Event Replay: Game Time With the San Antonio Spurs: Engaging Fans with ChatGPT & Sora

[00:00:00] Hi everyone, welcome, happy Wednesday. [00:00:14] I'm smiling so big right now because, one, I don't know if you guys know this, but I'm from San Antonio, Texas. [00:00:22] I might be the only person that works at OpenAI originally from San Antonio, Texas, so I'm really thrilled to be hosting this first today. [00:00:30] But also, I'm seeing all of these new folks in the OpenAI Forum community. [00:00:34] I want to say hi to Gareth Jobling, Senior Marketing Technologist at the Cleveland Cavaliers, so many sports teams, Nate Black of Sports at Adobe, and Jordan Weatherby, Senior Director of Analytics at the Miami Dolphins, that's really rad. [00:00:51] I think we might have a few more awesome folks in here. [00:00:55] Okay, sports, yeah, that's so awesome. [00:00:57] Thank you so much for joining us, and thank you to the San Antonio Spurs for boosting the signal and helping us expand our ecosystem of sports folks in the OpenAI Forum. [00:01:09] So, welcome everyone, I'm Natalie Cone, Head of the OpenAI Forum, the ECC community hosting our conversation this morning. [00:01:16] In the Forum, we spotlight discussions that reveal how AI is helping people tackle hard and we share cutting-edge research that deepens our understanding of this unprecedented technology's societal impact. [00:01:28] AI is an innovation on the scale of electricity. It's transforming how we live, work and connect. [00:01:36] OpenAI's mission is to ensure that as AI advances, it benefits everyone. [00:01:41] We're building AI to help people solve the toughest challenges because solving hard problems creates the greatest benefits, driving scientific discoveries, improving health care and education, and boosting productivity across the world. [00:01:54] We've also found that there are amazing use cases out there that leverage AI as a tool to bring people together and render creative expression more accessible. [00:02:04] Forum travelled to Singapore earlier this year to collaborate with the Peranakan Museum and Askmona to bring to life cultural artifacts and allow people to connect more deeply with objects imbued with the stories and history of their cultural heritage right around the 60th anniversary of Singapore's independence. [00:02:20] We've also hosted Sora Alpha artists such as Nick Klaverov who composed the intro to Netflix Narcos using OpenAI technology. [00:02:27] Today, we're going to explore more how my hometown team, the San Antonio Spurs, go Spurs go, are innovating with AI not only to drive business value, but to engage fans in super fun and creative and meaningful ways. [00:02:44] Old fans and new know that the San Antonio Spurs have always been ahead of the curve in relationship to technology adoption. [00:02:50] But they've also been using AI to train players for years. [00:02:54] And now, last year, the NBA Spurs became the first in the league to get chat GPT for their entire business organization. [00:03:02] Seven months in, and they're now experimenting with AI in pioneering ways that we'll talk about today. [00:03:08] If you're not from South Texas, this might be new information. [00:03:12] But the Spurs are not only known as innovators in the NBA, but their brand is synonymous with social impact and civic engagement. [00:03:19] They're widely known as a trusted anchor in South Texas, a team that shows up for families, youth, and community organizations across our region. [00:03:27] We're gonna talk a little bit about this aspect of their culture and impact today as well, and share about our collaborations, the collaborations between the Spurs and the OpenAI Forum, and what we have planned on the horizon in the South Texas community. [00:03:41] Today we'll hear from Jordan Kolosey, VP of Business Strategy, Innovation, and Data Operations, and Charlie Kurian, Director of Innovation & Strategy for the San Antonio Spurs. [00:03:53] We're bringing on a very special guest as well, Johnny Rodriguez, the Chief Innovation Officer at Fresh Consulting, who's partnering with the Spurs on a fan-facing activation leveraging SORA for this session. [00:04:05] I think we're gonna get a sneak peek of that SORA activation today as well. [00:04:09] We'll be the first people in the world to see it. [00:04:11] So, without further delay, please help me welcome Jordan Colosi and Charlie Kurian to the OpenAI Forum stage.

[00:04:21] Hi, Natalie. [00:04:22] Hi, Charlie. [00:04:23] Hi, Jordan. [00:04:24] First, I wanted to show you, I'm gonna step back so you can see it, but thank you for my swag. [00:04:29] You look good! [00:04:30] Luscious! [00:04:31] Cold! [00:04:33] Representing San Antonio Spurs! [00:04:36] Thank you for this beautiful swag. [00:04:38] I'm gonna wear it everywhere. [00:04:41] Okay, fellas, let's just jump in. [00:04:44] So, Jordan, I know that values, like, everything revolves around values for the San Antonio Spurs. [00:04:52] So, how do the Spurs' core organizational values shape the way you're approaching change, experimentation and adopting value?

[00:04:58] experimentation, and adopting new technologies like AI. [00:05:02] Yeah, absolutely, and thanks again for having us. This is a lot of fun.

[00:05:07] Listen, this space is moving fast, it's accelerating. It was no different when we started our journey than it is right now. Just the possibilities are endless, and I think a lot of folks are excited, but still uncertain about what's going to happen. I think in these moments of uncertainty and when things are moving quickly, I think we as an organization, what I've personally learned since I've been here is it's really important to kind of anchor and tether yourself, as you step into the unknown, to your values and what your culture is, especially as we are working with a brand and an organization that is very intentional with its values, and we are helping guide our 500-plus people in our organization who are each individually being confronted with the new opportunity of AI.

