Leanne: Ready. Okay, great. Hello and welcome to our latest Tech Talk's podcast, 2023 Unwrapped AI's Media Revolution and a Glimpse into 2024. So today we'll be diving into the world of media and entertainment, reflecting on the standout trends of 2023 and making our predictions for the transformative year ahead in 2024. I'm Leanne Tomlin from Perifery and we have an expert panel of speakers today. If you'd like to introduce yourselves, Tina.
Tina: Hey, everybody. Tina Spinello. I work with DataCore in the Perifery Division. I work for the Americas and the sales side for the West and Central team, specifically with M&E.
Leanne: And David.
David: Hi. David Fabrizio, our manager of Business Development. I help align our partners and customers together with the appropriate Perifery personnel and speak to all our new opportunities. Thankful to be on the podcast.
Leanne: Thank you. And Jonathan.
Jonathan: Hi. So yes, Jonathan Morgan I am the SVP of Product and Technology in Perifery. And before that I was the founder and CEO of Object Matrix who were acquired by dates core/Perifery. And I'm very pleased to be here.
Leanne: Thank you, everybody. So, to kick things off, Colin's dictionary declared AI as the word of the year for 2023. How has it been? A major talking point? What's specific impact has it made on media and entertainment workflows? Jonathan, I'm going to fire this first question over to you.
Jonathan: my goodness. You can't talk about trend technology and 2023 without that slightly overused term of AI coming up in some way or form in media and entertainment. I don't know. It seems to be at the forefront of AI. It seems to be the focal points. But whether that's because of the strikes, whether it's because of just the multitude of ways it can be used, whether it's in creation of media, whether it's in metadata tagging, whether it's in AI ops, it just seems to be everywhere. When you come up and any conference, any expo, there's a Gartner charts of technologies and Gartner charts as this is a peak hype, the word AI.
And that is definitely where we are at today. I mean, you must have a very different view of it from in LA or maybe the same view of it in LA you probably come up with, you've been in the middle of all the strikes, whether for the good and the bad. You've seen good and bad arguments for and against AI and maybe perhaps in terms of the technology. How do you sit in sales? how do you sit in terms of what people are asking for and where this is going?
Tine: Yeah. I think Jonathan makes some really good points, right? This is obviously a very timely subject right now and impacts so many different people there. There is a lot of potential goodness, though, because AI is currently and has been helping companies in media, in the entertainment sector with their content production. These AI algorithms help to create art. They produce content and they can also reduce the time and cost associated with production significantly.
And being that can also increase the time to market for these productions. I think where the challenge lies, is it's timely. But also, there's so many unknowns with 2023 and having both the writers and the actors strike, it legitimizes some of these concerns about digital likeness and replicates for real. These concerns are much, much needed in these negotiations and are now part of these addendums in both the contracts for both the writers and the actors. But still, there's so many unknowns.
David, do you have anything you want to contribute?
David: Yes. When it when it comes down to it, there's definitely the concerns around the talent side of things, the creative side. But when it comes to specifically the back end technical operations, there's also some benefits where I think AI is seen today and beyond. And when it comes down to that in speaking with our clients, for instance, I see a lot on transcoding and transcription services, basically areas of the business where they would like to automate processes so that way they're able to spend their time on much more worthwhile benefits to the organization.
So, I do see a willingness to embrace it in certain areas. But of course, there is concerns that are very well legitimate in other areas where AI may or may not be a good fit. So, I think it just really depends on the benefit that’s being looked at.
Jonathan: You can't shut the books now, can you? I mean, the technology moves forwards. We quite often worry us because it's such change so fast. It's at times revolutionary. And this is probably one of those moments and the thing is, you can't ignore it. You probably aren't going to embrace that to do good things for sure. You have to embrace it. You have to do things more efficiently because if you don't, your competitors will.
But however, the human aspects of it, the not becoming automatons ourselves. allowing us to be human but enhanced with better tools and better technologies, that's where we have to be. And it's a very difficult balance. It's very difficult to move forwards at times. But however, it's inevitable that these technologies will be embraced, that's for sure.
David: I totally agree. I think, when it comes down to it, there's certain areas that that have a way to go, right? for instance, when it comes to, concerns around script writing, I don't think that really today we're at all there. can it be done in some form or fashion?
Yes. but it's not going to be replacing the creative mind of a writer. But when it comes down to perhaps doing some of that transcription, that may be a good fit for it. I think it very much is, from what I'm hearing in the field, and this is something that that folks are looking to do in the now and obviously we're happy to help them with that.
Leanne: Yeah. AI definitely brought efficiencies and quality improvements and there's certainly monetization opportunities across the M&E workflow. So, any examples or instances where AI has notably transformed the industry during the last 12 months?
