WEBVTT - An AI advantage for the US Open

0:00:03.360 --> 0:00:06.880
<v Malcolm Gladwell>Hello, Hello, Welcome to Smart Talks with IBM, a podcast

0:00:06.880 --> 0:00:12.600
<v Malcolm Gladwell>from Pushkin Industries, iHeartRadio and IBM. I'm Malcolm Gladwell. This season,

0:00:12.680 --> 0:00:15.840
<v Malcolm Gladwell>we're diving back into the world of artificial intelligence, but

0:00:15.960 --> 0:00:21.040
<v Malcolm Gladwell>with a focus on the powerful concept of open its possibilities, implications,

0:00:21.079 --> 0:00:24.639
<v Malcolm Gladwell>and misconceptions. We'll look at openness from a variety of

0:00:24.640 --> 0:00:28.600
<v Malcolm Gladwell>angles and explore how the concept is already reshaping industries,

0:00:29.000 --> 0:00:32.440
<v Malcolm Gladwell>ways of doing business and our very notion of what's possible.

0:00:34.200 --> 0:00:38.479
<v Malcolm Gladwell>I'm particularly excited for today's guest, Brian Ryerson. He's Senior

0:00:38.520 --> 0:00:43.080
<v Malcolm Gladwell>Director of Digital Strategy at the US Tennis Association, helping

0:00:43.120 --> 0:00:45.600
<v Malcolm Gladwell>to oversee one of the most iconic events in the

0:00:45.600 --> 0:00:49.600
<v Malcolm Gladwell>world of sports, the US Open. Brian sat down with

0:00:49.600 --> 0:00:54.320
<v Malcolm Gladwell>Pushkin's own Jacob Goldstein, host of the podcast What's Your Problem.

0:00:54.360 --> 0:00:57.600
<v Malcolm Gladwell>A veteran business journalist, Jacob has reported for The Wall

0:00:57.600 --> 0:01:00.960
<v Malcolm Gladwell>Street Journal, the Miami Herald, and was a longtime host

0:01:01.040 --> 0:01:05.360
<v Malcolm Gladwell>of the NPR program Planet Money. IBM has been the

0:01:05.400 --> 0:01:09.560
<v Malcolm Gladwell>official technology partner of the US Tennis Association for more

0:01:09.640 --> 0:01:13.839
<v Malcolm Gladwell>than thirty years, and the more recent evolution into generative

0:01:13.880 --> 0:01:18.680
<v Malcolm Gladwell>AI has enhanced the world class digital experiences that help

0:01:18.760 --> 0:01:21.959
<v Malcolm Gladwell>more than fifteen million fans from all over the world

0:01:22.319 --> 0:01:26.319
<v Malcolm Gladwell>enjoy the US Open Tennis Championships. In this episode, we

0:01:26.400 --> 0:01:30.399
<v Malcolm Gladwell>will explore how generative AI is being used to generate

0:01:30.520 --> 0:01:35.560
<v Malcolm Gladwell>match insights, spoken commentary for match highlights, and postmatch summaries

0:01:35.640 --> 0:01:38.640
<v Malcolm Gladwell>at scale for fans to enjoy through the US Open

0:01:38.640 --> 0:01:43.080
<v Malcolm Gladwell>app and website. We'll explore how these AI solutions enable

0:01:43.160 --> 0:01:46.600
<v Malcolm Gladwell>the editorial team to cover more of the tournament than

0:01:46.680 --> 0:01:50.800
<v Malcolm Gladwell>ever before, bringing fans even closer to the game they love,

0:01:51.200 --> 0:01:53.680
<v Malcolm Gladwell>and we'll learn more about one of the engines behind

0:01:53.760 --> 0:01:58.440
<v Malcolm Gladwell>this AI powered content creation, a large language model from

0:01:58.480 --> 0:02:02.480
<v Malcolm Gladwell>the IBM Granite family, which is trained and maintained using

0:02:02.520 --> 0:02:07.280
<v Malcolm Gladwell>the wantonex AI and data platform. Okay, let's dive in.

0:02:08.560 --> 0:02:09.640
<v Jacob Goldstein>Brian, welcome to the show.

0:02:09.880 --> 0:02:11.320
<v Brian Ryerson>Thanks for having me. I'm excited to be here.

0:02:11.720 --> 0:02:13.240
<v Jacob Goldstein>Can you say your name and your job.

0:02:13.600 --> 0:02:16.960
<v Brian Ryerson>Yeah, I'm Brian Ryerson. I'm senior director of Digital Strategy

0:02:16.960 --> 0:02:17.679
<v Brian Ryerson>at the USTA.

0:02:18.040 --> 0:02:20.400
<v Jacob Goldstein>Some question, what's the USTA.

0:02:20.200 --> 0:02:22.359
<v Brian Ryerson>The US Tennis Association.

0:02:22.680 --> 0:02:24.639
<v Jacob Goldstein>And tell me about the USTA, Like.

0:02:24.720 --> 0:02:27.480
<v Brian Ryerson>What is it? Yeah? So, the USTA is the governing

0:02:27.560 --> 0:02:30.960
<v Brian Ryerson>body of tennis in the US. Or mission is to

0:02:31.000 --> 0:02:34.720
<v Brian Ryerson>grow the sport of tennis across the US at all levels. Really,

0:02:34.919 --> 0:02:36.760
<v Brian Ryerson>I would say we're more like a health and wellness

0:02:36.800 --> 0:02:39.160
<v Brian Ryerson>company where tennis is the means to health and wellness.

0:02:39.600 --> 0:02:42.000
<v Brian Ryerson>And then the US Open is kind of our tenth

0:02:42.040 --> 0:02:44.919
<v Brian Ryerson>pole event that happens everyear in Flushing Meadows and is

0:02:44.960 --> 0:02:47.160
<v Brian Ryerson>really our chance to showcase the support of tennis at

0:02:47.160 --> 0:02:49.399
<v Brian Ryerson>its highest level to fans all around the world.

0:02:49.560 --> 0:02:51.720
<v Jacob Goldstein>Yeah, I mean the US Open. I assume most people

0:02:51.800 --> 0:02:53.920
<v Jacob Goldstein>know this, but it's Grand Slam. It's one of the

0:02:54.000 --> 0:02:57.040
<v Jacob Goldstein>what four biggest tennis tournaments in the world.

0:02:57.160 --> 0:03:00.400
<v Brian Ryerson>Yes, yeah, every year, we especially the past couple of years,

0:03:00.440 --> 0:03:03.680
<v Brian Ryerson>we've seen immense growth and you know, we are very

0:03:03.800 --> 0:03:05.840
<v Brian Ryerson>hopeful this year and our big goals have over a

0:03:05.840 --> 0:03:08.240
<v Brian Ryerson>million fans on site during the three week window this year,

0:03:08.280 --> 0:03:10.480
<v Brian Ryerson>so it's an amazing event. I always say it's a

0:03:10.480 --> 0:03:13.359
<v Brian Ryerson>food and wine festival where tennis is the main attraction

0:03:13.760 --> 0:03:16.000
<v Brian Ryerson>and it's a really fun, unique atmosphere.

0:03:16.680 --> 0:03:19.000
<v Jacob Goldstein>How did you get into the tennis business.

0:03:18.680 --> 0:03:21.400
<v Brian Ryerson>It's a great question. It's not where I thought i'd

0:03:21.440 --> 0:03:23.880
<v Brian Ryerson>end up for especially being there for fourteen years. So

0:03:24.040 --> 0:03:27.440
<v Brian Ryerson>I was a marketing and technology major in school, and

0:03:28.120 --> 0:03:30.720
<v Brian Ryerson>I also played college lacrosse and sports was always a

0:03:30.720 --> 0:03:32.359
<v Brian Ryerson>big part of my life and always wanted to be

0:03:32.440 --> 0:03:34.840
<v Brian Ryerson>in the sports and entertainment world. I'm here from the

0:03:34.880 --> 0:03:36.400
<v Brian Ryerson>New York area. This is where I grew up. So

0:03:36.440 --> 0:03:38.920
<v Brian Ryerson>I moved back home and had a few friends who

0:03:38.920 --> 0:03:41.560
<v Brian Ryerson>worked there, and I started out more on the number

0:03:41.640 --> 0:03:43.880
<v Brian Ryerson>side of things and really digital analytics. It was really

0:03:43.880 --> 0:03:47.440
<v Brian Ryerson>the start of when Facebook and Twitter is just starting

0:03:47.480 --> 0:03:50.320
<v Brian Ryerson>and digital marketing and all of that. And you know,

0:03:50.360 --> 0:03:52.520
<v Brian Ryerson>I went to my first year so Open not really

0:03:52.600 --> 0:03:55.920
<v Brian Ryerson>knowing what to expect, and again, I think the atmosphere

0:03:56.000 --> 0:03:58.320
<v Brian Ryerson>kind of captivated me and hooked me in. And I've

0:03:58.360 --> 0:03:59.600
<v Brian Ryerson>been there now fourteen years.

0:04:00.080 --> 0:04:04.280
<v Jacob Goldstein>And so your title is Digital Director. What does that mean?

0:04:04.320 --> 0:04:04.920
<v Jacob Goldstein>What's your job?

0:04:05.160 --> 0:04:08.160
<v Brian Ryerson>Yeah, so it's an interesting one because it's tough to

0:04:08.240 --> 0:04:10.760
<v Brian Ryerson>explain to folks who are not in the weeds on

0:04:10.840 --> 0:04:13.240
<v Brian Ryerson>all things US open or even in the sports world.