[00:05:57] And so as we step into the uncertainty and we meet the moment, really anchoring on our values of integrity, success, and caring, and how does that show up in the decisions that we make and what we prioritize in our strategy moving forward? I think they really manifest in three key ways. I think it's being human-centered, thinking about the people who are listening and they're reading all of the stories about the impact that AI is going to have on them personally while also being intrigued personally with their own creativity and what AI is enabling them. How do we help foster that creativity and empower them to have and build a culture of excitement and curiosity versus one of fear and kind of retractment and holding on to information that's the last thing that we want. So we want it, that's, human-centered is a big part of it.

[00:06:48] I think another part of Spurs culture that we really learned is not skipping steps, being very intentional about the phase of the journey that you're in. The Jacob Rees stone cutters creed is an integral part of essentially, you do the work, you pound the rock, and eventually when the rock splits, you know that it wasn't that one hit that did it, but all that had come before, right? So understanding what phase of the journey that you're in, knowing where you want to ultimately end up, focusing on the work at hand, and then really being anchored in what our vision is. We've set organizational strategic goals that we're aligning to. How is AI helping us reach those goals? So we're not just doing AI for AI's sake, not doing technology for technology's sake, we're using it for meaningful things that are driving our business forward.

[00:07:42] And as the technology shifts and changes in unpredictable ways, we know that we're getting wins toward a predictable path, towards a strategy that we had previously laid out, and then we can take advantage of any other opportunity. So, again, I think just being principled, being disciplined in moments of uncertainty is something that I think has really been imparted on us as we've joined this organization.

[00:08:09] Thank you so much, Jordan. And honestly, we've watched this entire process of adoption from the outside, as OpenAI stakeholders looking in, and really the way that you approached integration and adoption is a masterclass. We recently hosted the CEO of Stack Overflow, Prashanth. And he mentioned that new technology adoption in enterprise businesses, it's never really a technology problem. It's a human problem. And it feels to me like you guys have really set a standard, and that's one of the reasons why we wanted to tell this story as well.

[00:08:46] So, maybe this one's for you, Charlie. You've emphasized a human-centered approach to AI adoption. What does that look like in practice inside the SPURS organization?

[00:08:57] Yes, for us, it started with understanding the people at the center of it. And that meant let's do an org-wide observation and hearing from people. We asked four different things. What is the current usage? What is the desired usage? What are the concerns, questions, and where do they need most upskilling? And that single act was foundational to how we approached it forward.

[00:09:21] The biggest learnings from there were, 14% of our staff at that time were using it, but over 40% of people wanted to use it without direction on how to go about it. People had a lot of concerns. None of this is foreign for probably forum folks. Things like, is it safe? What is the privacy implications? What can I actually input? Is it gonna take my job? All of those questions we welcomed, and yet people still said, we wanna still learn. And that's where we really figured out that we needed to partner with an organization that prioritized and aligned.

[00:09:56] that prioritized and aligned with the culture building as we did.

[00:10:00] So we actually explored a plethora of organizations all the way from out of the box, pretty amazing OpenAI, and some of the other tools, all the way to building actually our custom tool.

[00:10:12] And ultimately we landed on partnering with OpenAI and you guys because you guys, we were surprisingly aligned on the mission. And also you guys had dedicated support around it.

[00:10:24] And that was the part where it started. And we ended up co-creating these learning modules for our entire staff together.

[00:10:30] I love that. And then, just maybe two or three months in, Wade and Sarah and more of the OpenAI team came to San Antonio and you guys hosted a hackathon for your entire business org.

[00:10:41] Oh my gosh, that was so fun. And we were at the practice facility and everybody had already begun at that point.

[00:10:51] I think Charlie, you had been asking every week for use cases and tell me about that, like what led up that hackathon. Because I know that you weren't super structured. You weren't telling people, you know, use AI, have an impact now.

[00:11:05] What was the message that made the barrier to entry and just exploration feel possible, accessible to your team?

[00:11:14] Yeah, I think we were focused on, we generally, organizationally like to approach things in a human-centered way. And as a part of that is understanding behaviorally how human beings process and learn.

[00:11:22] And one of the core things are different people learn in different ways. And if we wanted everybody to be able to use this, one, we needed to not be prescriptive. This is exactly how you should use it. This is exactly the ROI you should drive.

[00:11:39] Our job was to create pathways for every learning style to consume and learn. And our job was to remove all of the friction for the first person to press the first button.

[00:11:54] And I always refer this as, I came from Apple. So I refer this as your first iPhone box opening moment and feeling like, wow. And how can we invoke that wow moment?

[00:12:06] The hackathon was very intentional as in there were a lot of people that learned by cross-pollination. So there are people that were sitting at tables with their teams.

[00:12:15] And quite honestly, they were probably tired of hearing from Jordan or I. But hearing their peers or even their team thinking new and creative ways was incredibly inspirational and slightly more disarming than Charlie Jordan or open AI teaching them and them having to produce or them having to take their first step.

[00:12:34] And I remember specifically about that hackathon. I mean, the use cases were just so creative. And you could tell when people were able, this is a lot of this we've learned with our faculty forum members as well, they're leveraging AI for pedagogy to make it meaningful, like to render the content that they're trying to teach very tailored and meaningful to the stakeholder.

[00:12:59] And I also saw this come through on your team. I remember what somebody on your people team in HR, they created a custom GPT to help them write their annual review, like to help them outline why they should get a promotion.

[00:13:14] And I really loved that because it was a mix of personal and professional development, but also if you have to meet certain criteria and drive impact to get a promotion, then the GPT can help you, like to help you understand what you needed to be hitting criteria-wise to get that.

[00:13:32] And then to tell the story. What were some of the other awesome use cases that came out of that process that ended up being meaningful, that maybe you formalized and are in place now to drive business impact?

[00:13:45] Yeah.

