- 16 hours ago
In the latest installment of the TDN Business Hour, the conversation focused on one of the most rapidly evolving and disruptive forces across global industries: the integration of Artificial Intelligence (AI). Host Najja Thompson talks with three insiders actively driving the integration of AI into the horse industry: Josh Isner, Joe Appelbaum and Billy Paynter.
Category
🥇
SportsTranscript
00:00Artificial intelligence is rapidly changing the way businesses operate across every industry.
00:05From healthcare and finance to professional sports, organizations are using AI to improve decision-making,
00:11streamline operations, and create new opportunities for growth.
00:15The question for thoroughbred racing isn't whether AI is coming, it's if we're ready to embrace it.
00:20Most importantly, can AI help position the thoroughbred industry to become more competitive,
00:25sustainable, and attractive to the next generation of owners, breeders, horse players, and fans.
00:32My name is Nadja Thompson, Executive Director of New York Thoroughbred Breeders.
00:37And today on the TDN Business Hour, we'll explore those questions and more.
00:41With Josh Isner of Delta Squad Racing, who has a professional background in AI and uses AI technology in his
00:48horse ownership.
00:49Billy Painter of Ocean Reef Racing, who has developed Ecotrack AI, a tool for every participant in our game.
00:56And Joe Applebaum, who has created Wager, a betting tool using AI for professional horse players.
01:03Thank you for joining us.
01:10Joining us now is Josh Isner of Delta Squad Racing.
01:13Josh, thank you for joining us today.
01:15Thank you very much for having me.
01:18You know, you have an interesting story in how you got involved in thoroughbred racing.
01:21Can you tell our viewers just how you got to being an owner?
01:24Sure thing.
01:26I came out of left field a little bit.
01:28I've been a fan of the sport kind of recreationally for a long time, but my friend Dave was nice
01:34enough to identify me as a fellow degenerate
01:37and somebody that would immediately get addicted to horse racing.
01:41And so he invited me up to Whitney Weekend in Saratoga.
01:45We went to the Whitney, some other races that week.
01:48And then obviously the sale follows the next two days.
01:51And we bought our first horse together.
01:53I also bought my first horse on my own.
01:56And that was August of last year.
01:59And now our stable has almost 25 horses here.
02:01And we're off and running.
02:03We've won a stake, won an allowance last week.
02:07And we're rolling, looking forward to a great summer.
02:11So having the time of my life.
02:13Can you tell us more about your background, getting into thoroughbred racing and your day job as well?
02:18Sure thing.
02:18So ultimately, I run a company called Axon.
02:22It's the only company I've ever worked for.
02:24I've been at Axon since 2009.
02:27I graduated college that year and drove straight across the country from Boston to Arizona and got started there.
02:33At that time, we were manufacturing taser devices.
02:36The yellow guns you see that police officers wear.
02:39Our goal is to obsolete the bullet in policing, meaning we can make a taser so good that a police
02:45officer never has to take somebody's life with a gun.
02:49That business has evolved now into we started putting cameras on the tasers.
02:54And that started to show how they were being used in the field.
02:57And then we said, hey, this is kind of working.
02:58Let's put the camera on the officer.
03:00And we've become the market leader in body-worn cameras.
03:03So when you see all the body-worn camera videos on the news, you'll notice the little triangle in the
03:08top right corner, the yellow one.
03:10That's our company, Axon.
03:11And now that we've built up this big data set of all this data coming off the body cameras, we're
03:16the market leader in artificial intelligence and public safety.
03:20We build a lot of AI tools that police can use to, for example, write police reports faster or translate
03:28language in the field using the body camera or start to identify license plates using the body camera and so
03:35forth.
03:36So a lot of applications for AI and public safety.
03:39But ultimately, day-to-day, that's what we do.
03:41We serve all U.S. public safety and military as well as over 100 countries around the world.
03:46So a lot of fun.
03:48I thought it would be the most fun thing I've ever done.
03:51It's getting its run for its money now that I'm in the horse racing game.
03:54We're having a lot of fun doing this as well.
03:56And initially being from Boston and that area, I think your first racetrack experience was actually Saratoga for the Whitney.
04:03Can you talk about that journey?
04:05Sure thing.
04:06So I missed Suffolk Downs.
04:08So I was a little – it was closing as I was coming of age in Boston here.
04:13But I was able to attend the Whitney last summer.
04:18That was really my first foray into horse racing and ownership.
04:22My friend Dave correctly identified me as a degenerate that will immediately get hooked on horse racing.
04:29And we went up there.
04:31We saw Sierra Leone dominate that day at the Whitney.
04:35We went to the sales the next couple days, Monday and Tuesday that week.
04:39And we had a fantastic time.
04:42We bought a horse together.
04:43I bought my own first horse there as well.
04:46And then Keeneland was two weeks later, and we were off and running.
04:50So now the stable has over 20 horses about 10 months in, and that number will continue to grow and
04:56starting to have a lot of fun.
04:58Definitely.
04:59And congratulations.
05:00In your day job, you certainly use artificial intelligence and technology to help to keep things up to speed.
05:06How has that transitioned for you to Thoroughbred Racing World and as an owner?
