AI in sports: From Warner Bros. Discovery to NBA #1 pick Greg Oden’s AI investments

0:04Nick Meacham:Hello, everybody. Welcome back to the next episode of Streamtime Sports. My name is Chris Stone. I'm the community lead. And unfortunately, today, I'm not joined by our CEO and Mick Meacham. That's not going to stop me from bringing the hot fire today. Although Nick is taking some well-deserved time off before full speed, the Sports Pro Media Summit, which if you're a listener to the Streamtime Sports podcast, you should be attending our annual biggest event hosted in Madrid every single year. it's going to cover all the topics that we cover in the Stream Time Sports podcast but in much more detail with industry experts, lots of people to meet if you listened to last week's podcast you heard from our guest Dave Guinan, the people that he's met, the relationships that he's built by attending these events, it's certainly something you should not want to miss and speaking of the Sports Pro Media Summit over the next couple of weeks Nick and I have had the opportunity to interview a few of our key speakers for the event this will be a great opportunity for you to get a little bit of an insight into what we're going to cover over the course of the two days in Madrid.

1:02Nick Meacham:This includes interviews with Sky Deutschland, Tubi. We're also going to speak to Paris Saint-Germain. That's me trying to do my best French accent. We're also going to be speaking, hopefully, to Tim Cocker, the YouTuber who purchased the French rugby rights for their Division II football. Let's talk about how we're going in the creator space and them not just being content creators, but also broadcasters themselves. So the next couple of weeks, we're going to bring you some content solely dedicated to the Sports Pro Media Summit to give you a bit of a taste of what you can expect by attending the summit in Madrid.

1:35Nick Meacham:But for today, we're going to look back on some of the sessions we had at our Sports Pro AI event in London at the London Stadium. There's a couple of sessions we're going to bring to you today. The first one is all around how AI is being implemented, you know, case studies between broadcasters and streamers. That session was moderated by the legendary Peter Hudden, who's been in many roles across the sports industry and an incredibly intelligent individual. He's also joined by Dave Gibbs at Warner Brothers Discovery, Louise Lawler at Eurovision Sport, and Mihir Walavalkar of LiveLike. It's a really great conversation.

2:10Nick Meacham:We all are kind of aware of AI, but knowing and understanding how some of these major broadcasters are implementing AI, these group, this group, the four speakers. They give you lots of examples of different ways of how we're seeing AI here. So it's a really great conversation, one you should certainly tune into. And then the second one is an interview that I did, so very close to my heart. It's not every day that you get the opportunity to interview the former number one overall draft pick in the NBA, Greg Oden. He's certainly the tallest speaker that's ever been to a sports pro event. But it's honestly just a great conversation.

2:44Nick Meacham:um greg being a former nba guy he's very honest about his role with ai and you know he's not going to be the tech expert but it's a really great story to hear from someone going through a career transition and you know we talk about athletes getting involved with uh tech investment all the time now so to hear from him sort of what his transition is from nba superstar to tech investor um it was a really great opportunity for me to get to sit down with him also two great people representing the great state of Ohio. So that interview will come after the first one with the different broadcasters. So from here, I'll hand things over to Peter Hunnam and the conversation that he had.

3:21So my homepage on Max will be very different than your homepage, than your homepage, than your homepage. And all of that is based on the signals that we are picking up from our customers around there, how they consume content and the way that they consume content. So if you're more attuned to watching live sports, we will present you with live sports. We can go deeper. If you're more of a tennis fan, we'll present you with tennis. If you watch a little bit of sport, but you watch a lot of movies, we'll lead with Superman with a bit of sport. You get the picture, right? So really, we are investing quite heavily in how we optimize that content discovery experience for our customers.