[00:13:46] Yeah.

[00:13:47] Because we led with culture and democratizing access, our early use cases were actually intentionally super simple.

[00:13:55] We didn't start with advanced automation or any major technical lifts. We focused on building confidence and showing every single person that regardless of your role or your tech background, this is the tool that can help you immediately.

[00:14:08] So in the beginning, people actually used, to your point, that felt more natural, very natural. So part of that was writing supports or emails, recaps, et cetera, brainstorming.

[00:14:21] So it may not be brainstorming for your actual work in terms of creative activations or best ways we can deliver for a partner or even brainstorming support of how I can get a promotion where it lines up and how do I vouch for myself?

[00:14:36] To summarizing decks, white papers, or even cleaning data, et cetera, et cetera, et cetera. Those were important because those early use cases really built the trust and they showed the entire organization that this wasn't a tool for data people and that it was for every single person.

[00:14:52] But once people felt comfortable, the organization...

[00:14:54] People felt comfortable, the organization started asking a new question. [00:14:57] And the question was, how do we make this specific to our problems and our opportunities? [00:15:02] And that's when we moved from generic prompting to actually starting building custom GPTs that were grounded in first specific data needs and workflows. [00:15:07] And that unlocked a completely different level of value for us.

[00:15:11] So some of the GPTs that started then, that have kind of carried on and we've kind of formalized is something called a fan voice GPT, where we have tons of fans and our fans are great at communicating to us how we did. [00:15:20] Until now, this was hard to be able to understand how to make this actionable. [00:15:27] Now AI is actually really good at looking at all the information and being able to go through some proprietary ways of filtering that's internal to us, and then actually prescribe to us this is the best ways to make it actionable.

[00:15:34] We have a deep understanding of our fans in terms of how they move around the concourse, etc., and in excess of 9 million data points. [00:15:39] That was one person's full-time job of just crunching through and making actionable insights. [00:15:43] This is where AI has been really helpful in coming in and taking all of that across two seasons worth of data and now letting us know, here's how we can personalize the experience for a Gen Z person that's going to be coming in towards the holidays. [00:15:47] Those are some of the early ones that started on that have kind of carried along and have been formalized.

[00:15:54] I love that, and I know that you guys care deeply about your team as well. [00:15:58] Charlie, do you want to give a shout out to the folks on your team? Maybe who is the person now responsible for the team you just talked about? [00:16:02] About like organizing all the feedback from fans and translating it to actionable feedback?

[00:16:06] Yeah, huge shout out to our team. I could name them: Ragvi, who's an incredible behavioral economist; Daniel, who works deeply with data; I know Andre who works deeply with merch who has been an early mover in a lot of these things. [00:16:18] Our partnership folks have been great. Honestly, our entire org has kind of rallied around, and now it's kind of become the norm, actually the table stakes, and everybody's kind of been contributing to that.

[00:16:24] And Jordan, during that first hackathon with your entire team, I noticed that during the day you were working on an AI training. [00:16:29] You were leveling up your AI literacy in the midst of the hackathon. [00:16:32] So anything you'd like to add to your, you guys' approach to identifying the use cases now that are, you're really leveling up and formalizing as core to the practices of the business organization that you lead?

[00:16:37] Yeah. I think specific to the hackathon and that use case, I think I was getting questions that I hadn't ever conceived of before. [00:16:41] And I think that's just the fun part about learning and growing and being okay, sitting in uncomfortable places, is you're going to learn something that's a question that you didn't even know to ask at the time.

[00:16:51] And so as the kind of the expert, I thought I needed to take the onus on myself and try to figure out a few things and questions that were inbound when I thought, at a certain point, I had everything figured out, we were the experts.

[00:16:58] So I think the important takeaway for me in all of that is, and the whole season we've been in, of getting people comfortable and shifting the temperature around what AI is and giving that permission for folks to use it, giving them access. [00:17:04] There's so much telling people what they can do with AI, like what can they do.

[00:17:11] There were people that were going to be using the solutions anyway, why not build a compliant environment that legal's comfortable with, that gives them the conduit and place to move forward and experiment and learn themselves.

[00:17:18] And it's really, really, I think that helps you really set the table because you brought it up earlier, if this is really the change at the scale of electricity, right? If this is going to change the human experience at that level, it's going to impact everyone in some capacity.

[00:17:24] And so when you go, you don't go to an order, there's a lot of questions of like what's your AI strategy right now? [00:17:29] There's not very many questions of what's your organizational approach to electricity, right? Like, it's just, that is the table stakes that you're operating at.

[00:17:35] And so, kind of allowing everyone to experience and not staying there. I think this is where we move to like not skipping steps, being aware of where you are, having the KPIs of usage early on and adoption and all of those things. [00:17:43] Then, as you've given that permission, moving to, I think what Charlie was alluding to a little bit more of, measuring the actual hard use cases of how is this driving our business forward. So, those KPIs.

[00:19:52] So those KPIs might evolve from usage to something more along the lines of hours saved, right? Like this is how much more efficient I'm being because I found a use for this tool that makes me more efficient in my role. And then as you move beyond that, then you might shift to like actual revenue driving, right? But that's three or four steps down the road of KPIs after you've kind of moved through not being stuck in one place.

[00:20:20] And I think this is where it's super interesting that that MIT study that came out a few months back, or the Gen AI divide. And I think the findings were only 5% of organizations drive meaningful revenue within, you know, however, the immediate return of launching AI initiatives and a lot of the narrative around that was 95% of pilots fail, right? Only five actually produced. But like, is that a question of the pilots failing? Or is that an issue of understanding what phase of the journey we're in? What is the adoption of this electricity within the organization? And what is the arc of that, right?