05:10Sure thing.
05:11Yeah, there is a lot of applications for AI in horse racing.
05:15I've actually spent some time rebuilding the catalog app for the sales and building in a lot more functionality to
05:22it.
05:23Just me, one other person, and a lot of help from Claude Code.
05:27I use it for handicapping.
05:30I'm trying to start to experiment with machine vision to see if we can learn anything in terms of yearling
05:37confirmation.
05:38As an owner, certainly, you know, things like understanding the condition book and just understanding the background of the sport.
05:46I was interested in learning more about the Jockey Club recently.
05:49So I had ChatGPT make me a podcast script.
05:53I fed it into something called Notebook LM.
05:56It produced an hour podcast.
05:57And as I was driving, I just listened and learned a ton about the Jockey Club.
06:01So there are all kinds of applications for AI in horse racing, and I think they'll be welcome.
06:08This hasn't necessarily been an industry that's been well served by technology historically.
06:12So hopefully this is one where, you know, it can be a step function forward in terms of how people
06:18are able to deploy this.
06:20Yes, and getting into that, especially on the sales side, you know, there's so much of a human element between
06:25bloodstock agents and visually seeing confirmations in your use or where you think AI is going to go in the
06:32future, especially with that vision technology you talked about.
06:35Can you see that expanding in the next five to ten years and adopted on a wide scale platform?
06:41I hope so, and I hope it doesn't take five to ten years.
06:44I think if it does, there'll be people using it a lot sooner than that that will have a tremendous
06:49advantage, in my opinion.
06:51So I think this is one, just like in all things, when something new comes out and people figure out
06:56exactly how they can create an edge for themselves or some kind of application for it that others aren't going
07:03to be using, those are going to be the early winners.
07:05Now, with this, you know, to be fair, you know, it's AI is generative.
07:10It gets better over time.
07:12And so this is the worst it'll ever be today.
07:15And so, you know, there'll be some there'll be some misses along the way, undoubtedly.
07:19But I think this is something that will hit the mainstream in horse racing in the next 18 months to
07:2524 months.
07:26Five years would be a long time to wait.
07:28And were you trying to solve a specific problem when you were looking to implement AI into how you operate
07:35your horse racing ownership experience?
07:37Or is that something that just naturally came along as a technology has developed rapidly through the last year or
07:44so?
07:44Yeah, I think it's been a phenomenal tool to get up to speed in horse racing.
07:49You know, it's it's one where being an owner and learning about the sport and all the ins and outs,
07:56it is it's overwhelming and it's it's a pretty high barrier to entry.
08:00And so when you have a tool that can literally teach you about anything regarding the sport, it's the biggest
08:06source of intelligence on the planet.
08:09Ultimately, it's something that you can catch up quickly.
08:12You know, I put little videos together that just teach me about certain elements of the sport.
08:16There's there's all kinds of things you can do.
08:19But personally, one of the things I hope AI really contributes to is the idea that essentially you can you
08:30can learn a lot more than what's on the page.
08:33Just like in a, you know, they take a horse catalog page.
08:36It doesn't tell you whether the dam was a turf horse or a dirt horse unless, you know, the races
08:41that she ran in.
08:43It doesn't tell you if she was sprint or went long.
08:46Ultimately, that's one very easy thing to do where you can color code the races on the catalog page.
08:51And all of a sudden, as a new owner, you're not researching every race and how long it was.
08:56And if the surface changed or whatever the case may be, you just have it right there in front of
09:00you.
09:00And you can get all kinds of additional stats and information that, you know, really, for example, you know, bloodstock
09:12agents really have off the top of their head, but really nobody else.
09:15So as an owner that wants to feel like they're participating, not just paying the bills and seeing what happens.
09:21This is something that that's really helped me catch up and learn quickly.
09:25There certainly has been a wider conversation within the industry in terms of acquisition of data and who owns data.
09:33Have you run into any obstacles in trying to develop your algorithms or AI functions and the data that's presently
09:40available and what you'd like to see more available from the industry?
09:43I think so.
09:44And I do think the there is some there, you know, the APIs are available.
09:49So those are the essentially connectors of information using software.
09:53So you can derive some of the data in the catalogs, especially using the APIs.
09:59But I'd say, as I assess, you know, I think all of us want horse racing to grow as a
10:04sport and get in front of more eyeballs.
10:06Personally, I think the data is very central to that.
10:09And the way horsemen use the data today is very specific for things that knowledgeable horsemen would need it for.
10:17But it's not it doesn't translate to end users who would be prospects to start to be fans of the
10:24sport.
10:25And so hopefully between, you know, using the data and applying AI to it, you can really start to teach
10:32people much more simply what they're looking at.
10:35Even things like calling the horse, a mom or a dad, as opposed to a dam or a sire, things
10:40that just make the sport less intimidating for people to come into.
10:44I think AI can be a major translator in that way.
10:47But of course, it does.
10:48You need the data to be able to do that.
10:51And you've also had tremendous experience as a handicapper as well as a fourth place finish in the NHC, I
10:56believe.
10:57I'm a sport-ranked NHC rookie.