3:56I mean, I'll give you an example of what we're doing with video. So we've just launched a new version of Max, which uses a lot more video within the carousel and on the homepage to promote a series. so what the machine will do the machine will identify six or seven drop-in moments to that show and then we'll present our marketing and merchandising team with those moments they can then pick them before they were having to watch an hour two hours worth of a show three hour movie and identify what those moments were so that's one example of how we're kind of improving that efficiency right if i can just follow up from those answers a little bit i'll start with you Dave you know Louise talked about using the automatic translation tool clearly something like WBD you're in you know multiple languages in all sorts of markets what would hold you back from from using automatic translation I think I think is that quality threshold right I think that's the kind of number one and then two like we have established workflows we have established workflows that have been in place for years.

5:06And there is a cost to changing those workflows. The video workflow is very complex. There's no one standard. So actually going into that and changing anything is not as simple as it's perceived. So then it comes back to the ROI, so the quality threshold and the ROI, really. And Louise, for you, the question would be, working for a group like Eurovision, You've got all sorts of different stakeholders there. How much of it as a challenge for you is to bring people up to speed? You have the vision of what's possible, but you're dealing with a whole load of very conservative stakeholders. Yeah, I think to what they've said, it absolutely is the case.

5:47I think that the concern around quality is there, and the EBU obviously represents the members of public service broadcasters across Europe, and there is quite a bit of reticence around that and nervousness around the quality of the AI tools. I think where Eurovision Sport has the advantages, we're at such an early kind of stage in our evolution. We're able to take those risks. And with the boxing as the example, we only found out that we'd write in India quite close to the event. And their team was super strong. And we knew that there would be engagement. So we thought, right, how can we activate this quickly?

6:21So I think, and then from the ROI perspective, we knew that we would get engagement. and it was a very efficient way for us to deliver it. Sourcing Indian commentators who specialize in boxing, all of the challenges that you face as a traditional broadcaster. But I do think that I see anyway myself in the industry, people are really interested more now in the case study. They want to see what's working, what isn't working so well so that they can truly assess the risk against the investment. Because, you know, we've talked about the evolution of AI. for it feels like quite a few years now. I want to see it working.

6:58I don't want to talk about it anymore. I want to see it in action. And when you watch it in action, or you engage with that content, whether it's personalized feeds or gamification or all the AI tools you have available, if you're able to watch it and engage with it and you see the level that it's getting to, it's never going to be perfect, right? We know that. We know that there's human intervention required on quite a few of the tools still. but if you accept that risk and I think the younger generation are much more open to accepting the risk I think as long as you're really transparent about what it is sometimes it's actually attractive to younger people where they find it quite interesting to go and see what AI is doing around their events and so for me I think there is plenty of opportunity I think the risks are still there absolutely and for public service broadcasters often the risk is too high you know at this stage but I think you know We're lucky in that we can embrace this technology and really trial it and push it and see how far it can take us.

7:58Because ultimately, personalization, localization, they're critical to engagement now. And you've got to be able to serve the fans the content that they want, not that you might think they want. It's the stuff they're actually going to engage with. And I see that's where we've got to get to. Great. I'll come to you later, Mahir, if that's right. Because I want to just follow up on that, Louise. You talked about case studies are the best way of selling the story. So who's got a good case study? Who do you think outside your own work environment is using AI well? And you can't use Dave as the example.

8:33It's an interesting question. As a middle-aged woman, I am obsessed with trying to engage younger generations now because I think with a lean-back service that I grew up with, unfortunately subscribers and engagement audiences in that market are going to be dead in 20 years so we've got to figure out how do we engage the next generation because if we don't we don't have a business right and they don't necessarily have a lot of money but they can drive engagement and the way that they consume is so different now so I love what Nickelodeon did around the NFL Super Bowl because the NFL were challenged by a drop-off in engagement with the younger audiences and what they did around the Super Bowl like the first one was phenomenal you know the big splash and everyone was obsessed and then it was Spongebob and to me it's really using kind of AR and AI obviously but to drive an entirely different feat so it's appealing completely to a different audience formula one do it really well as well um and and i think you know federations and and governing bodies are seeing that they do have to start targeting younger audiences and i think that's where we here can come in again but the gamification is is critical to that group and and how do you they don't want to sit back and watch they want to be involved and you know we're all now trying to think of ways of how do you get the audience more involved in the content through gamification, through them being able to personalise feeds, through watch-alongs and creator-led content, because that lean-back service, as I said, still has a place.