Tina: Well, obviously, Leanne, we've been talking about it, right? It's the writer and the actor strike. That strike lasted for close to 150 days and the effect was across the entire industry. I mean, not just the studios and the actors, but everybody else that helped support that industry from the caterers to the set builders, makeup artists, drivers, caterers. We can go on and on. Right.
It's just it was a huge impact. And I think back to once what we spoke about earlier, that it's really affecting the human part of this. And so how do we make sure that the technology is there to support this transformation? Because it's going to really help with the streamlining of the workflows. And that part is very exciting. And how do we also balance that out with the human creativity?
David: I think a lot of the folks that I'm speaking with for instance, on the post-production side, this is, as Jonathan said, the ball is moving forward regardless, right? So, it's something that you have to be a part of. And like any other technology or new technology to one's organization, they're looking at their workflow going, how can this help me?
And I think that's what is coming down to And I'm speaking to clients, who are looking at their workflow and recognizing areas, that are part of it that can be automated and , and, and, just that example I mentioned earlier, that's really a real world example from many of my customers, who are looking to, do exactly that and looking to do it in the now because it's a needed part of that post-production process. And it needs to be done. But by automating that piece of the project, they're able to go to market sooner in regards to completing that project.
They're able, to invest in other areas that are just going to make them more competitive. When it comes down to it, for instance, in the post-production space, that's an area where you need to be ultra-competitive and it's not just from a pricing standpoint, but it's in terms of how you differentiate yourself as an organization. And this enables them to also stay at the forefront of their vertical as well. And when you look in Burbank, you're talking hundreds of companies in a small amount of miles, and again, that need for differentiation, has to be there. And that's why they're looking in the now at how they can remain ahead of their competitors.
Jonathan: And I totally agree. I mean, some of it is absolutely game changing. Some of it is hype, that's for sure. So, yeah, we've been editing films on our phones to a better and better degree over the last five years, let's say. So, some of it's just better algorithms, right? Some of it is just, hey, we can do this even better now than we could last year. And we've seen that natural progression where it's new and revolutionary certainly, in natural language translation.
And then you start to see, I can do a search in my asset manager tool that isn't going, I want to find camera X, I can say, give me a shot where Leanne’s walking through a green door and smiling as she goes through the door and suddenly you'll get 10 clips come up, give me a slam dunk where, the client shines that face into the camera and smiles and 10 clips come up. Right? So, this is amazing because we couldn't search like that previously, but we know search is going in that direction.
And yeah, it's exciting. I semi-apologize not entirely apologize this is technology that that Perifery is bringing to market as well with our AI+ and that is just to me it's super exciting. It's better ways of doing things that are big step changes. So, as you were talking about that, David and I, I'm excited as a technologist to see it. I think where we legitimately are worried is where it takes away jobs that used to happen. You have metadata extraction. You can take a film now maybe semantically analyze that film with AI. Extracts what's happening in that film, take it all without any metadata that was that previously, have a speech to text, have a plot summary, be able to search on that with natural language processing and be able to bring that to the forefront. Who wouldn't want to do that?
Of course, we want to do that. Of course he wants to such a video archive, which is exciting for a company again like Perifery, because we want you to keep your video archives, We want you to be able to reuse the monetize them, bring out the value from them, be able to use them in new productions or your, TV, maybe news programs or, or be able to do all of these things is so exciting because that opens up doors where it affects jobs that previously existed. Of course, everybody is going to that's make sure we do this in the right way.
David: That's a great point, I must say, because you got me thinking just back to that phone example. Right. And what you said in regards to why wouldn't, you want to you want to do this in regards to, certain benefits. Obviously, there's the concerns that are very, as I said, legitimate.
But those benefits and, when we think about it, we've now for years been able to go and search an individual. I was recently just searching an individual in my family and looking through thousands of pictures by just typing the three letters of that person's name and being able to look at literally seven years of pictures and bring in milliseconds, right to the forefront, those pictures that match that person's name and obviously their name wasn't on the picture, but through AI, through those analytics, it was able to know that this is what I'm looking for.
Normally I would have to go through those thousands of pictures if I wanted to do that, and that could take me days versus milliseconds? So, by being able to go and really automate certain parts of that workflow could be very beneficial to the organization. And as you said, Perifery is really at the forefront in automating workflows and leveraging these different AI-powered solutions, cloud-based solutions and really being able to be at the forefront of that AI and ML space. So, we're really excited about that.
Leanne: Okay. So just shifting on to on premise solution. So, in response to the growth of archives, media companies are shifting from public cloud to on premise solutions. Can you sort of elaborate on how this transition impacts the cost efficiency of utilizing AI algorithms for content analysis and repurposing? I'm going to go to Jonathan on this one first.
Jonathan: Yeah, and I'm really happy to take this question. It's interesting, right, with AI because AI encourages number one active archive, because you need you don't want an offline archive, whether that's LTO or deep cloud or whichever technology you want to refer to because you want to be able to go through that media and collect information that will make search better, that will make re-usage much easier for your editors.