10:59I came in about 105th out of 850-ish last year.
11:04Missed the cut in a heartbreaking race in the last race of the second day.
11:07But I'm having a lot of fun doing it.
11:10Sorry about that.
11:11But have you used AI in your handicapping levels?
11:14And is that a personalized system separate from your ownership side?
11:17Or have you merged both systems to work together?
11:20Sure thing.
11:21I have a few different methods, some more conventional using the form.
11:24But AI, I do believe, should be applicable to handicapping.
11:30I think like a lot of things in AI, it really depends on prompting.
11:36So prompting is essentially what you're asking the AI to do, what context you're giving it, being very specific about
11:43the outcomes you want to see.
11:45And so you could have 100 people using AI for horse betting, and each of those AIs is giving a
11:53different pick for a race.
11:54Because they're looking at different information or they've been prompted differently.
11:58So I've spent a lot of time kind of building up a program with very specific prompting that I'd say
12:08is still, it's a work in progress, but it's helped.
12:13I've won my two seats for the NHC already, you know, for next March already.
12:17So it's certainly helping in that regard.
12:19And I want it to be very dialed in, and we've got some work to do to do that.
12:24But we're on the right track, I think.
12:25Definitely.
12:27And what I've read is that you've used AI to, I guess, lessen your burden in other areas or like
12:34the manual labor that it would take to exhaustively study pedigrees and background.
12:39So would you say that's one benefit of AI is to streamline workload?
12:42Of course, of course.
12:45And just give you better ideas of what you're looking at, 100%.
12:49You know, I think a lot of people will use the formulator to kind of build out what the race
12:54should look like.
12:54Pace-wise, AI can do that in two seconds.
12:57And so even if you're not interested in the pick, if you're just interested in seeing how the pace will,
13:01you know, likely play out, there are very, very quick solves for that now.
13:08And then the more you do with it, the more just opportunities it opens up.
13:11Like when we built the new catalog app, it's got a vet workflow in it.
13:15So you see a horse you like at a sale, you hit one button, the vet gets a text saying,
13:19hey, you need to vet this horse.
13:21And they just text that number back and it populates in your app so you have all your vet notes
13:25associated with each catalog page.
13:27So there's just things you can do once you get going that are natural branches into new workflows.
13:33And, you know, it's certainly for someone like me who's trying to learn fast, it's been a tremendous tool.
13:40Have you seen widespread adoption from fellow owners and friends?
13:43I know you've recently come into the game and, you know, buoyed yourself with now a 20-horse stable and
13:48more coming.
13:49But from owners you've presently met in the game that have been longstanding, have they been adaptive to embracing AI?
13:55You know, my day job, I work with police officers and we always say the only thing they hate more
14:03than change is staying the same.
14:05And I'd say there's an element of that with horse owners, too, and just the industry as a whole where
14:10it's a very kind of set in its ways mindset right now.
14:14But I think ultimately the thing that has to really get people excited about it is they need to see
14:19people having success using these new tools.
14:22It's great if you can build a new catalog app, but if you're an expert at reading the catalog the
14:26way you've done it for 30 years, you know, it's probably not as exciting.
14:29But when people start having more success using these tools, I think that'll be the paradigm shift where people say,
14:36hey, I'm falling behind by not using these.
14:39But I wouldn't say the adoption is on the leading edge at this point.
14:45Is that something you're working to convince others or are you taking advantage of that edge that you see currently?
14:51Yeah, I think I'll certainly anyone who asks, you know, I want the sport to grow.
14:56I want people to have success in it. So I'll certainly be you know, I won't be hoarding the information.
15:00But I'm not sure that we'll be trying to gain mass adoption either.
15:05It is something that I view as an edge. And and so it's it's certainly helpful in some ways.
15:11And I think, you know, like I said, as soon as folks start to catch on to it, I think
15:15it'll be pretty mainstream because it's pretty clear that there's something here.
15:19But in the meantime, hopefully hopefully we we get a few prospects that might be flying under the radar here.
15:26Definitely. And I guess, Leslie, I'll ask you within 10 to 15 years or, as you mentioned, maybe six to
15:3412 months.
15:34What do you think are going to be the biggest improvements from A.I. adoption within thoroughbred racing?
15:40Sure thing. I think you will. I think you will see a revolution of all the tools that folks use.
15:47And that's something that I'm very passionate about.
15:51Like that should be the easiest thing to do is like you're using software that's 20 years old and now
15:56you can modernize it.
15:57Like that should be by the sales next summer. I'd imagine that that'll be in practice.
16:03Certainly, I think on the handicapping side, as folks start to start to understand what's possible there, I think you'll
16:11see more applications around that.
16:14Now, I think a pessimist would say the folks that are going to benefit first from that are the CAWs
16:19because they're going to be investing a lot more into it than, you know, than recreational betters.
16:24But ultimately, I think that will be a tool for everybody.
16:27I think the thing that will probably take the longest is the machine vision stuff.
16:31So things like not only examining horses and, you know, making predictions based on the physical examination, but also the
16:39dynamics of how horses move and being able to plot that out and use it to identify soundness or, you
16:47know, really whatever you're looking for.
