Fast fail, community personalisation and changing vendor relationships: What AI could mean for the future of the broadcast industry

The transformative impact of AI on the media and broadcast industry was discussed in detail during a collaborative event hosted by Diversified last week. During ‘How AI Will Shape the Broadcast and Corporate Production Industry in 2025’, Jared Timmins, vice president of media innovation, and Chris Keath, director of machine learning at the company, shared not just how Diversified is using AI to transform its business, but also where it’s having the most impact across the broadcast sector.

When it comes to the role of AI in media, Timmins highlighted the emergence of new entrants into the market who are disrupting the status quo, especially when it comes to sports broadcasts.

“We’re seeing new entrants into our market, both on the tier four side, which is the single operator production, all the way to what the immersive realities are building for us on the tier one side. If you’ve seen any of the football stuff that they’re doing, where your players can now be Nickelodeon players, or they can be Toy Story players, it’s this new defining of what tier one means around immersive reality. The problem is that it’s a different tech stack at tier one, a different tech stack at tier four, and both of these are pushing in on our core markets that we’ve had in media. So it really requires that we think in a way that we can expand our addressable markets out not just to media workflows, but to anything that’s media adjacent as well. We believe that AI is an extreme driver towards that.”

Keath highlighted specific areas where AI is already thriving in the M&E sector today – with point solutions and rapid application development leading the way.

He said: “Point source solutions such as automatic logging for media asset management and real time translation are mature in the market and are pretty reliable ways to take AI and see real results on the other side.”

He cited the example of Fox Sports who no longer have a Browse view in their MAM, instead opting for a YouTube-type setup that allows for incredible accuracy. “Not only can you type ‘LeBron James three pointer’, you could type ‘LeBron James three pointer after he claps his hands with dust in the air when he’s playing against Houston and it’s the fourth quarter’,” he expanded.

“So in a lot of ways, they’re changing these high end models, reindexing all of their content and then using that to drive all of their search and modeling. The interesting question that I ask all the vendors when I see them who do this kind of stuff is, ‘can you take all the existing information that we have inside of those systems and incorporate it into your model?’ The answer is always no, but in terms of point focus solutions, that’s the kind of stuff we’re seeing happen out there right now – direct applications that very quickly and easily slot into media workflows that we already have.”

When it comes to application development, not only are advances here creating a completely new way of looking at software, they’re also enabling new levels of customisation.

Keath explained: “As we move towards zero-cost software development, customisation becomes trivial. No longer do we have to get a piece of software to perform some kind of media-centric business operation and conform your business operation to how that software is telling you to do your job. Now we can come in and say, ‘Okay, tell me what you’d like the user interface to be, tell me what you want the workflow to be’. Two days later, I’m going to show you that user interface, and then we can start moving on from there. I always enjoy doing bespoke, clever integrations so I’m really excited for this new world where we can go back to really embracing all of these in depth, customised ways to help everybody work. The real question is, how do we take this next paradigm of tools and use it to 10x and magnify and really drill in and make those efficiencies and make those systems and workflows operate at a high level so it changes the way of working.”

Rapid app development combined with the fact that coding has now become so easy and accessible that companies such as Google and Salesforce have made many of their coders redundant, means Diversified itself is moving from being a system integration company to a software integration company. “That makes up a not unsubstantial part of our go forward path,” said Keath.

Looking ahead and both agreed that further change is inevitable, however. “I’m 99% sure in the next two years, the number of vendors that serve this market is going to increase five to 10 fold,” said Keath.

“Rapid application development means that your vendor relationships are going to start to change, prices are going to drop, and new people are going to be popping up on a regular basis with highly competitive and highly capable products. The question then is how we start to build systems that are reusable and that do not constantly evolve with the underlying platforms. And that goes back to the AI identity operating model and open-source solutions.”

The transition of AI to the operating system level is also an area to watch, according to Timmins. “This means that AI can now do anything on the computer that I can do, and that changes the whole game, because now I have agents operating on my behalf, as me using my computer or your employees’ computer. That brings a bunch of challenges, but a bunch of opportunities.”

“We all talked about individual personalisation, but the truth is, I think that the real future is community personalisation”

Of course, how to monetise these new capabilities continues to be a challenge across the industry, but the potential is there, particularly when it comes to personalising content.

Keath continues: “We all talked about individual personalisation, but the truth is, I think that the real future is community personalisation. Why not have a disposable application that is Manchester versus Liverpool only served to people in Liverpool. Manchester versus Liverpool only served to people in Manchester? And now at least you’ve got a clear, addressable audience. You’re not trying to integrate various demographics, it’s an identifiable market segment of people who care that can be sold to in bulk, managed, and so on and so forth.”

Flexibility in terms of space configuration was also cited as a way to better use available resources. Whereas the norm has always been purpose-built spaces capitalised against the cost of a show, the future, according to Keath is “dynamic reconfigurability” which will enable “fast fail” to take hold in the broadcast sector. “If you can’t fast fail, then you can’t trial and error, and you can’t figure out what’s going to work and stay competitive in the market. This is something we want to see the broadcast and media industries be able to embrace – we put the show on the air for six weeks and it didn’t work, so we’re taking it off, and now we’re doing another one. It’s really hard to imagine in the current paradigm, being able to switch things around that fast, but I really think in the next three to five years, if not less, that that should be fully possible and realistic in our industry. New architectures with AI on top of them just give us another level of momentum that we can take into our markets.”

“I’m excited about all the ways that this industry can use AI tools to better perform what we are doing, but I am more excited about this industry’s ability to broaden its reach by being able to engage with more of the social changes of our time and figure out how to go broader with the resources that we have. This ability to fire off custom applications, this ability to fire off all tasks and do that at a price point that was just unapproachable before, and then to quickly iterate on those things, I think is actually, at least for me, as exciting as figuring out how to slot things in,” concluded Keath.