Now, if you're going to go through your media, you could be going through a lot of data, hundreds of terabytes potentially. And every time you pass through that data, you're potentially facing the egress fees, you're potentially facing compute power fees. If you're doing that in the cloud. And if you add up all of those fees, they can that can be a lot of money. Right. And then what happens the next day you have another AI algorithm because there's something even cooler that you want to be able to do or you want to retrain your model and be able to do another AI thing on your media library.
And then you pay that fee again. and then what's really cool is your results are good people reuse your data more. So, you pay more fees. Guess what? Egress that content again when you use that. So, all of this is a vicious cycle of fees because even though it's your content, you're paying to access your content potentially, depending which solution you're going with, you're potentially paying to do that.
So now look at that same archive or media library back on premises and you can run a thousand AI retrained models against your media library without paying a single penny other than maybe a few electricity charges as you do this potentially, so I'm not saying that's one argument But however, when you look at the cost models that go with large media sets, large content sets and all of this AI stuff that's happening, it can completely change the dynamic of the cost model you might have made even a couple of years ago.
And as people are doing, that's that very nice. And perhaps maybe the best solution here is to have that content on premises and be able do this quickly, efficiently, cost effectively on prem.
David: Yeah, that's a that's a great point. I mean, just to kind of focus on the on prem aspect of that question for a second. as Jonathan mentioned before, the content analysis capabilities with AI are just are just endless. So that's a benefit that folks are able to leverage. But when it comes down to it, why are folks looking to go on prem?
Some of those folks were folks just a few years ago who were looking to go all cloud. And the reality is, is that those ingress and egress fees, every time you access your content, your data, that that cash register is ringing and it's becoming, for many organizations, more beneficial for them to either go completely back on prem or most are looking to go into a hybrid workflow scenario. That's really what I've seen. and when it comes down to it, when it comes to archive, what often happens is, a cheap and slow SLA is what's being utilized. But if I'm waiting perhaps for days because that's on some cold archive tier, it could be a very long time before I have what I'm looking for. And if I'm not utilizing AI capabilities and I may have gotten the wrong data, now need to wait another five days.
So, it's really important to be using, those particular tools. But when I need my content, when I need my data, I need it now. if the client is asking a post-production house, for a particular project because they're doing a remastering of a certain film and they need to see all of that, it's really important that they're able to search that metadata and have that at their fingertips. Today, that is see it as sitting, as I said, potentially on cloud. It's sitting on potentially on LTO, which has to be digitized every few years. Hopefully they have the player to play that old tape. Hopefully those contents are accessible and hopefully I'm able to locate that tape in a reasonable timeframe.
Hopefully it's not in a vault, states away from where I'm sitting today and hopefully it's not on my production storage still because then I'm using that inefficiently. And what I just mentioned, those are all issues that folks are facing today regarding that inefficiency, whether it be to retrieve that content, ability to retrieve that content or utilize the production storage efficiently. So, what we're seeing to just sum it up is folks looking to have basically an on prem instance of their archive-worthy content.
And then if it's going to be a set and forget it, yes, then they'll put it on to a type of media that perhaps would make sense. But people, I think, are looking to change, into an object-based solution for those particular needs where they're able to access that data, not get hit with those ingress and egress fees. And know that the time to do that is milliseconds. And when you take the benefits of A.I. and you couple them with the benefits of an on prem object storage solution, you get exactly that - access to your content nearly immediately where the data, the content is secure, it's protected, and it's immediately searchable.
Tina: I think that's a great point David makes, because really at the end of the day, it's really having a choice based upon what's best for your IP and your data, whether it is in cloud or a hybrid approach. And obviously our solution at Perifery can help you with both of those scenarios. So, I think both are both of those options are definitely on the table and, as this technology grows faster and faster every day, and it's important to understand that it's really about the flexibility and the choice.
Jonathan: So it's a good point Tina. It reminds me actually of a phrase that I was around a couple of years ago, Cloud first, and I hated that phrase. And the reason I hated that phrase was because it was almost like saying, we don't care about the realities. We're just going to go with cloud. we're not going to think about it.
We're just going to go the cloud, unless you really force me to go in the direction. That's not how the world works, right? I mean, you have to be a bit more adaptable. You have to put the right workflows in the right places. Unfortunately, I wish there was a single solution that just solved everything. But it's just not the case that there are economic reasons. There's good point, David, about the time to get to your content and be able to do that in an efficient manner.
These all factor in to it. There isn't just one place to keep a library and every company out there has its own unique and special and brilliant attributes that they need to deal with, with their media libraries in their archives and cloud first, hopefully is in the dustbin. It's more about do the right thing. And that is not just defaulting to one solution.