16:48Those are the things that are getting going, but probably two or three years away from from really meaningful changes.
16:56But like anything, it's going to be a process, but it isn't going to be a linear process.
17:00That's for sure. It's going to be exponential where, you know, I would say that the utility is going to
17:05double every year.
17:07And so folks, you know, that are, you know, folks will be taking notice, I think, relatively quickly.
17:12That would be pretty cool and interesting, like those Ray-Ban meta glasses, if you could just put those on
17:17and look at soundness or just holding your eye.
17:21Like when you're at the sale with your iPad, just holding the camera up to the horse.
17:25And as it's walking or standing still, we should be able to build in a lot of the vision components
17:30into that.
17:31Now, it's going to be some trial and error, right?
17:33There are people who are more like artists that are very, very good at this.
17:36I work with one of them named Liz Crow, and she's fantastic.
17:39And so, you know, they're pretty good at their job.
17:43So it's going to take some time to be able to supplement that work.
17:46But, you know, I think folks like Liz are already trying to investigate how AI can help them.
17:53And I think she's smart to do that.
17:55Definitely. And lastly, I'll ask you, how did you come up with the name of Delta Squad Racing?
17:59Where did that come from?
18:00Yeah, of course. So our company is named Axon, and our logo is a Delta.
18:05So if you took the horse head away from my shirt and just had the triangle behind it, that would
18:11be our corporate logo.
18:12And so Delta Squad Racing is an ode to our company Axon.
18:17And we're the Delta Squad at work.
18:19It's got some ties into the military and so forth.
18:21We serve public safety and military.
18:23And so it was a pretty easy one to decide on in terms of starting my stable.
18:28That's great. Well, thank you for joining us, Josh.
18:30Really appreciate the time.
18:31I really appreciate it as well.
18:33Thanks for having me.
18:35Hi, Joe. Thanks for joining us today.
18:37No, it's great to see you.
18:39Is that a real background or a fake one?
18:41It is indeed real and not artificial, which getting into today's topic, you have more than 20 years experience in
18:48the industry as an owner and executive.
18:51And of course, now an entrepreneur.
18:54Can you tell us what excites you about artificial intelligence and its impact on third-party racing?
19:00Well, I think it's the ability to deliver to consumers the sort of tools, the sort of data and insights
19:08that professionals have been using probably for that whole 20 years.
19:12But now we've had a democratization of these analytical tools that basically make them available to everybody, or that's what
19:22we're trying to do at Wager is make them available to people.
19:26So I think it's an exciting time.
19:29It's a little scary for people because you're going to be asked to do things that you've never done before.
19:36You can't just crack open the paper with a felt tip pen and think you're going to gain those insights.
19:44Yeah, so talk about that development of Wager.
19:46What led you to the concept and how you're using it as a company and as a tool for bettors
19:52to democratize our industry?
19:54So I started thinking about this kind of in just in my head way about 10, 12 years ago, mostly
20:02because I'm extremely competitive, like many horse owners out there.
20:06That's why you get into it, right?
20:09And I realized what I was doing betting-wise, which had been fairly successful and got me into the game,
20:17I started buying horses because I won money gambling, wasn't going to cut it anymore.
20:23That I was up against a group of competitors who had better tools than me, better data than me, better
20:29analysis than me.
20:30And I didn't like that so much, right?
20:32So I started figuring out what it takes to be really analytical in your approach and to be able to
20:42deliver that in a way that people might enjoy.
20:47And so I read and watched videos of a guy named Bill Benter, who really started this off 30 years
20:53ago, and did a lot of reading of academic papers.
20:59I did all the reading of the academic papers, people, so you don't have to.
21:04You just got to listen to what we say.
21:06And this is going on like under the hood for years and years and years, just trying to figure out
21:14what these tools should look like, what real analysis is.
21:18And then about two years ago, we started developing Wager as a consumer product, something that can produce for people
21:27a couple of things.
21:29One, to give you an accurate read on where the market's headed and what the fair value of every horse
21:35should be.
21:36Two, to give you access to strategies that will put you in the best positions more frequently.
21:46Three, to try to change your conception of who you're betting on to as a value hunt or searching for
21:56value as opposed to loving a horse.
22:00So if you're in the love the horse camp, you're probably not going to like what we're doing.
22:04However, if you're in the I just want value from my money team, you're going to really like what we're
22:11doing.
22:12And so we started this years ago, and now we've gotten to a spot.
22:18We've debuted on Naira Bets on their wager pad.
22:21At the bottom of every race, you can see some wager suggested bets, which blend in our odds, the live
22:30odds, and a whole bunch of strategies that we know to be successful over time.
22:35Right. It may not be successful every single race, but over time, we know that these strategies are real successful.
22:43Obviously, in the betting space, you know, average bettors or everyday bettors having to compete against computer assisted wagering teams
22:50is a hot button issue for the industry.
22:53Tools such as wager, which you have developed.
22:55How does that help democratize and, I guess, level out the playing field for an everyday bettor versus these supercomputer
23:03teams?
23:05Well, it's a matter of access, right?