[00:21:05] And so I think just being intentional, giving yourself permission to be where you're at in that journey, and setting ourselves up and but also feeling that pressure and momentum to mature and find those meaningful use cases after you've kind of created an environment in the culture that's curious and excited about the new technology that is among us.

[00:21:31] Jordan, I love all of that. And it just really resonates with almost every single conversation we've had about adoption in the forum for the past two and a half years. So I hope that I hope that the recording of this session spreads far and wide because that advice that you just gave, it maps across industry. And I also hear this sort of slow down to speed up because of the scale, the magnitude of this transformational shift to this technology.

[00:22:05] There's so much work to be done at the cultural level to lay the foundation. You guys have done such an incredible job, and I'm really excited to boost the visibility of the job that you've done. So before we jump into, I know you guys have some really amazing fan experience use cases on the horizon. It's going to be leveraging OpenAI technology. We're going to dig into that and we're going to dig into how that maps directly to your core values and principles.

[00:22:32] I also just wanted to surface that during that full day of the Hackathon in San Antonio, you guys also had Wade and I sit down and give an AI literacy, like chat GPT 101 course to maybe 30 nonprofit leaders from across San Antonio. So tell us, please. I am not sure who is the best stakeholder to touch on this. Why did you invite OpenAI to do that piece? It doesn't really seem like there are a lot of parallels between the NBA and basketball and nonprofit leaders, but you guys capitalize on the opportunity. OpenAI is here, let's make sure that our nonprofit leaders have an hour with them. What was that all about?

[00:23:19] Jordan, why don't you start? I'll tag along.

[00:23:22] Yeah, absolutely. I think this goes back, so we talked about being human-centered, we talked about not skipping steps, but then now also, how are we using this new technology to achieve our business goals? Yes, we have championship teams as one of our strategic objectives both on the court and off the court. We need to be successful there. We also have financial strength and there's a lot of business applications, obviously, that we'll talk about and exciting things we can do for our customers and fans.

[00:23:48] But equally as important to our success criteria as an organization and the role that we play in our community, et cetera, is our impact. We have KPIs that we need to hit, et cetera, et cetera. So really, I think driving and using this as an opportunity to only use the resources and the access that we're able to create to empower our staff and our team, but also how are we being a conduit to not-for-profit leaders in the community whose budgets are as tight as they've ever been, and now they can figure out how to be much more productive within their own capacity to do better work.

[00:24:33] And one of the taglines for that is we build leaders and legacies in our community, and so we're building up our leaders so that they can be amplified in their work. And then that becomes exponential, the work. So it was a no-brainer for us. Not only is this a new conduit to...

[00:24:50] Is this a new conduit to provide this service and empower our leaders so we can drive towards our KPIs, but it again has this exponential effect. And I think the other kind of very cool thing, we built up a practice facility very recently, opened it up, called the Rocket Lock and Tear of Pound the Rock, all right. But the old practice facility, we didn't sell it. We've decided to convert it into a community hub where basically all of our community activity, programming, et cetera, can happen out of the centralized location. Because again, we have very aggressive impact goals we have set for ourselves and we needed a centralized place to do that.

[00:25:34] This AI has become, out AI for our community has become one of the foundational curriculum pieces of that institution as we're setting it up and building it out. So we're, it's kind of becoming the foundational element of our community impact. And so just another way that I think this opportunity, the resources, the access that we've been given in our partnership with OpenAI, we're trying to get the most and hit as many of our business goals as possible because it is just an exciting time given that we've been given the permission, the access and the education to utilize this technology.

[00:26:17] The only thing I may add is, I think it's important to highlight that this, we leveraged the time that you guys were here to start the process, but this process has been, this was not a one time done in, done and over with, we've kind of continued this journey with non-profit leaders ever since. We actually hosted another one two days ago, shout out to John Mansur on my team who's been leading that effort so thoughtfully. And just to show that basketball and all the sports, soccer, and all the events that we host, there are means to an end for us. And the end really is, as our purpose says, to inspire humanity through sports and entertainment, and it's bigger than basketball.

[00:27:00] And just like that, we're carrying that same message with AI, AI is a tool, a tool that is incredibly powerful and beneficial, but at the end of the day, this is not something that we will not just use for our organization to get better. But we are going to use our influence and the space that we occupy to make sure this can be used for good. So with nonprofit leaders, with parents, with fans, with any and every single stakeholder. Yes, and we have awesome plans that we're developing right now to continue to collaborate OpenAI and San Antonio Spurs in that practice facility on a few mini jams. AI literacy for youth sports, parents, and also those nonprofit leaders in San Antonio.

[00:27:45] And Charlie, if anyone who's watching from the other sports organizations, for instance, the Miami Dolphins are here, who can they reach out to on the San Antonio Spurs team for a sort of a playbook around social impact? Who's leading that arm of the organization?

[00:28:01] Yeah, Patricia, who's our chief impact officer is great, but her time is really busy. So maybe Jordan and I can steward that relationship. This way, it's thoughtfully done. Hit us up.

[00:28:13] Perfect. Okay, now let's jump into where are these initial use cases? You're laying the groundwork for AI adoption, comfort literacy, you've identified an area that aligns with your core principles and it's fan engagement. So Charlie, maybe you can tell us a little bit about some of the use cases that you've brainstormed and you plan on standing up over the course of the season of home games in San Antonio that are leveraging AI.

[00:28:41] Yeah. Thank you for that, Tia. We kind of approached this in phases. So we've kind of established a phase about foundation setting or if I'm using basketball terminology for anybody that plays basketball, we are learning how to dribble a basketball as an organization with AI before we should like Steph Curry. From there we kind of...