0:04:13.240 --> 0:04:16.960
<v Brian Ryerson>But really I oversee all of our consumer facing digital property.

0:04:17.080 --> 0:04:19.599
<v Brian Ryerson>So that's the us open dot org, our website built

0:04:19.600 --> 0:04:22.440
<v Brian Ryerson>by IBM, as well as our mobile app. I oversee

0:04:22.480 --> 0:04:27.520
<v Brian Ryerson>our content strategy, our sponsorship integrations. So really anything consumer

0:04:27.600 --> 0:04:30.640
<v Brian Ryerson>facing that happens on the web is under my purview.

0:04:30.920 --> 0:04:33.719
<v Brian Ryerson>Even some of our new platform extensions and gaming and

0:04:33.760 --> 0:04:37.000
<v Brian Ryerson>things like that. Anything that you can physically interact with

0:04:37.360 --> 0:04:38.800
<v Brian Ryerson>is kind of under my purview.

0:04:39.520 --> 0:04:43.360
<v Jacob Goldstein>And so you've been there now for fourteen ISSU years,

0:04:43.560 --> 0:04:47.160
<v Jacob Goldstein>which in the digital world is a long time. How

0:04:47.200 --> 0:04:51.560
<v Jacob Goldstein>has that sort of digital experience of sports changed over

0:04:51.600 --> 0:04:52.080
<v Jacob Goldstein>that time.

0:04:52.320 --> 0:04:55.880
<v Brian Ryerson>Yeah, it's obviously grown digital now, is what we say

0:04:55.880 --> 0:04:58.000
<v Brian Ryerson>and what my team says. It's the number one way

0:04:58.040 --> 0:05:00.320
<v Brian Ryerson>to engage with fans that can't make make it to

0:05:00.360 --> 0:05:02.160
<v Brian Ryerson>the event, as well as those fans who are at

0:05:02.160 --> 0:05:03.760
<v Brian Ryerson>the event, and how to enrich their stay. So it's

0:05:03.760 --> 0:05:07.359
<v Brian Ryerson>really kind of you're tackling multiple fan personas. It's the

0:05:07.640 --> 0:05:11.200
<v Brian Ryerson>international fan who's staying up late to watch in other countries,

0:05:11.240 --> 0:05:13.919
<v Brian Ryerson>to the fan here who's maybe watching on broadcasts, and

0:05:13.960 --> 0:05:16.160
<v Brian Ryerson>we go on in a company and enrich that broadcast of

0:05:16.279 --> 0:05:18.560
<v Brian Ryerson>new stats and insights to the on site fan who

0:05:18.640 --> 0:05:20.839
<v Brian Ryerson>bought a ticket and maybe doesn't know what match is

0:05:20.839 --> 0:05:23.640
<v Brian Ryerson>happening on what court. We do have twenty plus courts

0:05:23.680 --> 0:05:26.000
<v Brian Ryerson>happening at a time with all different matches, so we

0:05:26.040 --> 0:05:29.240
<v Brian Ryerson>really try to help all fans navigate the US Open

0:05:29.240 --> 0:05:30.200
<v Brian Ryerson>the best way possible.

0:05:30.560 --> 0:05:32.680
<v Jacob Goldstein>And so, like, what are some of the sort of

0:05:32.720 --> 0:05:34.680
<v Jacob Goldstein>problems you're trying to solve what are some of the

0:05:34.720 --> 0:05:36.040
<v Jacob Goldstein>hard things about your job?

0:05:36.320 --> 0:05:40.440
<v Brian Ryerson>Yeah, obviously technology changes at a rapid pace, right, So

0:05:40.520 --> 0:05:42.400
<v Brian Ryerson>I think part of it is how do we stay

0:05:42.400 --> 0:05:44.560
<v Brian Ryerson>on the forefront of that, and how do we do

0:05:44.600 --> 0:05:46.720
<v Brian Ryerson>that in the best way and make the best fan

0:05:46.839 --> 0:05:50.320
<v Brian Ryerson>experiences possible and the best user experience as possible. That's

0:05:50.360 --> 0:05:53.680
<v Brian Ryerson>always kind of driving factor number one. Then number two,

0:05:53.720 --> 0:05:57.320
<v Brian Ryerson>it's understanding and listening to our fans and what kind

0:05:57.360 --> 0:05:59.520
<v Brian Ryerson>of content they want. You hear me talk a lot

0:05:59.520 --> 0:06:02.040
<v Brian Ryerson>about storytelling. I feel like there's a lot of storytelling

0:06:02.120 --> 0:06:04.039
<v Brian Ryerson>that happens around the years open that really really want

0:06:04.080 --> 0:06:07.080
<v Brian Ryerson>to bring to fans, and that can be as simple

0:06:07.120 --> 0:06:10.400
<v Brian Ryerson>as storytelling of what's happening today and what you should

0:06:10.440 --> 0:06:13.760
<v Brian Ryerson>be watching too. Maybe it's your favorite players and what's

0:06:13.800 --> 0:06:17.359
<v Brian Ryerson>going on behind the scenes with them, to even introducing

0:06:17.560 --> 0:06:20.320
<v Brian Ryerson>I want to say, the casual fans to who they

0:06:20.360 --> 0:06:22.800
<v Brian Ryerson>should be watching, why they should follow certain players, and

0:06:22.839 --> 0:06:25.720
<v Brian Ryerson>more bringing that player's story to life. Yeah.

0:06:25.760 --> 0:06:28.000
<v Jacob Goldstein>I mean, I feel like almost the whole point of

0:06:28.040 --> 0:06:31.280
<v Jacob Goldstein>sports is to create stories for us to follow, right,

0:06:31.400 --> 0:06:33.760
<v Jacob Goldstein>Like they're engineered to be stories.

0:06:33.800 --> 0:06:34.720
<v Brian Ryerson>It's exactly this.

0:06:34.720 --> 0:06:36.920
<v Jacob Goldstein>Thing is happening in front of you and there are

0:06:36.920 --> 0:06:40.600
<v Jacob Goldstein>two antagonists and the stakes are high, and you don't

0:06:40.640 --> 0:06:42.520
<v Jacob Goldstein>know how it's going to end, Like it's built to

0:06:42.560 --> 0:06:43.279
<v Jacob Goldstein>be a story.

0:06:43.520 --> 0:06:45.760
<v Brian Ryerson>Yeah, and that's the main challenge of the job is

0:06:45.760 --> 0:06:48.560
<v Brian Ryerson>you can plan, plan, plan, but once you get on

0:06:48.680 --> 0:06:50.599
<v Brian Ryerson>two players on court and you don't know what that

0:06:50.640 --> 0:06:52.760
<v Brian Ryerson>outcome is going to be, it's now sitting and waiting

0:06:52.800 --> 0:06:55.200
<v Brian Ryerson>and watching and you become a fan yourself. And then

0:06:55.240 --> 0:06:58.640
<v Brian Ryerson>it's how do you really captivate that story and how

0:06:58.680 --> 0:07:01.120
<v Brian Ryerson>do you narrate it? How do you like translate up

0:07:01.160 --> 0:07:02.000
<v Brian Ryerson>to fans.

0:07:01.800 --> 0:07:03.240
<v Jacob Goldstein>And it's like you kind of have to do it

0:07:03.279 --> 0:07:05.320
<v Jacob Goldstein>in real time, right, Like the whole point of sports

0:07:05.360 --> 0:07:06.279
<v Jacob Goldstein>is you don't know what's going to.

0:07:06.279 --> 0:07:09.200
<v Brian Ryerson>Happen exactly, and that's the excitement. And it's also there's

0:07:09.240 --> 0:07:11.120
<v Brian Ryerson>so many different types of fans. You know, there's the

0:07:11.720 --> 0:07:13.800
<v Brian Ryerson>fans who want a lot of enriched data and their

0:07:13.880 --> 0:07:16.040
<v Brian Ryerson>tennis nerds for lack of better of saying it, and

0:07:16.040 --> 0:07:18.920
<v Brian Ryerson>that they really want to dive deep into the intricacies

0:07:18.960 --> 0:07:21.440
<v Brian Ryerson>of the game, versus the casual fan who maybe just

0:07:21.440 --> 0:07:23.520
<v Brian Ryerson>wants more of this high level storyline of what does

0:07:23.560 --> 0:07:26.920
<v Brian Ryerson>this mean? Why is it important? So it's really trying

0:07:26.960 --> 0:07:29.280
<v Brian Ryerson>to figure out how to deliver that at scale and

0:07:29.560 --> 0:07:32.120
<v Brian Ryerson>really help fans get what they're looking for and the

0:07:32.120 --> 0:07:33.400
<v Brian Ryerson>type of content they're looking for.

0:07:33.720 --> 0:07:36.960
<v Jacob Goldstein>So, are there specific examples of, you know, how fan

0:07:37.040 --> 0:07:41.560
<v Jacob Goldstein>feedback has led to specific features digital features you build.

0:07:41.960 --> 0:07:45.000
<v Jacob Goldstein>Are there, like particularly popular features you've come up with, Like,

0:07:45.080 --> 0:07:46.120
<v Jacob Goldstein>what are some specifics?

0:07:46.240 --> 0:07:48.800
<v Brian Ryerson>Yeah, some low hanging fruit type things that came from

0:07:48.800 --> 0:07:52.200
<v Brian Ryerson>fan feedback. Is simple things sometimes like managing time zones

0:07:52.200 --> 0:07:53.480
<v Brian Ryerson>and when matches start.