23:07So the stuff that you're not going to really want to do on a daily basis, we do for you,
23:15right?
23:15And we make available.
23:16So to give you an example, we produce a set of odds for every single horse in every single race
23:22in the United States of America, soon to be international.
23:26Although I guess we don't do the tracks in Texas, like Lone Star or Sam Houston, because they don't export
23:35their signal.
23:37So we do that.
23:38So you have an immediate touchstone.
23:41I think today we're covering 12 tracks, right?
23:44So, you know, you couldn't do that physically, you know, even if you wanted to.
23:51Also, like, do you really have the discipline to do that?
23:55Like, that's what makes the computer great, right?
23:58The computer does it quickly.
23:59The computer doesn't get angry.
24:02The computer isn't tired.
24:04The computer doesn't have another job.
24:06So we can produce all this for you.
24:08So in democratizing things, the first thing we want to do is give you a touchstone with some odds that
24:14you can look at, right?
24:15So often you'll see odds on the board and you go, wow, that horse is 10 to 1.
24:19And I can look at my odds and say, you know, he's probably not going to stay at 10 to
24:231.
24:23He's probably going to be more like 4 to 1.
24:25I say, well, how do you know that, Joe?
24:27Well, because we have a machine learning model that's looked at, you know, decades and decades of results and pricing.
24:37And, you know, there's not a lot of hidden secrets in the world, Naj.
24:42It's kind of – you have a good idea where that stuff's going to be.
24:45So that's the first thing.
24:47The second thing, and this goes underappreciated by everyone, is how much constructing your wagers is probably more important than
24:58picking the right horse.
25:00And we have looked at our own wagers.
25:04We've looked at a lot of other people's wagers, and we've looked at the markets.
25:07So we have a really good idea of what sort of wagers are typically successful and what sort of wagers
25:16are value producing, meaning they pay more than they should, right?
25:21So when you hit that trifecta and you're expecting $100 and it paid $140, you're happy, right?
25:30You're happy that you won, period, but you're extra happy when you get a little extra money.
25:33So we're trying to steer people in the direction of those wagers.
25:38So those are the sorts of things.
25:40Sorry.
25:41Will it also be tiered in that beginner bettors can be able to use wager and not go so in
25:47-depth versus if you're a seasoned and well-calculated tournament player or, you know, hardcore bettor where you can use
25:54wager as a tool with the odds with, you know, horses and races all over the country?
25:59Right.
26:00So our first product rollout is very simple, and it's for really a retail bettor.
26:05It gives you some suggestions.
26:06Not that a more sophisticated bettor wouldn't benefit from it, because I think they would.
26:12But for the more retail-oriented players, we're giving you some suggestions right away that you can see.
26:21I'm hopeful later this summer we'll be rolling out on our own website some more sophisticated tools for players so
26:29that they can see – they can take into account their preferences, their risk tolerances, their budgets,
26:36and we can really help create more sophisticated tickets in the Superfecta and the PIC-4 and PIC-5, et
26:43cetera.
26:44We can really help guide people along those lines to, you know, to really steer them to maximize their preferences.
26:52Right.
26:52It's like a decision-making support tool is how we frame it.
26:56Right.
26:57You have some ideas, but it's hard to execute on those ideas, mostly because you've got to punch in a
27:04lot of tickets.
27:04And, you know, these days most guys are doing that on their phone.
27:07It's hard to punch in the preferable number of tickets.
27:10So, to give you an example, we'll take your preferences – like in a PIC-4, a typical player will
27:18play maybe three tickets at most, probably one ticket.
27:21But, like, at most you're playing – if you think you're pretty cool, you play, like, two, three tickets.
27:26Hey, if you give me a $100 budget in the PIC-4, we're likely to play you 20 to 25
27:32tickets.
27:33But we're going to pinpoint exactly which horses you're interested in with which risk tolerances, meaning odds levels, you want
27:43to play with, right, and match that to your budget and try to create paths because we can do that,
27:48right?
27:49Our execution tools are superior to you and me punching things in with our fingers on a phone, right?
27:56And we will operate within the guidelines that Naira and the other ADWs said, but we want to really make
28:04ticket construction and execution as efficient as possible.
28:10And I guess it's also the goal to make that retail better or feel that they're still fine-tuning and
28:16making the final points and saying, you know, one of the great things about betting for me was I had
28:23to cap this race.
28:24I made the decision. I made the correct pick.
28:27But with using products such as Wager, are you saying it's an assistance tool and helping you set the parameters
28:34of how risky you want to get, how entailed you want to get for the assistance of using the product?
28:41Well, we're flexible enough. We can go either way.
28:45If you just want to turn your mind off, you don't have an opinion about this race, but you want
28:49a little action, we'll give it to you, okay?
28:52If you do have an opinion, though, we have a system, a proprietary system of asking you who you like
29:00in the race, who you don't like.
29:02Often that's more important, right?
29:04It's not so much who you are.
29:06Hey, that horse has no chance to win, so let's just take them out of the equation.
29:11So we have some flexibility in our approach and how we designed our systems so that you're able, as a
29:18consumer, to have as much input or not as you possibly want.