[00:29:02] Or Victor Romagnolo.

[00:29:03] Or Victor Romagnolo.

[00:29:04] Or Victor Romagnolo. Pick a player. That may be actually through the roof and that could be super advanced.

[00:29:11] Yeah. The second step is to create these high impact use cases while the culture is right. So that's actually the area where it's going to thrive. We moved from there to starting automating operations and then ultimately we actually infuse it into our products. That's where we actually bring this to improve our fan experience. Uniquely thinking about where AI is beneficial. It can be fun. It's cool. But it also really gives an ability to personalize the experience in a meaningful way.

[00:29:39] So we've actually piloted this dating back to maybe about a year, 12, 15 months ago in terms of what does it look like to have personalized messages out to our fans?

[00:29:48] have personalized messages out to our fans, which is cool. I would say one of our largest click-through emails in the Spurs history has been through some of those efforts. But we're actually thinking of something really unique. Thanks to OpenAI and Sora launch, I think our brains have kind of exploded in terms of the possibilities that are available. And we've been able to think about how do we actually have a lot of images of our fans interacting and having fun throughout the games. How do we enable them to personalize that and change that up, align with, let's say if it's a theme night, if it's a Disney night or a Peanuts night or a Star Wars night, how fun would it be that Natalie is able to personalize herself to whatever character that she prefers? And what a fun way to engage. And we've also been able to think about this may be actually a really great opportunity for us to deliver additional value to our partners, create these activations and ways of engaging, which is fun and new and innovative and hasn't really been able to be done because the tech just hasn't existed. I will only say that because I don't want to ruin the surprise. I will actually give it back to you, Natalie, because we have the team that we're partnering with to bring this to life, to show a peek behind the curtains.

[00:31:08] Yes, so that's so exciting. Just to close that loop, I wanna say, I wanna make it clear that the way that the Spurs are taking on AI's option and partnering with OpenAI is also leveling up AI literacy for everyone, all of their fans. How big is your fan audience, Charlie? How many people come to a stadium? Every game, you can expect about 18,000, times that by 40 to 44 home games. And so if nobody's ever downloaded the Sora app before or even used it, this is the first opportunity. Oops, I think I, okay. So before we spoil the surprise, let's bring on Johnny Rodriguez, Chief Innovation Officer of Fresh Consulting to talk about the new creative experiments that the Spurs are undergoing with the Sora app.

[00:32:05] Hello, Natalie.

[00:32:06] Hey, Charlie.

[00:32:06] Hi, Charlie. Welcome to the OpenAI Forum, nice to see you.

[00:32:11] Yes, thanks for having me. So Johnny, can you please tell us how did the Spurs reach out to you? What was the problem that you've been engaged to solve? Like just, and why? Why did they reach out to you, Johnny, and Fresh Consulting?

[00:32:28] Great questions, yes. Let's see, where to start here? I'll kind of keep this short a little bit, but yeah, I mean, I think as a lot of the folks that are probably on the call know, the Spurs is the NBA's most well-known and far-reaching team. They've been an organization that's been at the forefront of emerging technology. I think we saw that pretty early on in our kind of relationship and partnership and working on a few other fan-facing and internal-facing kind of opportunities where we leveraged emerging technology. So they really, really are radically kind of reshaping how teams unlock the potential of emerging technology like generative AI. That was really kind of appealing to us. And of course, as you've seen, hearing Charlie and Jordan here, they're amazing to work with. We've loved the collaboration and the actual people, some of the people that were even mentioned by them.

[00:33:27] And so there's just a lot of overlap in that spirit of research and development, tapping into this emerging technology. So we have a few years of working together on a few different types of emerging technology, including spatial technology with Apple Vision Pro and other kind of VR, AR kind of opportunities. And so this was just kind of another seamless opportunity to kind of dive in, to see how we could tap into the technology that OpenAI has been working on. And yeah, definitely agree with the excitement Charlie shared about the SORA by OpenAI app and how that kind of triggered a lot of opportunities to be able to kind of deeply reconnect with fans.

[00:34:08] I love that. And I know early on Charlie shared that they were working on a chat cheating team that was gonna help fans navigate the arena landscape where they should go gift shopping, where they should get their favorite snack. And now we're evolving it to these themed nights like maybe transform yourself into your favorite Hello Kitty or Disney character, or perhaps, I don't know, the ideas are endless. But Jonny, would you like to give us a sneak peek of what specifically you're working on in relationship to the Sora application?

[00:34:43] Absolutely. Yeah, I would love to do that. Great.

[00:34:46] Great, so we're looking here at three quick points that I want to cover as we think about the kind of creative process to experimenting with this technology and ultimately getting to something that goes external facing.

[00:35:00] We do have a process to everything that we do and so, you know, our creative process does start with, you know, understanding and researching and kind of organizing some of the structure. I thought it'd be helpful to kind of look under the hood a little bit and to how we actually tap into the open AI ecosystem to help with this creative process.

[00:35:18] So, I actually want to start with that first bullet point here of, you know, using a custom GPT for prompting in Sora. I'll dive into that here, but at a high level this is allowing us, you know, as we think about the idea of, hey, if we're going to be creating these Sora kind of enabled videos, how do we get some consistency around this? How do we get some creative control?

[00:35:43] And I think that's something that really stood out to us from the very, very beginning with Sora is the fact that there is a lot of cutting to different angles. We can be very, very specific. There's actually quite a lot of good documentation that open AI has put out. This is an example of one of those, the Sora 2 prompting guide that came out in October that talks about, this is about 12 pages of content that talks about how you can prompt and how you can go from something more basic to something a little bit more advanced.