0:07:53.520 --> 0:07:57.080
<v Jacob Goldstein>A persistent problem where those of us can work across.

0:07:56.800 --> 0:08:00.960
<v Brian Ryerson>Times exactly, and we do have like a twenty plus

0:08:01.000 --> 0:08:03.480
<v Brian Ryerson>courts happening at a time, so it's a lot to follow,

0:08:03.520 --> 0:08:05.960
<v Brian Ryerson>and how do you translate that to a fan whether

0:08:06.000 --> 0:08:08.560
<v Brian Ryerson>it's to their native language or to their time zone

0:08:08.640 --> 0:08:10.840
<v Brian Ryerson>or things like that. So that's one thing that came

0:08:10.840 --> 0:08:13.800
<v Brian Ryerson>through fan feedback, and another one a three to five

0:08:13.880 --> 0:08:16.480
<v Brian Ryerson>hour match, especially when you're having twenty plus of them

0:08:16.520 --> 0:08:18.680
<v Brian Ryerson>happening at a time, is there's too much for one

0:08:18.720 --> 0:08:22.160
<v Brian Ryerson>person to follow. So how do you start from an

0:08:22.280 --> 0:08:25.640
<v Brian Ryerson>editorial perspective really helping with that storytelling and guiding a

0:08:25.680 --> 0:08:28.280
<v Brian Ryerson>fan to like, all right, whether there's an upset about

0:08:28.280 --> 0:08:32.200
<v Brian Ryerson>to happen, or here's your matches to watch or even

0:08:32.240 --> 0:08:33.959
<v Brian Ryerson>some of the predictions we're starting to put in is

0:08:34.000 --> 0:08:36.000
<v Brian Ryerson>we really want to guide the fan before a match,

0:08:36.360 --> 0:08:39.120
<v Brian Ryerson>here's where you should tune in to even after a match,

0:08:39.160 --> 0:08:42.360
<v Brian Ryerson>of here's what's happened, here's what's important. And we're really

0:08:42.400 --> 0:08:43.839
<v Brian Ryerson>excited with some of the features we built in the

0:08:43.920 --> 0:08:46.640
<v Brian Ryerson>last few years that I would say really helps us

0:08:46.679 --> 0:08:48.400
<v Brian Ryerson>do that at more scale than what we were able

0:08:48.400 --> 0:08:50.920
<v Brian Ryerson>to do with just writers following a match and covering

0:08:50.920 --> 0:08:51.720
<v Brian Ryerson>every single match.

0:08:51.920 --> 0:08:54.199
<v Jacob Goldstein>Uh huh. So I want to talk a little bit

0:08:54.320 --> 0:09:00.480
<v Jacob Goldstein>about the partnership between IBM and the USTA. Tell me

0:09:00.480 --> 0:09:01.920
<v Jacob Goldstein>about the work you do together.

0:09:02.320 --> 0:09:05.560
<v Brian Ryerson>So IBM is our official digital and technology partner and

0:09:05.640 --> 0:09:08.600
<v Brian Ryerson>innovation partner of the US Open. They predate me. It's

0:09:08.640 --> 0:09:11.520
<v Brian Ryerson>a thirty year partnership and it truly as a partnership.

0:09:11.559 --> 0:09:14.559
<v Brian Ryerson>So I view the IBM consulting team as an extension

0:09:14.640 --> 0:09:18.120
<v Brian Ryerson>of my USTA team, So we work with them year round.

0:09:18.400 --> 0:09:22.760
<v Brian Ryerson>They design, develop and deliver the digital properties. They help

0:09:22.840 --> 0:09:25.520
<v Brian Ryerson>us provide the tools to create content to do things

0:09:25.520 --> 0:09:27.960
<v Brian Ryerson>at scale. They help us from stats and information and

0:09:28.280 --> 0:09:30.920
<v Brian Ryerson>really help us push from an innovation standpoint to make

0:09:30.960 --> 0:09:33.240
<v Brian Ryerson>sure that we are staying on that cutting edge of technology.

0:09:33.320 --> 0:09:37.800
<v Brian Ryerson>So I would truly say it's much more than a sponsorship,

0:09:37.800 --> 0:09:40.559
<v Brian Ryerson>where it's truly a partnership to deliver that fan experience.

0:09:41.000 --> 0:09:43.640
<v Jacob Goldstein>And so what are some of the specific things that

0:09:43.880 --> 0:09:45.040
<v Jacob Goldstein>you have done with IBM.

0:09:45.320 --> 0:09:49.480
<v Brian Ryerson>Yeah, so, I mean there's countless ones to talk through. Obviously,

0:09:49.520 --> 0:09:52.080
<v Brian Ryerson>they thirty years ago, they helped us build our first

0:09:52.160 --> 0:09:55.640
<v Brian Ryerson>website and it's kind of grown from there over the

0:09:55.800 --> 0:09:57.920
<v Brian Ryerson>past few years. I would say I think was twenty

0:09:57.960 --> 0:10:00.760
<v Brian Ryerson>eighteen as we started AI Highlights. That was really when

0:10:01.120 --> 0:10:04.360
<v Brian Ryerson>we were able to have all twenty matches going at

0:10:04.480 --> 0:10:07.640
<v Brian Ryerson>a single time. We were able to quickly deliver succinct

0:10:07.679 --> 0:10:10.440
<v Brian Ryerson>highlights to fans to our digital platform so they could

0:10:10.600 --> 0:10:12.280
<v Brian Ryerson>see highlights for every single core.

0:10:12.960 --> 0:10:15.920
<v Jacob Goldstein>Is that video highlights? Is that tech summaries? What does

0:10:15.920 --> 0:10:16.360
<v Jacob Goldstein>that mean?

0:10:16.880 --> 0:10:19.360
<v Brian Ryerson>At the time, it was video highlights, Okay, so it

0:10:19.400 --> 0:10:22.280
<v Brian Ryerson>was really taking that three to five hour match, let's say,

0:10:22.320 --> 0:10:24.840
<v Brian Ryerson>and cut it down to a three minute highlight that

0:10:24.920 --> 0:10:27.320
<v Brian Ryerson>could show up within moments after a match, ending to

0:10:27.400 --> 0:10:29.840
<v Brian Ryerson>our website and our mobile app, so fans could see

0:10:29.840 --> 0:10:31.520
<v Brian Ryerson>that all around the world and really kind of get

0:10:31.559 --> 0:10:34.079
<v Brian Ryerson>that three minute overview what happened in a match?

0:10:34.480 --> 0:10:37.360
<v Jacob Goldstein>And was that AI enabled. Was AI a piece of

0:10:37.760 --> 0:10:38.400
<v Jacob Goldstein>how to do that?

0:10:38.559 --> 0:10:41.520
<v Brian Ryerson>It was it was probably our first foray into AI.

0:10:41.800 --> 0:10:45.280
<v Jacob Goldstein>Back twenty eighteen is relatively early.

0:10:45.640 --> 0:10:50.079
<v Brian Ryerson>Yeah, exactly for tennis exactly. Yeah. It really I want

0:10:50.120 --> 0:10:54.400
<v Brian Ryerson>to say, opened up our ability to one again storyteller,

0:10:54.440 --> 0:10:56.720
<v Brian Ryerson>but attract new fans too. Is video has actually been

0:10:56.760 --> 0:10:59.640
<v Brian Ryerson>our number one growth area since twenty eighteen. I think

0:11:00.080 --> 0:11:01.559
<v Brian Ryerson>of that has to do with the scale of how

0:11:01.600 --> 0:11:02.960
<v Brian Ryerson>we deliver that content.

0:11:02.960 --> 0:11:05.880
<v Jacob Goldstein>Using AI and being able to deliver the sort of

0:11:06.000 --> 0:11:09.240
<v Jacob Goldstein>video highlight reels at scale.

0:11:08.960 --> 0:11:11.559
<v Brian Ryerson>Yeah, and do it quickly. Right. We've always had highlights,

0:11:11.559 --> 0:11:13.559
<v Brian Ryerson>but it was a manual process where you had a

0:11:14.080 --> 0:11:17.280
<v Brian Ryerson>video at or cutting through you know, a three hour match,

0:11:17.400 --> 0:11:19.679
<v Brian Ryerson>selecting the right scene, stitching together. It would have to

0:11:19.760 --> 0:11:22.960
<v Brian Ryerson>get voiced over, et cetera. We really have used AI

0:11:23.040 --> 0:11:24.640
<v Brian Ryerson>to make it, I want to say, much more efficient

0:11:24.760 --> 0:11:27.760
<v Brian Ryerson>and speed up that process and deliver it more quickly

0:11:27.760 --> 0:11:28.400
<v Brian Ryerson>to our fans.

0:11:28.760 --> 0:11:30.679
<v Jacob Goldstein>I mean, it would be a bummer to get scooped

0:11:30.840 --> 0:11:33.920
<v Jacob Goldstein>by whatever NBC News or Yes Pen or whatever. I'm

0:11:33.920 --> 0:11:35.439
<v Jacob Goldstein>sure there are all your partners and you love them

0:11:35.520 --> 0:11:38.240
<v Jacob Goldstein>most likely obviously you want to have the video first, right,

0:11:38.280 --> 0:11:39.000
<v Jacob Goldstein>it's your match.