29:26And that can differ on a race-to-race basis.
29:31What we're really looking to do is meld your preferences or lack thereof with what we're seeing through our strategies
29:41and through our odds balancing.
29:44And as a 2016 winner of the British Cup Battling Challenge, how would Wager have assisted you on those days
29:51of racing to get to the top?
29:53So first off, I can't believe it's 10 years ago already.
29:56Thank God our geek got his head up there over California Chrome.
30:01It was really an amazing day at Santa Anita, be there with our friends, and we just had a great
30:08day that day.
30:11Well, it started, I'll give you an example, and we weren't doing it at that time, but that's when we
30:18started thinking about these things, and it's exactly why we won.
30:25We played a straight Arrogate California Chrome Xacta.
30:30Now, yes, did we really like Arrogate?
30:33We did, but what I did at the time was I just looked.
30:38I said, hey, I'm going to need like $50,000, $60,000 here just to compete for the title, and
30:45I think I had $13,000 on at the time.
30:49And we just read the board, right?
30:53I had a bunch of friends with me, and we just read the board and found every combination that paid
30:58more than we thought it should.
31:02And that's what we got out of that, right?
31:06That combination paid almost $10 for two, and I think I'm getting that right, and one horse was like the
31:15eight-to-five favorite, and one was the nine-to-five second choice.
31:18So we got something like 30% value on our money there, and anyone statistically inclined, don't hold me exactly
31:29to that number, but it's close enough.
31:30So we got a huge value, and then on top of it, there's the equity in the contest.
31:37So we're always thinking about value at the end, right?
31:40You're always thinking like, okay, how do I leverage my selections by finding value?
31:48So while we didn't use a computer then, we used that style of analytical thinking that's now embedded in wager,
31:57right?
31:57So if you want to win a contest, you need to find value and leverage, right?
32:03Tony Zhao, who's probably the greatest contest player of them all, okay, he's looking for value constantly, constantly looking for
32:12value.
32:13You want to get paid $5 when you should be getting paid $4.
32:17That's how you leverage your advantage in those contests in particular, but it's that same sort of analytical value hunting
32:29that we're hoping to ingrain in wager.
32:32And as artificial intelligence is more accepted in general society, do you think it'll be adaptable to horse players who
32:39are quite oftentimes, you know, let's say I know young horse players,
32:42but mostly times it's, you know, close to retirement or just pre-retirement age players every day at the track?
32:49You know, so first off, I'd like to say, and I'm sure there's some marketing people out there think this
32:55is wrong.
32:55Like, I don't really love the term AI.
32:58It's just kind of a catch-all, right?
33:00For machine, are we talking about machine learning?
33:03Are we talking about chatbots that can have a conversation with you?
33:08Like, what are we talking about?
33:09Are we talking about a neural network?
33:12I would say this, is that, and I referred to it earlier, no one wants to get into the details
33:19of what our neural network looks like.
33:21I mean, I do.
33:22I find that fun.
33:24And our analytics team thinks that's fun, right?
33:27And trying to decide, should we use XGBoost or should we use PyTorch or what package of analytical language and
33:40software we should use is interesting to me.
33:42But to the consumer, even a technically adept consumer, they don't want to hear any of that, right?
33:49They want to know that these picks are coming from some sort of analytical framework, that the analytical framework's rigorous.
33:58And quite honestly, they just want to know if the picks are competitive or not, right?
34:03So what we try to do is translate the math, we try to translate the code to something that a
34:13consumer can use, kind of regardless of age, you know?
34:17Now, having said that, I do worry horse racing fans are typically not, or at least the ones who've been
34:25around a while, are typically not that amenable to new things and new thinking and new technologies.
34:33And just, you know, I know, like when you look at the racing form and formulator's an amazing tool, most
34:41of these guys just want their PDF.
34:43Like, can you print out my PDF, right?
34:47Okay, so I get it.
34:48I assume there's going to be some, like, time delay in the uptick, but all the better for early adopters,
34:56all the better for people who come to wager early and get the benefits of that.
35:02Definitely.
35:02Well, I appreciate your time, Joe.
35:04And, of course, best of luck with wager and we'll be on the lookout and seeing it on the keypad
35:08for successions.
35:09Okay, Naj, good to talk to you.
35:12Billy, thank you for joining us.
35:15I know many in the industry are doing fine.
35:18I know many in the industry know you, of course, through your connections through Ocean Reef Racing.
35:22But can you tell us about your background?
35:24And for today's topic, how do you feel artificial intelligence will best help their bread racing today?
35:31Yeah, no.
35:32Great.
35:32Thanks for having me on.
35:35Honored to be on a podcast talking about AI and horse racing.
35:38You know, a lot of people don't know my background in racing.
35:41We started off racing a couple of guys at my club back home when we exited our last software company.
35:47Asked us to get involved and we were a very small part of one of the first five horses was
35:51part of Flyline, which was a lot of fun.
35:54And then fast forward a little bit.
35:55My father and I had a farm in Versailles, Kentucky, and we befriended a couple of guys, Charlie O'Connor
36:01and a few others, where we ended up having racehorses with them.