10:15There's no doubt in my mind that live sport trumps almost everything else, but it's about how do you offer an additional feed? Either they watch the main feed on their phone. I always think of it that they're watching the main feed on the TV, and my nephew flipped it completely in my head the other day because he was like, no, I watched the match on my phone and I watched the creators on the big screen. I was like, yeah, you do. So that's the kind of, that's what's happening. And I think we as broadcasters and as content creators have got to move into that realm if we're going to keep the engagement levels high and justify the subscription fees.

10:50So Dave, two questions for you then. One is just following on that, talk about how you used gamification because you clearly have used it very well within WBD, but also a third party that you're impressed with. So we were talking about this, one of the questions backstage was like, who are you impressed with? And I said Wimbledon, and I've just looked up the next panel is serving up AI innovation with Wimbledon and IBM. So they are going to be able to tell the story far better than me. But I think from that perspective, I'm coming at this from as much as a product person as a sports fan. Like, I like context, right?

11:22There's all of this data that we can collect now, but it's what does it mean and why does that matter to me? and actually I think the work that we would have done with IBM in taking some of that data and that advanced data they're picking in and then being able to tell the stories with it I think is really interesting right if someone is serving you know their return percentage is x what does that mean like is that more than where they were is it less what does that predict and how does that tell stories and I think from a fan's perspective you you want to understand more and you want to be told the story so I think um yeah I think what they do is uh is is really impressive I mean I think coming on to the the gamification piece we're all in the engagement game right we all want to keep people hanging around for longer we know if they stay for longer they consume more if they consume more they spend more money with us and they're less likely to leave us right so so we're actually working with um Live Like and this product on yeah how do we and this is more on our kind of web products at the moment rather than our streaming service but how do we engage fans for longer and what live like provides us with is a set of tools that allows us to do things like polls and predictions and quizzes and that kind of stuff um automated now it clearly needs a human to check that it's the right thing and it's engaging and valuable for our fans but that that type of thing where you know here's a story here's a video like here's here's a related game about that here's a quiz or a poll or a prediction just very simple stuff that keeps you engaged when some of my ex-sky colleagues are here and they do something very similar within their live blogs that type of thing just keeps the user there for longer it doesn't take a huge operational or editorial effort to do um and you can monetize it right and you you know and and that's what we're in we're in the engagement game and longer we can keep you there the better and the simpler that is that's a good job for us and you know it's a good partnership with live like i'm not just saying that because he's here yeah i mean even when i was at meta we did a lot of a b testing where we said okay we give a poll to x amount of people and no poll to y amount of people and you clearly saw those engagement metrics go up once they'd chosen a side or chosen a position which which obviously is food for you mihir but rather than have you talk about how great live like is, who else is doing it well?

13:45You know, who do you get impressed by? I mean, I'll take my, I'll give two examples, one at a macro level and one at a micro level. As a consumer myself, you know, I'm really impressed with what Google is doing just because it's infusing AI into every product that we use every day. So AI to summarize your emails, AI to compose your emails. Now they've just introduced the operator agent in the Chrome browser. so that'll drive adoption more because I feel like Apple and Google in the consumer space are just best positioned to drive widespread adoption because it doesn't change my habit. I don't need to download a new browser, an AI browser.

14:23Chrome is bringing AI in it and that's gonna let me experience for the first time what an operator agent is going to do on my browser versus me downloading a new browser and being an experiment on it. So, and that's sort of been our thesis as well. we want to drive workflow automation. I mean, Dave mentioned that, you know, the biggest hindrance to adoption is going to be that you have established workflows. So how do we bring the innovation, these AI capabilities in those existing workflows is what's going to drive most widespread adoption in our industry, at least, at the very least. And more like in our technology level, I'm a huge fan of CAM AI.