0:11:39.240 --> 0:11:41.720
<v Brian Ryerson>Yeah, And I think It's also important to us as

0:11:41.800 --> 0:11:45.959
<v Brian Ryerson>being the USTA is ensuring that it's not just you know,

0:11:46.120 --> 0:11:49.720
<v Brian Ryerson>the main marquee players, that every player and all those

0:11:49.720 --> 0:11:53.200
<v Brian Ryerson>storylines and that whether it's you know, the main singles,

0:11:53.280 --> 0:11:55.880
<v Brian Ryerson>draw to our mixed doubles, et cetera. They all need

0:11:55.960 --> 0:11:58.400
<v Brian Ryerson>highlights and they all have their own stories to tell,

0:11:58.440 --> 0:12:00.199
<v Brian Ryerson>and how do we do that at scale? It was

0:12:00.240 --> 0:12:02.640
<v Brian Ryerson>something that before we had that product was not something

0:12:02.679 --> 0:12:03.400
<v Brian Ryerson>you were able to do.

0:12:03.840 --> 0:12:07.280
<v Jacob Goldstein>Great, So let's let's talk in some more detail about

0:12:07.320 --> 0:12:10.720
<v Jacob Goldstein>what you're working on. Let's start with the app. Tell

0:12:10.720 --> 0:12:13.760
<v Jacob Goldstein>me about the us Open app and the Companion website.

0:12:13.840 --> 0:12:16.080
<v Brian Ryerson>Yeah. So, so I'll start with the app, and I

0:12:16.120 --> 0:12:19.439
<v Brian Ryerson>feel like they serve similar needs, but they're a little

0:12:19.480 --> 0:12:22.760
<v Brian Ryerson>different in their own respective manners. Is the app everybody

0:12:22.840 --> 0:12:24.560
<v Brian Ryerson>has a phone in their hands at this point. The

0:12:24.600 --> 0:12:27.040
<v Brian Ryerson>app is kind of their guide to when I say

0:12:27.040 --> 0:12:29.680
<v Brian Ryerson>a million fans on site, we view the app as

0:12:29.720 --> 0:12:31.800
<v Brian Ryerson>we want that to be their on site guide and

0:12:31.840 --> 0:12:33.240
<v Brian Ryerson>Companion a million.

0:12:33.400 --> 0:12:36.080
<v Jacob Goldstein>Let's just pause on a million fans on site, right,

0:12:36.120 --> 0:12:39.839
<v Jacob Goldstein>because like a big professional whatever, an NFL game or

0:12:39.880 --> 0:12:43.360
<v Jacob Goldstein>something that's like one hundred thousand, this is ten x that.

0:12:43.720 --> 0:12:47.199
<v Brian Ryerson>Yeah, and a three week window and a very succinct, tight,

0:12:47.280 --> 0:12:51.319
<v Brian Ryerson>action packed window. There's a lot of action logistics.

0:12:51.400 --> 0:12:53.120
<v Jacob Goldstein>Okay, so keep going.

0:12:53.200 --> 0:12:56.000
<v Brian Ryerson>So the app, you know, whether it's finding the schedules,

0:12:56.080 --> 0:12:58.960
<v Brian Ryerson>the live scores, what's happening on court. That's really the

0:12:59.000 --> 0:13:01.640
<v Brian Ryerson>focus point of the app app and what we're really

0:13:01.679 --> 0:13:04.240
<v Brian Ryerson>focused on this year is how do we build in

0:13:04.280 --> 0:13:06.720
<v Brian Ryerson>some of those match summaries into the app, into our

0:13:06.800 --> 0:13:10.200
<v Brian Ryerson>Slam Tracker experience. So again, before match, that kind of

0:13:10.200 --> 0:13:12.440
<v Brian Ryerson>match preview of here's maybe if you have a ticket,

0:13:12.480 --> 0:13:15.559
<v Brian Ryerson>here's what to expect, here's you know are likely to win,

0:13:15.679 --> 0:13:18.160
<v Brian Ryerson>who we are predicting, so you can kind of get

0:13:18.160 --> 0:13:21.520
<v Brian Ryerson>some information heading in, and then after the match it's

0:13:21.559 --> 0:13:24.760
<v Brian Ryerson>more of what just happened, what it means for the

0:13:25.040 --> 0:13:27.800
<v Brian Ryerson>rest of the draw, who they're playing next, is this

0:13:27.880 --> 0:13:30.080
<v Brian Ryerson>the first time this has happened, et cetera, and really

0:13:30.160 --> 0:13:33.199
<v Brian Ryerson>enriching that experience as well. So the app is one

0:13:33.280 --> 0:13:35.600
<v Brian Ryerson>your guide to what you should be watching, but also

0:13:35.640 --> 0:13:38.199
<v Brian Ryerson>then giving you that insights and context to what's happening

0:13:38.240 --> 0:13:39.200
<v Brian Ryerson>on that court as you're.

0:13:39.080 --> 0:13:42.720
<v Jacob Goldstein>Watching, like the commentator in your pocket exactly. So you

0:13:42.880 --> 0:13:45.480
<v Jacob Goldstein>used a phrase in there as if I already knew

0:13:45.480 --> 0:13:47.960
<v Jacob Goldstein>it and I love the phrase, but I want you

0:13:48.040 --> 0:13:51.000
<v Jacob Goldstein>to talk more about it. That phrase is slam Tracker.

0:13:51.200 --> 0:13:56.960
<v Brian Ryerson>Yes, so slam Tracker is our long standing live scores.

0:13:57.000 --> 0:13:59.240
<v Brian Ryerson>I want to say match Center. It is, okay, where

0:13:59.480 --> 0:14:02.200
<v Brian Ryerson>every single data point, where every single match lives. And

0:14:02.240 --> 0:14:05.520
<v Brian Ryerson>it really it helps showcase what's happening to match. I say,

0:14:05.520 --> 0:14:08.360
<v Brian Ryerson>it's our broadcast companion. So if you're watching live, it's

0:14:08.360 --> 0:14:11.000
<v Brian Ryerson>our in stadium companion. It's also the best thing to

0:14:11.200 --> 0:14:13.000
<v Brian Ryerson>have if you aren't able to watch.

0:14:13.080 --> 0:14:14.840
<v Jacob Goldstein>And so, like, I'm on the app and there's a

0:14:14.880 --> 0:14:17.760
<v Jacob Goldstein>thing called slam Tracker, and it like taps slam Tracker.

0:14:17.800 --> 0:14:19.560
<v Jacob Goldstein>What do I see on my phone when I tap

0:14:19.600 --> 0:14:22.960
<v Jacob Goldstein>slam Tracker? You know, midday when the tournament's happening.

0:14:23.000 --> 0:14:24.880
<v Brian Ryerson>So before match, that's where you get a lot of

0:14:24.880 --> 0:14:27.080
<v Brian Ryerson>pre match content. That's where those live kind of our

0:14:27.120 --> 0:14:30.480
<v Brian Ryerson>predictions are. Likelihood to win lives within that So likelihood

0:14:30.480 --> 0:14:33.080
<v Brian Ryerson>to win essentially pulls in a bunch of data points.

0:14:33.120 --> 0:14:36.200
<v Brian Ryerson>So pass matches, how many times these players have played

0:14:36.240 --> 0:14:38.920
<v Brian Ryerson>each other against each other, Even some punditry and other

0:14:38.920 --> 0:14:41.720
<v Brian Ryerson>written articles that maybe our editorial team put out and

0:14:41.800 --> 0:14:44.360
<v Brian Ryerson>really kind of puts a prediction out there.

0:14:44.280 --> 0:14:46.400
<v Jacob Goldstein>And so it's just a percentage chance.

0:14:46.320 --> 0:14:49.400
<v Brian Ryerson>Yes exactly, but it uses millions of data points to

0:14:49.440 --> 0:14:51.720
<v Brian Ryerson>come up with that. Yes, so it really helps you

0:14:51.800 --> 0:14:54.760
<v Brian Ryerson>kind of understand what you're getting into for that match.

0:14:55.120 --> 0:14:57.880
<v Brian Ryerson>During a live match, it is every single point, so

0:14:58.240 --> 0:15:01.400
<v Brian Ryerson>point by point scoring as well as in depth analysis

0:15:01.400 --> 0:15:04.320
<v Brian Ryerson>in point commentary where also this year have a live

0:15:04.400 --> 0:15:07.360
<v Brian Ryerson>visualization that accompanies that that will really help bring the

0:15:07.520 --> 0:15:09.880
<v Brian Ryerson>match together. And what I mean by that is it

0:15:10.000 --> 0:15:13.240
<v Brian Ryerson>uses our ball tracking technology to really showcase the match

0:15:13.560 --> 0:15:16.480
<v Brian Ryerson>in near real time, so within seconds delay of where

0:15:16.520 --> 0:15:19.000
<v Brian Ryerson>the ball is being hit, where the players are, and

0:15:19.040 --> 0:15:21.840
<v Brian Ryerson>really bring a visualization to life and layered stats and

0:15:21.920 --> 0:15:22.640
<v Brian Ryerson>data on top of it.

0:15:22.760 --> 0:15:24.640
<v Jacob Goldstein>Huh. If that's sort of like when I'm watching a

0:15:24.720 --> 0:15:27.480
<v Jacob Goldstein>match on TV and there's like a close call as

0:15:27.520 --> 0:15:29.120
<v Jacob Goldstein>the ball in or out and they do that thing

0:15:29.160 --> 0:15:31.160
<v Jacob Goldstein>where they kind of show a sort of video game

0:15:31.280 --> 0:15:33.400
<v Jacob Goldstein>version of where the ball landed. Does it look like that?