36:06Next thing you know, we had seven partners down in Florida, where I lived down in Key Largo, and up
36:12to 50 horses in racing.
36:13So a couple of years later in the Kentucky Derby, and now I think we looked at it today.
36:17We have parts of 72 horses, either in breeding, racing, foaling, pin hooks, the whole nine yards, parts of two
36:23stallions.
36:23So it's been fun to get involved.
36:26But my background is very interesting to others, not to myself.
36:30But I started off in technology and telecom back in 2008, came back to work with my father in technology
36:39and a managed service provider in 2011, where my team actually invented a software that we ended up taking globally
36:45to market.
36:46So our job was to move data packets back and forth from multiple locations across the world.
36:53So they would call it big data and software at the time, but now people would probably call it similar
36:59to AI, but we were basically an engineer in a box.
37:01So all software really is, is you do something faster, better, stronger, quicker, and cheaper.
37:07And AI is just that on an exponential scale, right?
37:11So AI and horse racing, I think, is a very interesting topic.
37:15We use a lot of it.
37:16But if you're not using AI in today's world and you're not learning, you're not living, in my opinion, because
37:24this, if you look back in 100 years, you're going to see that the invention of AI is going to
37:29end up being bigger than the invention of electricity, the automobile, and the Internet combined.
37:34I mean, that's what's happening right now.
37:36This next two years is going to be incredible.
37:39From what we've seen two years ago to where we are today is just, it's just night and day, just
37:43night and day.
37:44And you've been able to use AI technology to also develop Equitrack AI.
37:49Can you talk about that company, which you've established and how it's going to be used to help everybody who
37:54participates in thoroughbred racing?
37:56Yeah, for sure.
37:57Sure. When I was out of this, when I exited tier three in the software business, you know, we went
38:02into the venture capital side, so making early on stage investments and pre-seed series A.
38:08We did a lot of stuff with the Elon Musk ecosystem.
38:10So the last couple of weeks has been great in that capacity.
38:13And I said I'd never go back into operating another software business just because of the 89 hours a week
38:18in the business travel, unless something presented itself to me that I really couldn't want to do.
38:24And most good businesses are built out of something that occurs or a problem that somebody is trying to solve.
38:30For me, when we got into the horse racing, A, I had no idea what I was doing from the
38:34bill pay perspective.
38:35I'll be first to admit, I was one of the most habitual bad payers of all time because you're sending
38:41me snail mail to my farm.
38:42I live in Florida, to my house.
38:44I'm getting email.
38:45I'm getting vet bills.
38:46I have no idea.
38:47I have no idea what vendors are approved or not.
38:49It was a nightmare.
38:50So it was taking me 90 days to pay bills.
38:52In still some cases, we've almost fixed that at this point.
38:56But, you know, what we did with Equitrack was we went and built an operating system that will ingest your
39:03bills in.
39:03So our bills gets into a system.
39:05It goes through with AI.
39:07It'll turn around and read them, checked by an equine bookkeeper who has the basis and the skill set to
39:14do so.
39:15They press approve.
39:15It's been looked at five times.
39:17It comes to me.
39:18I press approve as ocean reef racing.
39:20It goes out that night through what's called the EquiWallet, which is an FDIC-backed bank.
39:25And it syncs back to QuickBooks, IES.
39:29And then I have a real-time profile, a real P&L for that horse on what looks like a
39:35Facebook profile for a horse.
39:36So I've saved 70%, 80% of my time on a day-to-day basis on managing our syndicate or
39:41our interest.
39:42Or I would have left the sport.
39:43I mean, just point blank.
39:45Can you talk about the efficiency of especially people being able to use Equitrack or AI in general over just
39:53creating more data points, but the actual usefulness that you say in the productivity of, you know, having your operations
40:00more streamlined?
40:01Yeah, 100%.
40:02I mean, outside of Equitrack specifically, I mean, AI just in general in the day-to-day workspace.
40:08I mean, I have two little kids.
40:10Timing is the essence.
40:11My best work productivity or if I want to have me time, it's before the kids get up or the
40:15kids go to bed.
40:15The reality of it is with AI tools like Otter, Loom to do videos, Grok, ChatGBT, Claude.
40:22I can name another 15, 15, 20.
40:25They're really efficiencies that you can apply in your day-to-day life.
40:28You know, if you were to use Otter to record a Zoom call you're on or a conversation, you can
40:33go through and it can have notes for you.
40:35You can go in and create a standard operating procedure for you.
40:38It can turn around and send that email out for you as well, and then it can be tracked and
40:43turned around.
40:44That would take an hour in previous.
40:45You can upload it into HubSpot.
40:47Say you have sales calls you want to do, those type of things.
40:50So it's all about efficiencies back in somebody's day.
40:52I implore people to use it.
40:53Actually, I live in a place down in Florida where there's a lot of older people, and I say, hey,
40:59use the Otter app or use an app when you're sitting there talking to your doctor, and then you can
41:02take it.
41:03You remember what they're going to say.
41:04They're going to send it to you.
41:05Go upload it in Claude and say, hey, I want a second opinion of a doctor.