[00:36:12] We basically took that and turned that into a document, exported. We also fed it cinematography terminology, which we know Chad GPT already has familiarity with it. But by being able to have these two knowledge files, these two PDF files, we then put that into a GPT. Very much like the GPTs that were talked about earlier for saving time and bringing more productivity and things like that.

[00:36:42] These tools are also part of the creative process, and so this is just a look behind the scenes into a GPT we created called Sora2 Long Scene Prompting. I'm just going to show some of the prompting that was used in this for the instructions, but it's basically saying, look at those two files and structure out a prompt that will fit within the constraints of the Sora2 API with the character limitations, assumptions of timing and being able to include things like camera details, dialogue, sound, sound effects, which are all things that Sora2 handles really, really well.

[00:37:16] You can see that those two files were added in my knowledge files, so Sora2 best practices and the cinematography terminology, and we know GPT 5.1 just came out, so we made sure that we have this recommended model for thinking. And so ultimately when you start to put that all together, you can get to something like this, where you give it something really short, say, hey, I want to create something like this.

[00:37:36] And again, to personalization, this is just showing one of the use cases kind of explore, but you can see that it took a small prompt and then it brought out a very, very detailed within the constraints of Sora. You can kind of see here, 1250 characters tells you how many tokens will be used. You know, it's in that Sora2 prompt comfort zone and you can see all the details.

[00:37:58] So almost second by second, it's saying, okay, here's the camera angle that's going to be leveraging that document. The action that's going to happen, the lighting, the sound, the sound effect. And so what you kind of end up with is again, going from something small to something longer. And that allowed us to do a lot of testing.

[00:38:19] You can imagine how you can kind of start to scale that and say, hey, let's tweak it. What if it's, they added smoke as they're kind of walking down the hallway, or what if we change the lighting position of things? You can see that some of the things that are underlined here are kind of a reference to the character. This is also talking about, this is mentioning a jersey with number one.

[00:38:41] So there's different parameters that can kind of be tweaked. But as part of our creative process goes here, we like to test. And one of our principles is paper is cheaper than code. It's a principle that we have, one of our kind of UX principles. And so that was us saying, how can we iterate quickly?

[00:39:04] And so kind of Sora, the Sora by OpenAI app was a great place for us to kind of quickly test some of those video prompts. So you can see here on the right hand side, this is just a screenshot, but this often will end up in our drafts. So it doesn't end up being something we publish, but you can see that we'll do some early testing.

[00:39:17] There is a difference between the Sora by OpenAI app and some of the limitations and timing and quality outputs and things that you can do. It's a really, really great place to test for anybody that hasn't already been in there. And that prompt that you just saw is the direct prompt that was used for some of our tests.

[00:39:33] And we would tweak, you know, little by little to kind of see, OK, what happens if we took it this way or what if we change this angle. And as you can imagine, again, fan facing, depending on what the use case might be, right? If we wanted to be able to.

[00:39:44] might be, right? If we wanted to be able to create a dropdown that shows, hey, select the jersey number you want to have, or hey, select the theme or select, you know, those are parameters that can kind of be checked, changed to kind of update in the prompt. So ultimately we end up with an output that's consistent, that's personalized, and it really takes this concept of taking a photo from a fan and then turning it into, you know, something that's very, very specific.

[00:39:56] So I'm going to tab over to a quick example of what that output looks like. And you'll see kind of two examples of this. So let me share my tab for that. And this is just a 10 second version. I click play and it'd be signed so we can hear a little bit of all the camera angles, the sound effects, and the audio. And I'll show you just a quick second one that we did where we tweaked a few other parameters.

[00:40:43] And so you can kind of get a sense for the ability that we have from a creative direction standpoint of kind of specifying those details, right, of saying, hey, we wanna tweak some of the parameters. And so it makes it really, really, really powerful for us to be able to kind of, again, have the consistency in the video that gets created, consistency on the sound effects, the camera angles, branding, those types of elements are all kind of part of that.

[00:41:15] And, yeah, we're excited to see how, you know, as we start applying this, we can see there's gonna be a lot of opportunities to kind of bring that personalization to the fans. And just to kind of echo what, you know, we heard Jordan say, and some of what Charlie was saying, you know, there are elements of this, as you can tell, that are very human centered, and making sure that the fan is at the center of that, you know, and as these experiences get built, this open AI technology, specifically the Sora video kind of technology allows us to really bring that personalization and be human centered, and not skipping any steps along the way, being really intentional about that creative process.

[00:41:53] Again, kind of reviewing there. We looked at that custom GBT that was created for some of that prompting and consistency, testing early in the process, again, in this particular case, using the Sora app, and then kind of getting to some high quality outputs for consistency past that. Obviously, there's a lot more to that process, but I thought that would be helpful for some of the demo today.

[00:42:14] Yeah, totally. I love that, Johnny. How exciting. And so for the Ultra Drip Tunnel, Charlie, can you give us a little bit of history about the drip tunnel? And then I'm also curious, will fans be able to impose their own image in the drip tunnel? Or what does the actual personalized engagement look like?

[00:42:35] Yeah, drip tunnel for all the sport fans who are probably already familiar, this is actually those cool pictures you see of all the players walking with their amazing drip. And Ultra has been someone that has sponsored that for a long time. And the hope here is, and the goal is, how can we bring this exclusive experience that fans love to consume and see and dream about potentially walking down there one day, how can we mass scale it and bring it to everybody so that like Johnny showed, I can just take a picture and that will allow me to create my own version of a drip tunnel walk without needing to actually be in that space?