0:15:33.640 --> 0:15:36.360
<v Brian Ryerson>It's like that before every single shot, So it's not

0:15:36.480 --> 0:15:38.960
<v Brian Ryerson>just those close ones. It's our first foray to bring

0:15:39.000 --> 0:15:40.000
<v Brian Ryerson>that match to life.

0:15:40.520 --> 0:15:42.360
<v Jacob Goldstein>Huh. And so what do I see on that kind

0:15:42.360 --> 0:15:44.520
<v Jacob Goldstein>of view that I don't see from whatever watching the video?

0:15:44.640 --> 0:15:47.320
<v Brian Ryerson>Yeah? So one you'll be able just to see more

0:15:47.360 --> 0:15:50.520
<v Brian Ryerson>of the ball trajectory and where the ball is being hit.

0:15:50.560 --> 0:15:52.840
<v Brian Ryerson>But then you can also start layering things in stats

0:15:52.840 --> 0:15:55.360
<v Brian Ryerson>and insights on top of that, So how many times

0:15:55.400 --> 0:15:58.520
<v Brian Ryerson>has player a hit the ball on a certain baseline,

0:15:58.600 --> 0:16:01.560
<v Brian Ryerson>how fast are they hitting it, maybe their serve percentage

0:16:01.560 --> 0:16:03.480
<v Brian Ryerson>and a certain side of the court, etc. So you

0:16:03.520 --> 0:16:05.920
<v Brian Ryerson>can really start layering in for the ones that really

0:16:05.960 --> 0:16:07.600
<v Brian Ryerson>want to dive deep into the For.

0:16:07.560 --> 0:16:11.520
<v Jacob Goldstein>The nerds, it's for the information rich exactly.

0:16:11.560 --> 0:16:14.040
<v Brian Ryerson>It's the strategy of tennis. It really should be an

0:16:14.120 --> 0:16:16.160
<v Brian Ryerson>interesting way to slice and dice a match.

0:16:16.360 --> 0:16:17.200
<v Jacob Goldstein>Huh.

0:16:17.400 --> 0:16:20.720
<v Malcolm Gladwell>It's remarkable how the USDA is leveraging AI to enhance

0:16:20.800 --> 0:16:26.080
<v Malcolm Gladwell>fan engagement and deliver immersive experiences both on site and online.

0:16:26.480 --> 0:16:32.000
<v Malcolm Gladwell>Brian's emphasis on storytelling really underscores the evolution of sports marketing.

0:16:32.640 --> 0:16:36.440
<v Malcolm Gladwell>The slam Chakra feature particularly caught my attention. It's essentially

0:16:36.520 --> 0:16:39.800
<v Malcolm Gladwell>bringing the excitement of a tennis match to life in

0:16:39.840 --> 0:16:44.040
<v Malcolm Gladwell>your palm, moment by moment. As someone who appreciates the

0:16:44.120 --> 0:16:47.840
<v Malcolm Gladwell>narrative intricacies of sports, I find it compelling how AI

0:16:47.960 --> 0:16:51.800
<v Malcolm Gladwell>helps predict and analyze matches in real time.

0:16:52.840 --> 0:16:55.240
<v Jacob Goldstein>Tell me about the AI commentary feature.

0:16:55.440 --> 0:16:59.080
<v Brian Ryerson>Yeah, I know, I mentioned AI highlights back in twenty eighteen.

0:16:59.160 --> 0:17:02.120
<v Brian Ryerson>It's now progress for us. And again if we go

0:17:02.240 --> 0:17:05.200
<v Brian Ryerson>back to before we had a highlights, to have a

0:17:05.240 --> 0:17:08.359
<v Brian Ryerson>highlight ready for the site was a video editor cutting

0:17:08.359 --> 0:17:12.160
<v Brian Ryerson>the highlight and getting voiced over and then being published aside,

0:17:12.200 --> 0:17:15.840
<v Brian Ryerson>and it took probably an hour plus for that highlight

0:17:15.880 --> 0:17:19.520
<v Brian Ryerson>to really be created. Now with AI commentary, not only

0:17:19.560 --> 0:17:22.679
<v Brian Ryerson>are we creating and cutting the highlights using our AI technology,

0:17:22.720 --> 0:17:24.919
<v Brian Ryerson>but it's now using all the data points that we

0:17:24.960 --> 0:17:27.200
<v Brian Ryerson>have around the match, whether it's our live scoring data,

0:17:27.520 --> 0:17:30.919
<v Brian Ryerson>our ball tra directory data, etc. And it's really creating

0:17:30.920 --> 0:17:34.439
<v Brian Ryerson>a script that helped storytell around that match. That's all

0:17:34.520 --> 0:17:38.399
<v Brian Ryerson>using Watson X technology and then using text to speech

0:17:38.480 --> 0:17:40.960
<v Brian Ryerson>we're able to actually then create the commentary on top

0:17:41.000 --> 0:17:44.239
<v Brian Ryerson>of that, which all happens now within minutes. So our

0:17:44.280 --> 0:17:47.160
<v Brian Ryerson>team's able to now create fully voiced highlights for every

0:17:47.200 --> 0:17:50.399
<v Brian Ryerson>men's and women's singles match to our site within minutes.

0:17:51.160 --> 0:17:53.679
<v Jacob Goldstein>So I know there's a new feature you're working on

0:17:53.720 --> 0:17:58.000
<v Jacob Goldstein>for this year called match reports. What are match reports?

0:17:58.320 --> 0:18:02.919
<v Brian Ryerson>It's our ability to succsickly tell the story of a match,

0:18:03.400 --> 0:18:06.920
<v Brian Ryerson>so everything happens in five hours within that match down

0:18:06.960 --> 0:18:10.880
<v Brian Ryerson>to a couple paragraphs. That really helps a user understand

0:18:10.960 --> 0:18:14.560
<v Brian Ryerson>or a fan understand what just happened. Again, some key

0:18:14.600 --> 0:18:18.159
<v Brian Ryerson>stats what's upcoming really help us with that storytelling. In

0:18:18.200 --> 0:18:20.840
<v Brian Ryerson>the past, when we have twenty two courts happening at

0:18:20.840 --> 0:18:23.040
<v Brian Ryerson>a certain time, we would have to pick and choose

0:18:23.080 --> 0:18:25.560
<v Brian Ryerson>which stories we think or which matches we think are

0:18:25.560 --> 0:18:27.600
<v Brian Ryerson>going to have the best stories, and that's a really

0:18:27.600 --> 0:18:30.560
<v Brian Ryerson>hard thing to predict from an editorial perspective. With our

0:18:30.600 --> 0:18:32.960
<v Brian Ryerson>Match Reports now we'll be able to have full coverage

0:18:32.960 --> 0:18:34.920
<v Brian Ryerson>of every single match during the main draw.

0:18:35.680 --> 0:18:38.280
<v Jacob Goldstein>So, of course I want to talk about jeneritive AI.

0:18:38.560 --> 0:18:41.399
<v Jacob Goldstein>How could we not talk about generative Of course, what

0:18:41.480 --> 0:18:42.840
<v Jacob Goldstein>are you working on with jenerative AI?

0:18:43.160 --> 0:18:45.560
<v Brian Ryerson>So match Reports is the prime example of it. So

0:18:45.640 --> 0:18:48.760
<v Brian Ryerson>Match Reports will be completely using Watson next genera of

0:18:48.760 --> 0:18:53.320
<v Brian Ryerson>AI technology, and really, again to us, it's how can

0:18:53.359 --> 0:18:56.639
<v Brian Ryerson>we do that storytelling at scale? Tennis is such a

0:18:56.720 --> 0:18:59.960
<v Brian Ryerson>data rich sport. All sports have data, but tennis has

0:19:00.080 --> 0:19:02.200
<v Brian Ryerson>a lot of shots and different shot types and ball

0:19:02.240 --> 0:19:06.040
<v Brian Ryerson>trajectory and live scoring data and umpire chair data and

0:19:06.080 --> 0:19:10.160
<v Brian Ryerson>crowd and all that factoring in jenera AI really helps

0:19:10.240 --> 0:19:13.119
<v Brian Ryerson>us take some of that structured and unstructured data really

0:19:13.480 --> 0:19:16.960
<v Brian Ryerson>one organize it in a way, but then help us

0:19:17.200 --> 0:19:20.359
<v Brian Ryerson>quickly tell that story at scale to all of our fans.

0:19:20.560 --> 0:19:23.239
<v Brian Ryerson>And I think we're really just starting to scratch at

0:19:23.280 --> 0:19:26.520
<v Brian Ryerson>some of the capabilities, and we're really excited about where

0:19:26.560 --> 0:19:28.560
<v Brian Ryerson>we're being, but we also see the opportunity of even

0:19:28.840 --> 0:19:31.040
<v Brian Ryerson>how we can grow to new fans and new fans

0:19:31.080 --> 0:19:33.560
<v Brian Ryerson>around the world using jennal of AI in the future.

0:19:34.840 --> 0:19:38.600
<v Jacob Goldstein>So I'm curious, and you alluded to this a moment ago,

0:19:38.640 --> 0:19:40.240
<v Jacob Goldstein>but I'd like to talk a little bit more about

0:19:40.280 --> 0:19:44.000
<v Jacob Goldstein>it because it seems interesting as a technical problem. Right,

0:19:44.200 --> 0:19:49.480
<v Jacob Goldstein>is the nature of turning tennis matches into stories, which

0:19:49.520 --> 0:19:51.919
<v Jacob Goldstein>is fundamentally what we're talking about here in different ways

0:19:51.960 --> 0:19:57.200
<v Jacob Goldstein>in different media, is about taking both structured data, right

0:19:57.320 --> 0:20:02.880
<v Jacob Goldstein>like the stats, you know, points, matches, and also unstructured data,

0:20:03.000 --> 0:20:06.800
<v Jacob Goldstein>right like commentary and articles and the kind of fuzzier

0:20:06.840 --> 0:20:10.760
<v Jacob Goldstein>parts of storytelling. And so I'm curious how AI kind

0:20:10.800 --> 0:20:14.200
<v Jacob Goldstein>of helps you manage both the structured and the unstructured data.