15:01I think you guys are working with them. Their CTO, Akshat, is a genius. A young guy came out of Apple and just like on it, like just give a problem statement, solves it. The use case is obviously translation, but it applies in many, many different ways. I think in our industry and broadcast, you know, now there's federal governance. You know, you have to have subtitling for people who can't hear. And so subtitling as a federal requirement can be solved through AI. You know, historically that might have required a lot of resources that becomes suddenly the cost has dropped significantly. So I feel like those are the types of use cases.

15:38They are micro-innovation. It's not massive change in stuff, but how do you drive more efficiencies, operational efficiencies? And that will, over time, free up the editorial teams to think more creatively. And what else can we be doing in changing the broadcast? Like, you know, Luis mentioned about the slime meter with Nickelodeon or whatever, right? Like, it's because they weren't focusing on it, or hopefully others will be able to do the same thing because they're not focusing on the everyday stuff that can be solved by AI. You showed us the genie example earlier. So you're in a world now where you're having to pitch these services to broadcasters.

16:13What's the response genuinely like? You've clearly got believers here, but in the broadcast industry as a whole, are people looking to take on these solutions? Is it a difficult pitch or an easy pitch? Honestly, once people see it, and they can, I think from a pitch perspective, making it relatable to their day-to-day lives is what makes the most difference. Now I am, you know, our board member asked me the same question when he saw our press release, like is this actually going to drive impact? And my response was, I hope so, but it won't happen until people actually start using it and behavioral change is difficult.

16:53You know, getting people to adopt something that is not part of their day-to-day routine is difficult. So I think everything that we have done from an integrated workflow standpoint is already being used. If you're spending 30 minutes coming up with a question bank for trivia, the right answer, the wrong answers, the two other wrong answers, and spending 30 minutes on it, and that's now being done in three seconds, of course you're going to use it. It's just, you know, it's self-preservation, right? Like just making your life easier. The genie is more of a standalone product offering that allows you, it's more like an operator agent, which is going to require meaningful effort from a partner to start using it day over day.

17:34Where we are spending time on it, and I think this is the part that's resonating most with people, if I came to you and I said, Peter, you're doing this job, I'm going to help you do your job better, that's a personal choice. You can decide whether it works for you or not. Where we are focusing on is actually coordination. It's more, enterprises are messy, right? Like you have siloed departments, editorial product, marketing, BI. They all have to work together. And that part is where things fall down mostly. So they might want to launch something, but the editorial team may not be totally ready to adopt it.

18:07And hence the feature can get launched. Now if these agents can actually do that job for you, so the editorial agent and the analytics agent and the marketing agent are all in unison at all points in time, that to me is the biggest impact. It's not single use agents. It's those agents all talking to each other and this network of agents working together. Dave, is that fair of the world that you have to work in? Is there a separation between digital and editorial that you need to get across? So yes, to an extent. I mean, we're all kind of after the same thing, which is to increase our number of subscribers.

18:46I mean, that's our job. I think the thing for me is you're asking these editorial teams to do more things. So if you think going back to the customer again, things like an indexable timeline, key moments, recap, whatever we want to call it, they're kind of table stakes now. I mean, that's not innovation. That's an expectation from the customer that's there. What do you mean I can't come into a game 20 minutes late and just find out what's happened? like maybe that's how I want to consume it and maybe that's where I see the real value so things like that table stakes and then we talk about alt casts and all this kind of stuff and you know of course you know YouTube are gonna win in that space in terms of scale right stick the NFL with Mr Beast or speed or whatever of course so but what can we do when we have a user base that you know starts sort of finishes sorry 75 80 85 and starts at nine or ten right so how do you create an experience that works for all of those audiences and to try and do that manually and create four or five different versions of the same stream is just hard it's just hard you're asking the same team to do five different things so you know where what we've done is try to invest in those tools that drive the biggest impact you know things like our key moments we can roll those out across multiple sports in multiple languages in different countries you know things like rapid recaps you know it needs to be more than just stitching those key moments together because that's not particularly good edit if you're paying us 30 pounds a month so i think you know i think i think as long as we are working on tools that make that editorial team's job either easier or they don't have to do the same thing for different things five times i think that works for us yeah you you have a unique position in that you're also working across the u.s and what tnt are doing there.