[00:43:14] I love that. That's so exciting. Great job, Johnny. It seems like this is the perfect collaboration. Okay team, so we're actually right on time. We're gonna take a few questions from the audience. And let's start with Daniel Green. He's here all the time. It's so good to see you, Daniel. From Kansas City, he asks, this is for Charlie and Jordan. You last joined the forum in February, which, in AI times, feels like two years ago. What's changed most in your workflow or approach to chat GPT since then?

[00:43:48] I can start. When we last joined, we were at our inception stages, going through the process of driving adoption. Since then, what has changed for us, we're at an amazing place internally where using AI is something celebrated and something that is talked about. Candidly, I sit on my desk and I smile when I hear somebody 10 desks across saying, hey, why don't you try using chat GPT for that? Kind of have this proud dad moment a little bit.

[00:44:20] So that's incredible, and I think the ROI of that has been really good to see efficiencies gained and people talking all the time about how you use it for creative brainstorming. When I put these pitches together, we even heard that yesterday, but more than anything, we've crossed the threshold in terms of adoption, which has kind of set a culture for us of how do we use AI, how do we prioritize and how do we leverage this?

[00:44:42] and how do we leverage this? That is now enabling us to go deeper into use cases. So the itinerary building agent that Natalie referred to is something in the works of that. And along with a couple of other agentic workflows, we're now able to partner with departments and actually create these high-impact use cases. And they're actually receptive to us coming in and creating this and helping them. And that's the nod to the culture that's been created. If that wasn't there, we would have been in a place where all of those concerns initially being said about, is it gonna take my job? Why should I collaborate here would still be relevant. So culture has enabled us to partner and create high-impact use cases. And that's kind of where we're diving deeper.

[00:45:20] Yeah, the only thing to piggyback on there, I think simply put our KPIs and how we're judging success has shifted. I think we were trying to measure how much usage the tool was getting initially, probably back in February. Now, as I think is being alluded to, this is moving to the work that Johnny's doing. Are we able to drive incremental revenue and create new, sane experiences? So it's come a long way. And I think another use case that is small and in its infancy, one small leap for AI, but big, giant leap philosophically, I guess. We five-coded a Spurs version of an arcade favorite I'll say, and we launched it in our app. It drove 700 new sign-ups for our database and people had a blast with it. And it was just five-coded in 20 minutes. The process to get it live and in the app and authenticated and all that. I think we've created the process, created the framework for onboarding the agent that the itinerary planning agent that Charlie alluded to. So some pretty cool wins there.

[00:46:24] That's awesome, Jordan. And what was the old, you said archaic. Like what was the old thing that you guys were bipoding to make more relevant today? We joke internally. I guess we can't get a cease and desist because we've already deprecated the game. But it was Dribble Dash, which was a version of Flappy Birds only you were playing with the basketball kind of navigating. It was actually very hard. It was two different things.

[00:46:50] Take it down a notch, John. For everybody who's joining us here today and those who will also just watch the replay, December 3rd, if you're interested more about bi-coding and bi-engineering, we're gonna be doing demos actually with one of our software engineers, Aaron Freel, led by our head of developer experience, Roman. So if you wanna get a behind-the-scenes look at how the Spurs might have done that, December 3rd will be a good day for that to be demoed.

[00:47:19] Okay, so this is from Andre Holtz, AI tech lead from EPAM Systems. If you could change one rule about basketball, what would it be, Jordan?

[00:47:34] One rule about basketball? Jordan, you are a collegiate basketball player.

[00:47:38] I was, I was, yeah. You know, that's one I was not, I was preparing for a lot of different questions, that you could put your best player in bubble wrap so they never get hurt out there, I don't know. Maybe a five-point spot, just one spot on the court randomly that you could shoot from.

[00:47:57] Oh, that's a cool one. Can I have my own version of that?

[00:48:03] Of course, please. That we would get to create humanoid versions of Wemby all across the floor, across all positions.

[00:48:11] Yes, oh my gosh. We have, my son has Wemby up in a full life-size poster on his wall right now.

[00:48:20] Okay, so Mike, the academic tech director at SMU asked, well please clarify if you mean creating a standard custom GPT by adding instructions or building a more advanced GPT that uses underlying model with your own wrapper or integration?

[00:48:38] And I think maybe this is related to business use cases but also some of the fan engagement activities we've discussed. Yes, we've kind of only approached it in a standard way. So we haven't gone much more deep in customizing that. So yeah, simple answer. Pretty standard out of the box leveraging it against uploaded knowledge base or connected knowledge base.

[00:49:02] Cause everyone, AI is for everyone. So yeah, allow everyone the opportunity to create custom GPTs. Johnny, I think I'd love to hear your voice in this because it can be brand-specific but you work with a lot of brands.

[00:49:16] So this is from Jason DeLuca. He asks with Sora you're letting fans become creators. How do you balance creative freedom and with brand guardrails to ensure that fan-generated content stays authentic to the Spurs culture?

[00:49:32] And so I'd love to hear from Charlie and Jordan on this but also Johnny, just if you could give us a more broad perspective on how other clients including the Spurs are.

[00:49:40] Including the Spurs are managing to be able to balance this. [00:49:46] Go for it, Johnny, you can speak from all- [00:49:48] Sure. [00:49:49] From a brand perspective first. [00:49:50] No, that's a great question. [00:49:51] Yeah, I mean, I think that balance is really, really important. [00:49:55] So there's a lot that you can get directly with these tools from a consumer facing side. [00:49:59] What's really powerful about the API side of it is that, for example, if we wanted to have the same outro part of the video with the client logo at the end, or we wanted to have a consistent underlying music that gets added, that can be part of the pipeline that gets created in the process.