0:20:14.680 --> 0:20:18.520
<v Brian Ryerson>Yeah, so I think the structured data is pretty self experimentatory,

0:20:18.840 --> 0:20:20.600
<v Brian Ryerson>but when you get into the unstructured data and some

0:20:20.680 --> 0:20:22.520
<v Brian Ryerson>of the punditry. That's where you get more of the

0:20:22.560 --> 0:20:26.000
<v Brian Ryerson>opinion pieces into it, like a specific player matchup. This

0:20:26.080 --> 0:20:28.879
<v Brian Ryerson>player always plays well against so and so, or as

0:20:28.880 --> 0:20:30.720
<v Brian Ryerson>they play always played well at night, or they're a

0:20:30.760 --> 0:20:34.520
<v Brian Ryerson>fan favorite, and the crowd, you know, adrenaline and the

0:20:34.560 --> 0:20:36.880
<v Brian Ryerson>crowd being behind you can really motivate you to play

0:20:36.880 --> 0:20:40.200
<v Brian Ryerson>a lot better. So it pulls in all those unstructured

0:20:40.240 --> 0:20:43.000
<v Brian Ryerson>pieces and helps us really put some more rigor around

0:20:43.040 --> 0:20:45.680
<v Brian Ryerson>it and help add and enrich our storytelling with it.

0:20:46.080 --> 0:20:50.000
<v Jacob Goldstein>And so I'm curious when you're starting to use generative AI,

0:20:50.280 --> 0:20:53.080
<v Jacob Goldstein>you know, over the past few years, like what were

0:20:53.119 --> 0:20:54.439
<v Jacob Goldstein>your concerns going into that.

0:20:54.840 --> 0:20:59.360
<v Brian Ryerson>I think our biggest concern is ensuring that one factually

0:20:59.600 --> 0:21:01.520
<v Brian Ryerson>it is because it's only as good as the data

0:21:01.520 --> 0:21:03.439
<v Brian Ryerson>you feed in. And how do you really ensure that

0:21:03.480 --> 0:21:06.119
<v Brian Ryerson>your model's working right and that the output and the

0:21:06.200 --> 0:21:08.919
<v Brian Ryerson>data you're feeding it matches the output and how do

0:21:08.960 --> 0:21:11.040
<v Brian Ryerson>you do that at scale? So we do have a

0:21:11.080 --> 0:21:14.160
<v Brian Ryerson>lot of human intervention. That's where the IBM consulting team,

0:21:14.280 --> 0:21:16.520
<v Brian Ryerson>they're on site with us for those full three weeks

0:21:16.560 --> 0:21:19.919
<v Brian Ryerson>really helping us review everything. And we're constantly learning, especially

0:21:20.000 --> 0:21:22.720
<v Brian Ryerson>early in the tournament and I would say the other

0:21:22.960 --> 0:21:25.600
<v Brian Ryerson>big concern, again it goes around to the data, is

0:21:25.920 --> 0:21:28.600
<v Brian Ryerson>what data do we have available that is trustworthy? So,

0:21:28.800 --> 0:21:30.800
<v Brian Ryerson>you know, we are feel very confident with the data

0:21:30.800 --> 0:21:32.440
<v Brian Ryerson>that comes off of Cork, but when we get into

0:21:32.440 --> 0:21:36.160
<v Brian Ryerson>that unstructured piece, what are the right data sources? How

0:21:36.200 --> 0:21:38.560
<v Brian Ryerson>do we validate those data sources and how do we

0:21:39.000 --> 0:21:41.800
<v Brian Ryerson>ensure that they're accurate Because if the data that has

0:21:41.840 --> 0:21:43.440
<v Brian Ryerson>to go in has to be accurate for the for

0:21:43.480 --> 0:21:44.480
<v Brian Ryerson>the output.

0:21:44.480 --> 0:21:46.879
<v Jacob Goldstein>So how do you do that? That's the concern? How

0:21:47.200 --> 0:21:48.000
<v Jacob Goldstein>how do you address it?

0:21:48.200 --> 0:21:50.800
<v Brian Ryerson>Yeah, so I think there's a number of tools that

0:21:50.880 --> 0:21:53.560
<v Brian Ryerson>we use all within the Watson X umbrella. We do

0:21:53.640 --> 0:21:56.280
<v Brian Ryerson>a lot of training with the IBM team, so we

0:21:56.359 --> 0:22:00.119
<v Brian Ryerson>have to constantly train and retrain that model. I think

0:22:00.119 --> 0:22:02.960
<v Brian Ryerson>the other piece that we're doing is again as we're

0:22:02.960 --> 0:22:05.720
<v Brian Ryerson>creating that content and we have the IBM consulting team

0:22:05.760 --> 0:22:08.280
<v Brian Ryerson>on site helping us with that, is as we see

0:22:08.280 --> 0:22:11.280
<v Brian Ryerson>things and we see outputs, it's refeeding that back into

0:22:11.280 --> 0:22:13.080
<v Brian Ryerson>the model to make it better for the next time.

0:22:13.200 --> 0:22:16.680
<v Brian Ryerson>So it's a constantly learning process that we're undergoing.

0:22:17.280 --> 0:22:21.040
<v Jacob Goldstein>So I want to talk about scale. Yes, you have

0:22:21.200 --> 0:22:24.960
<v Jacob Goldstein>like what twenty two different courts with matches going all

0:22:25.000 --> 0:22:28.640
<v Jacob Goldstein>at the same time, you're trying to you know, approximately

0:22:29.000 --> 0:22:32.399
<v Jacob Goldstein>instantly generate summaries of all these matches in something like

0:22:32.480 --> 0:22:37.320
<v Jacob Goldstein>real time. And I'm curious in particular how the IBM

0:22:37.359 --> 0:22:41.560
<v Jacob Goldstein>models you're using, the IBM Granite models are helping you scale.

0:22:42.040 --> 0:22:44.800
<v Brian Ryerson>Yeah. So I think one of the big learnings we

0:22:44.880 --> 0:22:48.560
<v Brian Ryerson>had with IBM granted models too is that we're able

0:22:48.600 --> 0:22:51.399
<v Brian Ryerson>to run it, you know, against last year's tournaments and

0:22:51.440 --> 0:22:54.760
<v Brian Ryerson>see what the expected outputs could be and really help

0:22:54.840 --> 0:22:57.000
<v Brian Ryerson>train that model heading into the tournament. Because as we

0:22:57.080 --> 0:22:59.440
<v Brian Ryerson>talked about in the beginning, is we can plan play

0:22:59.440 --> 0:23:02.040
<v Brian Ryerson>and play. Once two players get on court, the outcome

0:23:02.080 --> 0:23:04.480
<v Brian Ryerson>is unknown. So how do we really run it through

0:23:04.480 --> 0:23:07.479
<v Brian Ryerson>its paces and really make sure that whatever that outcome

0:23:07.520 --> 0:23:09.840
<v Brian Ryerson>could be and whatever that scenario is, whether it's a

0:23:10.040 --> 0:23:13.760
<v Brian Ryerson>fifth set tie break that happens, or maybe there's a

0:23:13.760 --> 0:23:17.560
<v Brian Ryerson>a fault in the match or something that we're not anticipating,

0:23:17.720 --> 0:23:19.359
<v Brian Ryerson>that we have that accounted for and that the a

0:23:19.600 --> 0:23:22.240
<v Brian Ryerson>won't throw off that output. So we really try to

0:23:22.280 --> 0:23:26.840
<v Brian Ryerson>think through every scenario, which is sometimes difficult, right because

0:23:27.000 --> 0:23:29.560
<v Brian Ryerson>again live sports is the unknown is the unknown that's

0:23:29.560 --> 0:23:31.840
<v Brian Ryerson>what makes it fun. We do spend a lot of

0:23:31.840 --> 0:23:34.840
<v Brian Ryerson>time thinking through potential scenarios and ensuring that we have

0:23:34.880 --> 0:23:38.119
<v Brian Ryerson>the right data sets and the model to predict that.

0:23:38.920 --> 0:23:42.240
<v Jacob Goldstein>Tell me about match reports and the generative AI model

0:23:42.240 --> 0:23:42.919
<v Jacob Goldstein>you're using for that.

0:23:43.640 --> 0:23:46.080
<v Brian Ryerson>Yeah, so match reports will be new for us this year,

0:23:46.160 --> 0:23:48.680
<v Brian Ryerson>so we're in testing right now, so we're really excited

0:23:48.720 --> 0:23:51.000
<v Brian Ryerson>around it. But the model that we'll be able to

0:23:51.119 --> 0:23:54.520
<v Brian Ryerson>use using Watson X will use a bunch of different

0:23:54.520 --> 0:23:57.560
<v Brian Ryerson>parts of the suite of tools A, meaning that again

0:23:57.600 --> 0:24:00.560
<v Brian Ryerson>of taking some of that punditry and the unstructured and

0:24:00.600 --> 0:24:04.040
<v Brian Ryerson>the editorial spend, take our structured data as well. And

0:24:04.119 --> 0:24:07.240
<v Brian Ryerson>really what we're working on right now is figuring out

0:24:07.280 --> 0:24:10.880
<v Brian Ryerson>the right prompts for the AI to really ensure that

0:24:11.040 --> 0:24:15.320
<v Brian Ryerson>it tells the right structured story, meaning what just happened. Right,

0:24:15.480 --> 0:24:18.360
<v Brian Ryerson>So our recap is pretty standard. Here's what the data

0:24:18.400 --> 0:24:20.680
<v Brian Ryerson>is telling us, who won, who lost? How many sets?

