of production companies that have been serving the ad industry and the TV industry.  They're all freaking out and or regear and retraining, but a lot of disruption, a lot of fear.  Where do you see those types of businesses landing in a couple years?

So, I, I think a thing we hear commonly, and I, I do believe it is true, is that an AI won't replace you, but a person using AI will. And, and I think that bears out in, in a couple of different ways. The, the. The quality of output you get if you're classically trained and understand the, all of the elements that, that you're talking about, what you can then go do with an AI model is going to be far superior to, to what me or somebody off the street is going to do.

So, I, I think that that is a given. I, I always use the analogy of a guy in a basement in Pacoima. Yeah. 

You will suddenly potentially be competing in, in, in a broader field. There will be more people who can do what you're doing. And therefore, it, it's my hope that you can stand out in doing that. Because you, again, both the, the software, the, the aesthetics, the, you have that sort of, that language of motion graphics that, that, and, and an understanding of, of the AI that that goes into it. So, I think there is a certain level of retraining that goes on as part of what you're doing.  That said again, I, you know when I was in school, there simply wasn't anything around compression or encoding or live streaming, and that was stuff that I had to learn on the job’s multiple times over many, many years as, as we moved forward.

So, I, I feel like in. I, I believe this is probably true in all industries, but I, I know it is definitely true within media is that you're really never done learning correct. And, and AI is the new thing that we are learning and, and learning how, how we as an individual need to use it to, to move ourselves forward with, with what we're doing.

And I think that that is what we're seeing that we will see more of is, is that people will.  Feel threatened by it because now suddenly more people are in the, the pool for, for what they can do, but knowing how to use it will, will, will arm you and make you a more effective and, and more hirable asset coming out of that, I hope. 

Right. To me, it, it's like, it, it's a new level set, right? Yeah. You talk about democratization, which implies access, but then on the creative side, if there's a, a new level set, then if everyone knows it, then it will be the people who are using it creatively to take it to the next level. They will stand out and, and that's where creativity wins again. 

And is not crushed by ai but is empowered. Right? So, it's an interest. I, I do believe it's an interesting time. But there, there is this disruption that's going to happen. People need to learn. Maybe there's some shrinkage in terms of some of these agency sizes. Certainly, if you're managing campaigns, let alone just, doing the creative. 

A lot of that stuff is streamlined, so. Sure.   I, I think, I think social media agencies and social media teams are going to be devastated if they aren't already. But, but I also, I would, I would argue that maybe the, the, let's take the sort of the branding scenario. There are companies that never would've done worked in that world because it was simply too expensive for them to approach it because. 

It had to be done through an agency. It had to, it, it involved x amount of people and was an entire effort.  In a world where, where the, the tools have been more democratized and they're more sort of consumer level and I can go access pieces of it myself online. There, there is now a broader audience of, of companies out there that, that will potentially go work with a smaller agency on, on a branding exercise that again, never would've done that before because it was simply out of their reach again, much like there were you. 

We have entire people. We have an entire industry of people who make their living off of like a YouTube channel now, and that simply didn't exist 20 years ago. The infrastructure and the tool sets simply weren't there. And so, I. 20 years ago, could you have imagined saying that there's going to be a world where an independent producer is potentially making a living off of just putting their stuff online for, for people to go find and then be either advertised against, sponsored against, or, or simply pay for access.

So, I, I think we're in that same place with, with ai. I, I just keep coming back to that of, of historical examples of. We don't a hundred percent know how it's going to work, but the model is going to be different moving forward as well. We will, we will find patterns that work, and we will, we will have to teach ourselves the, the things that we need to do it, but we'll have to look for the new opportunities that don't maybe fit the old business model that we were using either. 

Yeah. So, I know you had a lot of experience with movie studios, sports leagues. Mm-hmm. What about the music industry? Yeah a little bit here and there, mostly probably earlier days of my career working with, with studios around how they manage the sort of both finished assets and, and the music videos in, in particular, I even did work with vivo years back, which I think of as basically the, the replacement for MTV  when, when MTV became more about reality shows and less about music videos vivo suddenly appeared and, and now I think even that's sort of been replaced somewhat just by YouTube. Yes, as a place.