[00:50:15] So then every single video that gets outputted has that branded audio, has that branded output, the logo's flash screen at the end, and so there's a lot of those types of things. Also from an experience standpoint, there are things that we could put some constraints around. So we're not getting a completely random, unexpected video at the end of it. Again, we're getting this, hey, 99, 98% consistent type of video, but they're unique. There is a one of one for every video that gets generated. And again, Sora really allows for that kind of flexibility around that, and there's some creativity that we put into that process.

[00:50:52] When you start to pair the API with kind of best practices from a user interface, user experience side of things, you can get to something that is hopefully in that balance of managing the personalization, but also keeping the consistency in the brand.

[00:51:08] Yeah, and I think the only thing I'll add, so from an app standpoint, from a technology standpoint, I know OpenAI has done a lot to give the freedom and the ability for the creators to choose, including cameos, and who is it available for, who can play with it, et cetera, et cetera, et cetera. From internal position, I think something to think about is more important than even before figuring out how do we establish brand guidelines.

[00:51:29] I think I'm focused or we're focused on making sure that all stakeholders or as many stakeholders get to touch and feel and play with it, and that's where it should start. That goes to help you understand the power, but at the same time, make a decision for yourself of where it may be right to protect yourself. In the absence of having personally touched and felt, a lot of times things can be driven by concerns only or cautiousness only. I think both are equally important. So it's an encouragement for everybody to get jump in there and play with it and interact with it, touch and feel it so you can make a decision for yourself.

[00:52:13] Well, I can't wait. Are you allowed to tell us when we might as a fan be able to engage in one of these first AI activations? [00:52:18] I'm not allowed to say that yet. [00:52:24] Okay. Not yet, but we definitely have one around holiday time. So I'm going to leave it open to drive our ticket sales. Everybody come to all of the games. [00:52:34] Yeah. And between December, potentially something in December, potentially something in March. So optimize for both of those times.

[00:52:41] Awesome. There are lots more questions, but we only have one more minute. So I think we can just address this one as a team here. And it's how, again, it revolves around how the Spurs are engaging the community with AI. And I think we've touched on that quite a bit, but you invited the nonprofit leaders to the first session in Hackathon with OpenAI.

[00:52:57] And we did a tailored tutorial for them. Charlie just mentioned that just yesterday, like you're continuing to activate that AI literacy in the community. And then just wanted to share with everybody, especially the other sports teams here, that we do plan on continuing to collaborate OpenAI and San Antonio Spurs on these community activations. We're bringing OpenAI's technical expertise.

[00:53:30] And then all the comfort and that sense of belonging that is the practice space at the Spurs, that home base everybody really feels welcome in and connecting with the San Antonio, South Texas, like local communities to continue that awesome work. So please stay tuned, friends. We would love for this to be a playbook that we take to other states and other cities.

[00:53:54] And as an access point for, I think the Spurs are really a role model for how you can lower barriers to access to social capital and technological literacy in so many ways. So that's coming. There's one more question, and I'm thinking maybe friends, you can drop this in the discussion chat later for this event.

[00:54:10] And we'll just leave everybody kind of hanging, cliffhanger. But how are the players using AI to interact with fans? And how is that lowering the barrier between fans and their favorite players? And is that even something that you guys are thinking about? So I'm gonna add that question as a follow-up and I'll drop it in the link.

[00:54:38] and I'll drop it in the link or e-mail so that you guys can access it really easily. [00:54:43] But we would love to hear if there are any plans in the works around that. [00:54:48] Okay, friends! So it was awesome to host you again.

[00:54:50] Charlie, Jordan, you know how much I just absolutely adore, respect, [00:54:59] admire the Spurs for the way that you are community anchors. [00:55:02] Thank you for being awesome partners. [00:55:04] I can't wait for 2026. All the cool things we'll be doing together.

[00:55:06] And, Donnie, thank you so much for jumping in last minute to give us this sneak peek [00:55:14] of what fans can be looking forward to. [00:55:17] And also, it was way more than I expected. [00:55:19] You really gave us a little bit of a Sora tutorial, [00:55:22] and I'd love to have you back in the forum again. [00:55:25] Of course. Really appreciate it.

[00:55:27] Thanks, Pam. All right, friends. Have a great day, and I'll see you on the court soon. [00:55:31] Thanks. Bye, guys.

[00:55:36] Yay! That was fun, and that concludes our morning session, [00:55:41] depending on where you're coming from. [00:55:43] If you're in California or the West Coast, you're here at 9 a.m. [00:55:46] If you're in Texas or Chicago, you are here at 11. [00:55:49] Thank you so much for joining us with your coffee, everyone.

[00:55:52] And then I just want to remind you that December 3rd, [00:55:54] if you are interested in learning more about VIVE engineering, [00:55:58] it's a little bit different than VIVE coding, [00:56:00] but come and learn more about it, [00:56:02] how our software engineer is now leveraging AI, [00:56:05] codecs specifically, in this new realm of coding in 2025-2026 [00:56:11] to do some of the work that the Spurs have been working on internally.

[00:56:15] And join us. We have really awesome software engineer, Aaron Friel, joining us for a demo, [00:56:20] our Head of Developer Experience, Romain Hewitt joining us. [00:56:24] So that's a wrap. It was really lovely to host you all. [00:56:27] I love seeing all of your faces in community, and we'll be back soon.

[00:56:31] I'm heading out, actually, in just an hour to Houston for the Small Business Jam [00:56:36] that's being run by the OpenAI Academy, our sister community, [00:56:39] all over the country tomorrow. [00:56:41] So I can't wait to give you updates on that. [00:56:43] Have a great day, everybody. Go Spurs go!