0:24:20.680 --> 0:24:23.360
<v Jacob Goldstein>Here's the story the structured data part, that's the easy part.

0:24:23.560 --> 0:24:26.280
<v Brian Ryerson>Yeah, and then really where it gets exciting is then

0:24:26.560 --> 0:24:30.200
<v Brian Ryerson>what does this mean? Meaning what's upcoming? So there's all

0:24:30.240 --> 0:24:32.440
<v Brian Ryerson>these different scenarios when you get into you know, two

0:24:32.520 --> 0:24:35.239
<v Brian Ryerson>hundred and fifty four players and a large straw. This

0:24:35.280 --> 0:24:37.880
<v Brian Ryerson>allows us to distill that down and really tell kind

0:24:37.920 --> 0:24:40.840
<v Brian Ryerson>of what could happen upcoming. The AI helps us do

0:24:40.920 --> 0:24:41.800
<v Brian Ryerson>that at scale.

0:24:42.040 --> 0:24:44.359
<v Jacob Goldstein>So I want to sort of generalize for a moment

0:24:44.400 --> 0:24:47.960
<v Jacob Goldstein>to talk about kind of you know, broader challenges with

0:24:48.000 --> 0:24:50.960
<v Jacob Goldstein>AI and how you've solved them. You know a lot

0:24:51.000 --> 0:24:55.880
<v Jacob Goldstein>of generative AI pilots fail because the data quality isn't

0:24:55.920 --> 0:24:59.199
<v Jacob Goldstein>high enough, because the risk controls aren't there, and so

0:24:59.280 --> 0:25:02.560
<v Jacob Goldstein>I'm curious how you dealt with those problems and are

0:25:02.600 --> 0:25:03.200
<v Jacob Goldstein>dealing with them.

0:25:03.560 --> 0:25:06.479
<v Brian Ryerson>Data quality, again, we feel common with the data that

0:25:06.640 --> 0:25:10.320
<v Brian Ryerson>is supplied from the US open and from the USTA, right,

0:25:10.440 --> 0:25:13.239
<v Brian Ryerson>so we have again that's our structure, scoring data and

0:25:13.280 --> 0:25:16.159
<v Brian Ryerson>all that. I think what we're constantly looking at is

0:25:16.160 --> 0:25:18.639
<v Brian Ryerson>when we get outside of our known sources and out

0:25:18.680 --> 0:25:20.840
<v Brian Ryerson>to third parties is that's where a lot of the

0:25:20.880 --> 0:25:24.320
<v Brian Ryerson>testing and model work happens. So we pull in different

0:25:24.359 --> 0:25:27.879
<v Brian Ryerson>data sources and really try to work through how it

0:25:27.960 --> 0:25:30.159
<v Brian Ryerson>changes that output. Again, some of that comes down to

0:25:30.280 --> 0:25:32.600
<v Brian Ryerson>where it's an open model and the transparency that we

0:25:32.680 --> 0:25:35.480
<v Brian Ryerson>have and the learning that comes behind it. That's where

0:25:35.520 --> 0:25:37.720
<v Brian Ryerson>a lot of that confidence can come from, and it

0:25:37.760 --> 0:25:40.800
<v Brian Ryerson>comes from a lot of testing and feeding it more data.

0:25:41.600 --> 0:25:44.000
<v Brian Ryerson>Your second question was a little bit more around the

0:25:44.040 --> 0:25:45.280
<v Brian Ryerson>output I believe.

0:25:45.040 --> 0:25:48.240
<v Jacob Goldstein>Right, yeah, and risks, Right, so risk I think of

0:25:48.320 --> 0:25:51.080
<v Jacob Goldstein>risk more in terms of output, right, But the obvious

0:25:51.119 --> 0:25:54.160
<v Jacob Goldstein>sphere is like what if it says something wrong? Yeah,

0:25:54.200 --> 0:25:57.840
<v Jacob Goldstein>inflammatory or whatever like that seems scary.

0:25:58.119 --> 0:26:01.480
<v Brian Ryerson>Yeah, it definitely is one of our largest concerns when

0:26:01.480 --> 0:26:04.040
<v Brian Ryerson>we first took this fora I would say a lot

0:26:04.040 --> 0:26:06.440
<v Brian Ryerson>of that comes through our work with IBM and IBM

0:26:06.600 --> 0:26:10.000
<v Brian Ryerson>consulting team and really ensuring that again they're an extension

0:26:10.000 --> 0:26:13.520
<v Brian Ryerson>and the partnership there of our team. So whenever we

0:26:13.560 --> 0:26:15.680
<v Brian Ryerson>are creating let's say it's the Match Report, and we're

0:26:15.720 --> 0:26:18.520
<v Brian Ryerson>going to be creating these extinct articles for every single

0:26:19.040 --> 0:26:21.560
<v Brian Ryerson>men's and women's single match that happens, is all of

0:26:21.560 --> 0:26:24.679
<v Brian Ryerson>those will have manual review and people looking through them

0:26:24.720 --> 0:26:27.840
<v Brian Ryerson>for accuracy to ensure that the model then hallucinat or

0:26:27.840 --> 0:26:30.080
<v Brian Ryerson>make up a fact or fill in the gaps from

0:26:30.119 --> 0:26:32.720
<v Brian Ryerson>things like that. That's the first step. And then also

0:26:32.800 --> 0:26:35.600
<v Brian Ryerson>when our editorial team goes to publish those of the website,

0:26:35.760 --> 0:26:37.520
<v Brian Ryerson>they're going to be checking it as well, So there

0:26:37.560 --> 0:26:40.879
<v Brian Ryerson>are manual interventions throughout that to really check that model,

0:26:41.320 --> 0:26:43.920
<v Brian Ryerson>but we feel that the ability to do it at

0:26:43.960 --> 0:26:46.520
<v Brian Ryerson>scale and with us more to check that is the

0:26:46.520 --> 0:26:48.680
<v Brian Ryerson>efficiency problem that we've been looking to solve.

0:26:49.480 --> 0:26:52.679
<v Jacob Goldstein>So the USTA and IBM have been working together on

0:26:52.920 --> 0:26:55.679
<v Jacob Goldstein>digital innovation for like thirty years from you know, the

0:26:55.720 --> 0:27:00.600
<v Jacob Goldstein>first website, yes for the USTA until now that's the

0:27:00.640 --> 0:27:04.120
<v Jacob Goldstein>past thirty years. If you look ahead, what's the next.

0:27:03.920 --> 0:27:06.960
<v Brian Ryerson>Thirty thirty years is a really long time.

0:27:06.840 --> 0:27:07.640
<v Jacob Goldstein>At A three.

0:27:07.800 --> 0:27:11.399
<v Brian Ryerson>Yeah, I think you know where I get excited, and

0:27:11.560 --> 0:27:13.680
<v Brian Ryerson>I think I alluded to it in the beginning about

0:27:13.720 --> 0:27:15.959
<v Brian Ryerson>how I feel like we're just scratching at the surface,

0:27:16.080 --> 0:27:18.240
<v Brian Ryerson>especially with journat of Ai, and where I see it

0:27:18.320 --> 0:27:21.440
<v Brian Ryerson>going is there's a lot of different fans out there,

0:27:21.720 --> 0:27:23.600
<v Brian Ryerson>and we're also very kindness in the US open that

0:27:23.600 --> 0:27:26.000
<v Brian Ryerson>we're a worldwide event and that there's a lot of

0:27:26.000 --> 0:27:30.240
<v Brian Ryerson>different fans that were not necessary creating content for bespoke

0:27:30.359 --> 0:27:33.679
<v Brian Ryerson>meaning in their native language or maybe it's in that

0:27:33.760 --> 0:27:35.560
<v Brian Ryerson>native players language and things like that.

0:27:35.720 --> 0:27:38.960
<v Brian Ryerson>Is where I get excited is we've seen immense growth

0:27:39.000 --> 0:27:41.280
<v Brian Ryerson>with A Highlights and the ability to now do highlights

0:27:41.320 --> 0:27:44.280
<v Brian Ryerson>at scale is the ability for us to start creating

0:27:44.359 --> 0:27:48.560
<v Brian Ryerson>content in different languages, maybe covering different parts of the match,

0:27:48.560 --> 0:27:51.040
<v Brian Ryerson>So maybe you do have that stats junkie really wants.

0:27:51.119 --> 0:27:54.120
<v Brian Ryerson>Just it's the fastest serve and here's the deep insights

0:27:54.200 --> 0:27:56.760
<v Brian Ryerson>versus the casual fan who's looking for more of the

0:27:56.840 --> 0:28:00.480
<v Brian Ryerson>storytelling around how a player trains and what up to

0:28:00.560 --> 0:28:03.400
<v Brian Ryerson>it was like and what it means for them afterwards

0:28:03.440 --> 0:28:05.639
<v Brian Ryerson>and things like that. A lot of that takes a

0:28:05.640 --> 0:28:08.320
<v Brian Ryerson>lot of time. Now we're able to solve that efficiency

0:28:08.400 --> 0:28:11.080
<v Brian Ryerson>problem and do it in multiple languages, we can really

0:28:11.119 --> 0:28:14.680
<v Brian Ryerson>create I want to say, personalized content to a lot

0:28:14.760 --> 0:28:17.960
<v Brian Ryerson>more fans all around the world, which again helps us

0:28:18.040 --> 0:28:20.280
<v Brian Ryerson>grow the sport of tennis great.