But I think they're, they're very sensitive about AI much in the same way the, the studios are, they're very concerned about their, obviously not only the sort of the visual IP, but like the, the songs themselves the voices of the artists and, and how that approaches so that I think they're doing the same process the studios are, which is fairly rigorous around. 

  we may, we may go explore particular areas, but we're going to, we're going to move slowly as we approach it because we want to make sure that both our intellectual property and ultimately the artists we represent are, are.  Are covered and secured in, in any, any plans that we make. I think the area they are moving forward on quicker is, is around things like production and virtual production, because that's dramatically decreasing the cost that it takes to do.

A music video shoot, which means that they can now take on more of them whereas they may have only had, they have a fixed budget for how many for how much they can spend. And as that price goes down, they, they can potentially go do more with the same artists or bring in a wider range of artists.

They never would've done a produced a, a music video for and, and take advantage of it. I do think they're also looking at things like live events online differently and, and how they can make those more interactive and, and more engaging. And that's a lesson they're sort of learning from the, from sports leagues who have had a super fan centric approach to how they use the data they're collecting to drive an interactive experience.

And, as, as sports leagues sort of take back the production control themselves from people, they've licensed it to globally they, they can reshape the experience of what it's like to consume sports. And I think live entertainment and concerts are, are looking at similar things.  Well, you speak about different areas of the music industry.

What about the estates of deceased musical stars, rap artists, Elvis Presley how do we bring them back using ai? The Beatles did it right and they want to grammy for it.  Beatles did it. I just spoke recently this week with ReSpeecher who had a lot of press coverage during this award seasons because they were used heavily in a, in a number of movies that, that were made.

But they've also worked with estates of music artists to, to, to bring their voice back for a number of, of licensing opportunities in Elvis Presley being one of them. That, that they were able to help assist and support. All of those tests and, and sort of larger names, I think that they have worked with so far.

It sounds like it's been very much a, a heavy hands-on process to get everything just right. It's not something where you're just going to upload a dozen clips and then suddenly, magically you've got an Elvis Presley powered singing machine. It's very much a hands-on process. It takes a lot of bodies to, to get right and, and a time. 

It sounds like a days and weeks and sometimes even months process to, to get the, the results that you want. So, it's again, something that only a, a big studio or a big production is taking on right now. But. The technology is getting commoditized and, and is coming down to a point where, they even showed off in the, the conversation I had with them where they had started working with in the medical field with people who create the, um.

The, the, the equipment for, for vocalizing the, the laryngectomy, if you've, if you've lost your voice for any reason, they effectively have a way to take that. And instead of sounding like a, a very tiny robot that we're used to with these types of medical devices, you could have a very natural sounding voice, or you might even have your voice that, that could power it.

And, and so I, I think that. Overlap in the technology, which probably never would've come around on its own without the media industry being there. In other words, they, they're able to take the learnings out of media from doing work with, with the states, with others to, to get those likenesses and the intonations right, and map it into these other areas and these other techs Well, that's fascinating. 

Speaking of avatars, which that was just about voice and music. Right. But you can have a digital twin of yourself. Sure. Which is certainly being exploited. I, I have an avatar. I, I, I use it and just feed it a script. And it's me talking and.  What the heck. Right. I gave it permission.  Right, exactly.

I'm using yeah, I'm using Hagen, which I think mm-hmm. Is an ethically sourced platform. So not all of 'em are. And they're not going to train on me, but who cares about me, my voice so I don't care. But what do you think about that future? I think there's ton of, 

there's a ton of use cases for it where. Content would simply never exist because there either wasn't enough time in the day or there was no justification for being able to reduce it. I, I remember when I was at Microsoft, there was a, a big European broadcaster who had trained voice models of a few of their primary anchors and news reporters. 

Whereas they had, they had tried to bring them in to, to record a couple of headlines and, and sort of the quick hit news items that they could use on the, their website. The reality was they were never going to be able to do everything, and they certainly weren't going to be able to do it in multiple languages, because often you're wanting to cater to a broader range of languages in, in the EU to make sure that you’re, you're doing something that, that the viewers can natively watch.

But by training the, the voices, you, you could now have a producer who was helping put together the articles and, and the headlines Just create a version that got uploaded to a content management system that would create now a, a, an audio file of that presenter reading, reading that news in the same style and intonation that they, they would've done in a,  in a, in a recording.