David: Exactly. And depending on the client I speak with, I have clients, as I said, whether they're looking to get out of cloud. And they may be looking to be on prem for that archive and Perifery can help them. It's very ideal, probably one of the top use cases I've seen. But then, as I said, I see others where, they realize that maybe they had that cloud first mentality, and it didn't work because from an English perspective it's gotten crazy.
Jonathan: But then they got the bill right?
David: Exactly. They got the bill and now they're going, what? I realized the benefits. But I now realize some disadvantages. So, I'm going to have a hybrid workflow where I have that on prem. And then when I'm not likely to access that access that I call it, set it and forget it, then I'll put it, in the cloud because, I'm just, dumping it off to a tier I'm not going to access at some point, but that tier is in today. It's maybe 12 months down the line, six months down the line.
So having it on prem and again, not on the production I think is very important. So, being able to help our clients, whether they want to have it in the cloud at some point or whether they're looking for an alternative to the cloud or an alternative to LTO, we can help them with all three of those options. And depending on the client, there's different preferences in where they want to have their data and for what time they want to have it. And it's great being able to help them with whatever that time frame may be.
Leanne: So, looking ahead into 2024, what trends do you expect to see dominate the media and entertainment industry? Tina, I'd love to hear your thoughts on this.
Tine: Well, we talked a lot about this today, both Jonathan and David. But make media workflows and conversational AI is obviously going to be impacted. honestly with the impact of the strike We'll just kind of have to wait and see how those negotiations go in to play because the contracts have just kind of been out and been blessed. So, we're not sure how that's going to either enhance the creative impact or hurt that from an AI perspective, there is a lot of unknowns based upon these recent legal negotiations. But 2024 is going to be an exciting year. And as they say in media, definitely stay tuned. Love to hear what David and Jonathan's thoughts are.
David: Yeah, I think getting that content, on prem, from on set are getting it on prem and then automating the capabilities around transcription, I think is going to be a big one. that's an area that I think is one of the first areas where again, post-production facilities are leveraging or looking to leverage AI and being able to get that, content, onto that repository and then being able to go and, automate, that part of the workflow in the different languages for the markets that they serve are going to be, I think utilized more and more, in the new year. And I think they're going to find other ways, themselves even in how to utilize the different technologies.
Jonathan: Good points. I feel like I should come up with something that isn't A.I., but I'll probably fail completely. I was thinking particularly about how Edge will come into play over the next year. We are getting stronger and stronger and this includes AI, so it's not really a non-AI answer, but we're getting stronger and stronger at having technologies that are self-contained and can do a job and perhaps can kind of prepare that locally before a move and the results of that job up to the next stage be that on prem or in cloud.
So you might have an edge device that is perhaps recording a high school football game and maybe being able to prepare that in a way that when it gets pushed back up to the next stage, it can be broadcast. You might have an edge device that as simply doing local transcode and then maybe passing that back up to the central on premises systems for news articles that can be immediately broadcast, etc. But the advantage of doing more on the edge there is that you can reduce down the amount of data, you can extract metadata quickly, and then you can pass on the benefits of the content to the next stage of the process in the most efficient manner.
And if you can do that with devices that don't require, you basically fire, forget almost you put them there and let them do that job and you don't have to manage them. That's a revolutionary way of being able to make production and to do certain parts of the workflows. I'm not talking about film; I'm not talking about high-end production. But in terms of certain parts of the workflow, I think Edge will come more and more into play and I predict ‘24 will really start to see that coming into to the sphere of what people are purchasing technology wise.
David: during the strike, here in L.A. and beyond, of course, non-scripted reality shows were one of the few things that were not affected by the strike and what Jonathan just mentioned, and kind of my answer as well, was it was just an example of what conversations I ran into one after the other. it was being able to go have that edge device in the field, as mentioned on set, if you will, and being able to go and take that content, upload it on prem, being able to extract that metadata, being able to add transcription, and then after that, do the rendering of the content and having that ability to automate those first parts of it in their workflow just gives them a tremendous advantage.
And, now that they've seen a better, more efficient, more cost effective, quick way to do it, the genie is out of the bottle. Right? And I think we're just going to see more of that and not just, with reality shows, but I totally agree with Jonathan, having that capability. And obviously we can help them, Perifery with that of providing an edge device, getting that content to the core. And then whether it, needs to be in cloud or on prem at the core, we're able to go and help them so that way they're able to render their content and then decide to archive it on the appropriate type of solution.
Leanne: All great points. Thank you so much for those. Well, that's a wrap on today's podcast. I think it's clear that AI continues to be a driving force in reshaping the media landscape, and I'm certainly excited to see what 2024 brings us. Thank you to our panel, Tina, David and Jonathan, and thanks for listening. And stay tuned for more insights into this ever-evolving world of media and entertainment. Thank you.