20:41Do you see a different attitude in the US? Are we behind in Europe? Are we ahead in Europe? I honestly don't think we're behind from everything that I've seen. Other people may have a different view. I think we're all in roughly the same place. Obviously some sports are different, right? You know, a highlight service from the NBA is very different than a highlight service from the Premier League. You know, a 0-0 draw versus a 90-86 points like event, there's a lot of different type of content that's produced. And so I think for the automation of those highlights, they have been more advanced just because of the sheer volume.

21:19But everything that I see, I think we're in a decent place. I mean, if you look at what ESPN have just launched, it's a good app. It's a good app with good stuff in it. But I don't think there's anything that we haven't seen in these markets. Louise, you've been through this change with DAZN, with now Eurovision, where do you think the biggest gains in efficiency have been that you've seen over that period? I think AI driving personalisation of the app is critical to what David said. I think, you know, you could now go onto an app and be served what you want to watch, what you traditionally engage with.

21:56And AI plays a massive role in that. Just the efficiencies of not having teams running, you know, apps in languages, as apps in, you know, I only follow Premier League or I only follow rugby or whatever your thing is. The metadata that you can gather and the algorithms that you can gather from your consumers allow you to drive a personalised feed. And to me, that's one of the biggest innovations that I think we have got to use because people are impatient. They don't want to log on to an app and scroll through tons of stuff to try and find something to watch. When it happens to me, I can tell you now, I close the app because I don't have the attention span to find something.

22:35Whereas if I go on Netflix and they say, you've watched this, you're going to enjoy this, I will watch the trailer. I might not watch the series there and then, but I will watch the trailer. And so by serving up content that you know people will engage with, you're just almost guaranteed that engagement level. And I think that what AI can do to serve up that personalized app experience is groundbreaking because the efficiencies that you can make across that and also make it really personalized. I am a certain team. I don't want to watch one team against another. I want to watch this team. I want content about this athlete.

23:14So being able to create content and serve content around what my algorithm, what I'm interested in, is critical. And I think AI allows people to do that at scale in multiple languages across multiple territories. And that's where I think for global companies, it's a game changer. Just to build on that, I kind of did some research before I came in here. And I was like, this is what we're doing. What's everybody else doing? Are we behind the curve? Is there FOMO? Are we out of the curve? And I saw a piece of research that a guy called Dan Rabin, if any of you are familiar with him, had done. I think you guys have, you've had him on the podcast.

23:52And he did a piece, I think he interviewed close to 200 execs, OTT execs, streaming, sports leagues, etc. And the question was like, are you using AI in your video workflow? I think 68 % of them said no. I think the 18 said yes, and the rest said, yeah, we plan to. I mean, read into that what you like, but I think the single biggest use case was content discovery. Yeah, and I'll go back to what we said at the start. When we are trying to make sports live within a much broader entertainment experience and give that customer all of that choice, the more that we can use those signals to optimise that experience is a real opportunity for us.

24:33Mihir, what about, we all talk here as like we're complete converts. What are the misconceptions about AI that you come up against? I mean, look, I think there's a, we all are sweeped, have been swept up by the hype and wave of AI. I think from my standpoint, a lot of people think that AI is going to essentially, we're going to get into this autonomous world where everything is being served up to you by AI, so to speak. And to me, I mean, I have two kids, six and three, I'm scared of what the world looks like in 20 years from now. But if that's what the reality is, but like, I personally believe AI should be more of a helper, rather than someone who's going to replace this conception that, oh, AI is going to replace all jobs.