And you haven't had to, to try to book that experience or bring them in for that time. And you can do multiple takes, or you can do way more content than, than you would've done before. And again, it. You might have had a little bit of that content, but you never would've had as much that was available without that AI technology being there in order to, to do it.

Even internally at Microsoft, we, we would use some of the, the localization pieces so that if there's a   presentation, I give a lot. Why don't we just put it into other languages so that the first time you're hearing it, you're hearing it natively so that you're not trying to understand tech and my bad English at the same time. 

That's funny. Yeah. So, it is empowering, it's amplifying I, I think there is a concern of flooding the zone with so much content that people can't even possibly consume it all. Right? And, and that becomes a problem. But people eventually find voices they want to listen to, and, and maybe AI can assist with how to navigate finding that.

Right. Which I think is the next evolution in, in television and streaming. Right? Very much so. It's like, how can you make that experience better? cause it, it sucks. Right now, I think we, you know.  I've heard the, there's going to be too much content argument before, and I feel like there’s al, there's al there's been a steady growth of content since the broadcast days through cable, through streaming.

There's been a continual sort of fragmentation of the audience, and which then allows you to.  Cater to, to, to a more niche group with, with what you're showing. In, in an age where you only had three or four channels, you were never going to be able to get to an anime specific audience or, or a, a classic film audience in, in the same way that you can when, when there's niche.

So, so I believe that the, the more content argument is, is a little hollow because. Are we going to try and watch it all? No. Are we going to find the things that we want? Absolutely. And you are a hundred percent right. An AI will have to play a role in, in helping solve that. And frankly, content discovery has been broken since.

The days of cable TV and TV guides and cable guides being, being there as a sort of single source of truth of, of the content. Oh yeah. That we knew we had access to. The EPG, the electronic program guide, preview 10. My family will literally forget, a show that we're watching because we just haven't logged into that app in a week.

And we'll go back and it's like, oh, there's two more episodes. We totally forgot that it's there because we don't have that, that nudge or that reminder and, and because things lay in a variety of different sort of data silos or entertainment silos. And so, I, I'm, I'm continually shocked. Nobody has.

Has tackled sort of the entertainment management in a and entertainment discoverability in a better way. But somebody's going to do it. I think, I think, I think it's going to happen with Alexa because Alexa and Prime, it's all same company. I think. I think they will absolutely be a contender in it. I would argue though that Apple and Apple TV is, is certainly looking to, to play a role in that space as well.

And so, I, I think there, but there will be other independent players that, that are out there that, that, that come to it. It, it can't just be, like a Google and or search mechanism that we're looking for. We're, we're looking for something that helps us curate that experience.  Yeah.

Right. Well, listen, we're getting towards the end of our little show here and I'd love to hear.  Where you're going to be next and where you're appearing. cause you're a busy guy in that way, and you're a very vocal advocate for ai. And so, what's your lineup? I've got a couple things going on, so I'm, I'm writing a lot more as you, as you called out I, I started a Substack when I left Microsoft.

And I'm just writing on the, the things that sometimes they never let me talk about. Sometimes they would. But I'm having fun doing it. And, and that's just@abe.substack.com A-B-E-A-C h.substack.com. I will be at NAB, which starts in about, about a week. I'll be heading that direction in a, in a week.

And then, I have a couple of panels both with the society for Motion Picture and television engineers and one with NAB specifically around deep seek and, and ethical ai and generally talking about content provenance. I. Other than that, I have a couple of speaking engagements. I'll actually be at Chapman University in, in April, talking to their entertainment law group and then making plans right now to go to DePaul University in early May because they have an AI and media symposium that their, their media team is putting on. 

Fantastic. Well, keep up the great work and really appreciated you talking with me today and our audience. And I just want to thank you so much and we want to have you back as.    things progress, certainly. And to all of you who are tuning in, catch more of our Realm IQ sessions on your favorite podcast platforms.

Please follow and subscribe and smash that subscribe button. You can also follow us on TikTok and LinkedIn. And now Blue Sky. Yes. Blue sky. So, thanks so much, Andy. Take care. And come back. Okay. Thank you, Curt. We'll come back anytime.