0:28:20.840 --> 0:28:24.240
<v Jacob Goldstein>Uh So I want to finish with a speed round. Okay,

0:28:24.440 --> 0:28:25.080
<v Jacob Goldstein>are you ready?

0:28:25.200 --> 0:28:26.000
<v Brian Ryerson>I am ready?

0:28:26.119 --> 0:28:29.640
<v Jacob Goldstein>Okay, first thing that comes to mind complete this sentence.

0:28:30.240 --> 0:28:31.960
<v Jacob Goldstein>In five years, AI.

0:28:31.840 --> 0:28:35.360
<v Brian Ryerson>Will transform many parts of the business.

0:28:35.600 --> 0:28:39.840
<v Jacob Goldstein>What is the number one thing that people misunderstand about AI?

0:28:40.400 --> 0:28:44.560
<v Brian Ryerson>That it's supplemental, not replacing, meaning that it helps it

0:28:44.600 --> 0:28:48.640
<v Brian Ryerson>with efficiencies, but it doesn't necessarily replace the creativity.

0:28:49.360 --> 0:28:53.040
<v Jacob Goldstein>Right now, what advice would you give yourself ten years

0:28:53.080 --> 0:28:55.920
<v Jacob Goldstein>ago to better prepare you for today?

0:28:57.000 --> 0:29:00.520
<v Brian Ryerson>I think it would have been especially now that we're

0:29:00.560 --> 0:29:03.000
<v Brian Ryerson>able to take so much of that unstructured data and

0:29:03.520 --> 0:29:06.640
<v Brian Ryerson>pass content that we were created to help tell stories

0:29:07.240 --> 0:29:10.080
<v Brian Ryerson>was to I want to say, archive more of that

0:29:10.440 --> 0:29:12.040
<v Brian Ryerson>in a way that we could be using that to

0:29:12.120 --> 0:29:15.560
<v Brian Ryerson>help pull from that now. So you know, we've seen

0:29:15.680 --> 0:29:17.720
<v Brian Ryerson>kind of a change in the guard from some of

0:29:17.760 --> 0:29:20.680
<v Brian Ryerson>our start players to now new and up and comers,

0:29:20.680 --> 0:29:22.680
<v Brian Ryerson>and it would be really fascinating to me if there

0:29:22.760 --> 0:29:25.440
<v Brian Ryerson>was a way to to cross sections some of that

0:29:25.600 --> 0:29:28.680
<v Brian Ryerson>and saying like what trajectories are certain up and coming

0:29:28.720 --> 0:29:32.360
<v Brian Ryerson>players maybe filing from others. So it's more I wish

0:29:32.360 --> 0:29:34.480
<v Brian Ryerson>we kept more of the content we created.

0:29:34.200 --> 0:29:39.120
<v Jacob Goldstein>Back fave the data exactly. Well are you saving it

0:29:39.160 --> 0:29:39.600
<v Jacob Goldstein>all now?

0:29:39.880 --> 0:29:42.880
<v Brian Ryerson>Oh yeah, one hundred percent learned our lesson? Yes, yes.

0:29:43.480 --> 0:29:45.880
<v Jacob Goldstein>So on the business side of AI, what do you

0:29:45.880 --> 0:29:47.120
<v Jacob Goldstein>think is the next big thing?

0:29:47.760 --> 0:29:51.040
<v Brian Ryerson>I alluded to it earlier. I think it's personalization and

0:29:51.080 --> 0:29:54.680
<v Brian Ryerson>getting content that's catered to you at scale, whether you

0:29:54.720 --> 0:29:57.720
<v Brian Ryerson>know that's across the sports sphere or or any type

0:29:57.720 --> 0:30:01.800
<v Brian Ryerson>of written content or or new I feel like the

0:30:01.880 --> 0:30:05.280
<v Brian Ryerson>ability to really get contentated to the type of fan

0:30:05.360 --> 0:30:07.959
<v Brian Ryerson>you are and the insights you have is where we're

0:30:08.000 --> 0:30:08.440
<v Brian Ryerson>all headed.

0:30:09.480 --> 0:30:13.320
<v Jacob Goldstein>And in terms of your non work life, how do

0:30:13.360 --> 0:30:15.000
<v Jacob Goldstein>you use AI? Day to day.

0:30:15.200 --> 0:30:17.560
<v Brian Ryerson>It's funny. I was just having this conversation with a

0:30:17.600 --> 0:30:20.840
<v Brian Ryerson>friend the other day and we were talking about sometimes

0:30:20.840 --> 0:30:23.920
<v Brian Ryerson>when you're starting something new, the hardest thing to do

0:30:24.080 --> 0:30:26.360
<v Brian Ryerson>is you have a blank piece of paper or a thought,

0:30:26.400 --> 0:30:30.520
<v Brian Ryerson>and how do you get started. Sometimes with these generative models,

0:30:30.840 --> 0:30:32.440
<v Brian Ryerson>the easiest thing and the best thing you can do

0:30:32.520 --> 0:30:35.200
<v Brian Ryerson>is it helps you get started. Meaning it may not

0:30:35.200 --> 0:30:37.160
<v Brian Ryerson>be one hundred percent with that first prompt, but it's

0:30:37.200 --> 0:30:40.520
<v Brian Ryerson>that efficiency of whether it's an outline for a new idea,

0:30:40.720 --> 0:30:42.920
<v Brian Ryerson>or it's a marketing brief you have to write, or

0:30:43.120 --> 0:30:45.280
<v Brian Ryerson>sometimes even if it's an email, you have to write

0:30:45.400 --> 0:30:47.600
<v Brian Ryerson>for a personal something and you're not sure how to

0:30:47.640 --> 0:30:49.760
<v Brian Ryerson>word it the right way. It allows you to have

0:30:49.960 --> 0:30:51.880
<v Brian Ryerson>a start and then you can edit from there. So

0:30:51.920 --> 0:30:54.960
<v Brian Ryerson>again going back to my efficiency point, it helps you

0:30:55.000 --> 0:30:55.920
<v Brian Ryerson>become more efficient.

0:30:56.200 --> 0:30:57.840
<v Jacob Goldstein>Solve's the blank page problem.

0:30:58.040 --> 0:30:58.400
<v Brian Ryerson>It does.

0:31:00.200 --> 0:31:01.600
<v Jacob Goldstein>It was great to talk with you. Thank you so

0:31:01.680 --> 0:31:02.360
<v Jacob Goldstein>much for your time.

0:31:02.440 --> 0:31:03.920
<v Brian Ryerson>Yeah, this was fun. Thanks for having me.

0:31:06.080 --> 0:31:08.320
<v Malcolm Gladwell>A huge thanks to Jacob and Brian for the deep

0:31:08.400 --> 0:31:12.520
<v Malcolm Gladwell>dive into the cutting edge innovations transforming the game of tennis.

0:31:13.080 --> 0:31:16.040
<v Malcolm Gladwell>Brian shed light on how the US opens partnership with

0:31:16.080 --> 0:31:21.360
<v Malcolm Gladwell>IBM is harnessing data driven insights to reshape storytelling in sports,

0:31:21.760 --> 0:31:27.280
<v Malcolm Gladwell>from AI generated commentary to match reports. As we look ahead,

0:31:27.360 --> 0:31:32.120
<v Malcolm Gladwell>I'm excited about the possibilities for personalizing content and reaching

0:31:32.200 --> 0:31:36.240
<v Malcolm Gladwell>fans in new ways. The future of AI promises more

0:31:36.280 --> 0:31:44.600
<v Malcolm Gladwell>than just efficiency. It's about enhancing fan experiences worldwide. Smart

0:31:44.600 --> 0:31:48.320
<v Malcolm Gladwell>Talks with IBM is produced by Matt Romano, Joey Fishground,

0:31:48.520 --> 0:31:52.560
<v Malcolm Gladwell>and Jacob Goldstein. We're edited by Lydia Jean Kott. Our

0:31:52.600 --> 0:31:57.240
<v Malcolm Gladwell>engineers are Sarah Bruguiere and Ben Tolliday. Theme song by Gramoscope.

0:31:58.240 --> 0:32:01.280
<v Malcolm Gladwell>Special thanks to the EightBar and IBM teams, as well

0:32:01.320 --> 0:32:04.920
<v Malcolm Gladwell>as the Pushkin marketing team. Smart Talks with IBM is

0:32:04.960 --> 0:32:09.200
<v Malcolm Gladwell>a production of Pushkin Industries and Ruby Studio at iHeartMedia.

0:32:09.720 --> 0:32:13.880
<v Malcolm Gladwell>To find more Pushkin podcasts, listen on the iHeartRadio app,

0:32:14.200 --> 0:32:19.960
<v Malcolm Gladwell>Apple Podcasts, or wherever you listen to podcasts. I'm Malcolm Gladwell.

0:32:20.240 --> 0:32:24.000
<v Malcolm Gladwell>This is a paid advertisement from IBM. The conversations on

0:32:24.040 --> 0:32:41.800
<v Malcolm Gladwell>this podcast don't necessarily represent IBM's positions, strategies or opinions.