25:22We see it on our own day to day, right? Like from a, you know, I'm running a startup. How many people are we hiring? How many? I don't think it's taking away jobs. I think it's making your existing team members do more, do better. Now, does that mean that I may not need to hire as many people if the same person has superpowers? Like, you know, that's probably true. But from a perspective of, like, especially in our industry, AI can do things on its own is just not going to happen. I mean, you talked about the Hindi use case in boxing. even the biggest tech companies' translations or automation from using AI is going to miss the nuance, is going to miss the context, which for a discerning audience, which the younger audiences especially are, but even the older audiences are, it just doesn't work.

26:11You have to adapt it. You have to train those models on specific things that are key to that sport in that country, in that language. It's not just translate how this sport is consumed or translate how this language, you know, go from English to Hindi. It's listening to boxing in Hindi. I mean, I would love to listen to the imagined translation in Hindi. And maybe, Peter, you can tell us with your Hindi knowledge how well it did. But, you know, I think that's the one misconception in my mind is that it's going to be autonomous and we're going to just replace everything. I don't think that's the case.

26:46On the other hand, though, people go on the other extreme. Oh, it's hallucinating. Oh, it's not accurate. Well, it's not supposed to be perfect. It's supposed to make your life easier. Don't throw away the good, the okay with everything else. From my perspective, that's where the adoption needs to be. On that it's not perfect thing, the best quote I always use about AI is, this is the worst that AI will ever be, and it will learn. Because we've got a very tech-savvy audience, You clearly don't need the instruction that there's a QR code there to submit questions because we've already got our first few questions in.

27:24Just running through some of them, I'm going to pin on you, Louise, because you mentioned Nickelodeon, the NFL. One of the questions is you mentioned SpongeBob and how it engages a new audience. How can you shed more light on how this engages new audiences? I don't have the viewing data, so I don't know. but I do think that offering kids a more kids-focused feed will keep them in front of the TV longer. Whether that's a good thing or a bad thing is a whole other question. I was hoping you were going to ask me the next one, which is about the duration of the content. You're allowed to answer that as well.

28:02So the question, just so in case someone can't read it, is right behind your head. I think the interesting thing is we all thought, even just two years ago, that it was about snackable content, and I hate that word. It's not, because they're watching full live streams with creators watching a football match. So I don't believe, I think it's a myth that their attention span is short. I think all our attention spans are shorter because we're all addicted to our phones. But I think if you're putting something that's interesting on the phone, people are staying on the phone. So I don't think that it is about duration.

28:34I think it's about engaging content where they feel part of the event and not just sitting back and watching someone serve them something. so I do know that for F1 they are getting more engagement across that piece that they're doing I don't again have the numbers but I think you are creating a bespoke feed that doesn't have an adult theme to it it's based on what the kids want to watch and if you can bring their world into a sporting world then you're uniting their interests and therefore that kind of cross-pollination of interest happens and to me, I think it's a great way to get kids into sport now.

29:15I mean, I can just quickly add to that. I think that was step one, right? NFL putting the stuff on Paramount, then Disney did the same thing by using their own IP in it, and it brought a lot of brand sponsorship integration in there. I know Sony is working a lot through Beyond Sports on this stuff, but it was step one. Step two, three, four, now you're bringing the creators in there. Now you're giving an NFL game internationally to speed to comment on. So I don't think that one thing changed behaviors. It brought a different type of audience into the world of NFL. And now you're keeping that audience by giving the content to the creators that they watch the most on YouTube and doing these co-viewing things with Speed and Mr.

29:53Beast and stuff. So I think for me, I think it's a series of experiments rather than one thing that drove a meaningful change. Dave, is that fair of what you're doing at WBD? You're looking at a series of experiments. And do you recognise that question as something that you're asked in terms of your normal job? Won't the younger gen grow up out of short and sharp content? I mean, ask that question. I mean, no one in here is 10 years old, right? And I bet we're all consuming short form video in some case. But I think the Olympics is a really good example of how we covered all of that choice, all of those audiences and all of those different consumptions.