41:08You know, what should I be doing if I want a second opinion or if I want to go take
41:12more medicines or whatever for my joint pain or whatever?
41:16I mean, use whatever application.
41:17It's an efficiency thing that you can call on experts because there's big data language in the background.
41:23So more opinions, more facets, right there at the tip of your tongue.
41:26And certainly in acquiring new owners is what thoroughbred racing needs to do, and especially growing our game.
41:34How do you feel that AI and other tools can be used to help acclimate owners, whether from the technology
41:40world or other worlds, such as, you know, Jason Wirth with baseball and, you know, diversifying and broadening our sport?
41:46Yeah, no, great question.
41:48I think that the best analogy I can give to people, especially Jason Wirth being mentioned, this AI can be
41:54money ball for horse racing, point blank.
41:58You're going to, within the next three to six months, be able to go and say, hey, I want to
42:03go run a mile race on the turf at Del Mar.
42:06Okay, what of these 5,000 horses entered at Keeneland in the sale?
42:11I want to upload those pedigrees into a tool coming out or into Claude, and it'll give you those 30,
42:18right?
42:18Now, you still have to vet it.
42:20You still have to go and have somebody look at it that knows what they're looking at.
42:23I'm not the person to go look at it, you know, a bloodstock agent or somebody like that.
42:28And then you have to put your value on it.
42:30But that, to a new person coming into the game, can be tremendous, right?
42:35I wish I would have had that years ago.
42:37It might have changed the way we bought certain things or whatever.
42:39I think to have other new people into the game, you're going to see more AI analytics that are huge
42:44in gaming.
42:44I mean, I got more people asking me right now.
42:46They want to go load the program because no one knows how to read the thing anymore.
42:49The DRF in the program, a 19-year-old doesn't know how to use that.
42:52A 15-year-old doesn't know how to use that and those type of things anyways.
42:57You know, they load those in and say predictive analysis.
42:59That's the way the sport's going.
43:01That's the way we got it.
43:01We got to be easy to buy from on the breeding side at the Keeneland sales so they can put
43:07it in nomenclature term.
43:08You got to be easy for the high school kids, middle school kids, or college kids to understand what they're
43:12looking and press the easy button.
43:15And the easy button, if you get it, AI will allow the easy button to occur in all aspects,
43:19whether you're young and just going to the track at $2 across the board or you're, you know,
43:23a billionaire spending $20 million and you don't want to spend money on tax incentives, you know, that easy button
43:31out there will just get more people into the sport.
43:34You know, hands down.
43:34Earlier, you referenced that we're kind of in the new frontier of AI and this technology, especially within the thoroughbred
43:42industry space.
43:43Where do you see things going in 10 to 15 years from now and how in-depth we can get
43:47with all these tools?
43:48Well, look, you know, I'm sure I'll offend some people when I say this.
43:52I mean, the old adage of, you know, the horse racing industry is, you know,
43:56they're high-end farmers at the end of the day that breed a very expensive asset class.
44:02You don't become a horse trainer because you love sitting behind a computer to do administrative work or accounting.
44:06I can tell you that.
44:07I don't really like doing accounting and I have an MBA from SMU.
44:10I mean, I hate it.
44:11But it's just part of what business is.
44:13The reality of it is the horse racing business has always been 15 to 20 years behind on technologies
44:20because for whatever reason, it's not forward-facing on technologies.
44:25It has to be.
44:27We're an agriculture industry.
44:29It goes into a gaming industry.
44:31And it's a very small sector of people that make money in those, you know,
44:35in between of, you know, getting from, you know,
44:38Inception to the racetrack to the Kentucky Derby, the Triple Crown,
44:42and beyond to Ascot and all those different things.
44:46Your question on the early aspect of it, I think that where we are now,
44:51I think in the next two years, you're going to see AI and the administrative side
44:56of being able to pay somebody as quick as Venmo-like payments is going to be almost done in the
45:02next six months.
45:03I mean, we do that now.
45:04But the idea of if I could win a race today at Belmont at the Big A on the purse
45:07money
45:08and I pass the drug test tomorrow and it hits the Encompass and the Paymaster
45:14and hits my Ocean Reef racing account,
45:16but today I happen to buy a horse in Arcana or Golfs or somewhere around the world,
45:21that I can move that around the world and pay by Tuesday with the same money,
45:25moving around in the flow of funds.
45:26That's where it's going to go very quickly.
45:28That's what fintech and actually blockchain was meant for, not cryptocurrency.
45:32That's what the idea was.
45:33Where we are in 20 years, that's, I mean, I can't look more than two or three years at it.
45:39I think rocket ships, I think we're going to be flying around on golf carts from racetrack to racetrack.
45:44I don't think that's far off either.
45:46I mean, I do see a bunch of venture, but specific for AI, for horse racing,
45:50I think you're going to see a lot of analytical tools that can go so far down into the bloodlines.
45:55They can go so far down into historical race results on return on investment
46:00and capital on broodmares on, should I put this stallion with this broodmare?
46:05Because historically what it can win on the track versus the analytics on the business side of it
46:10that you can get through companies like Intuit and IES, which is, you know,
46:15we use QuickBooks as IES license with Equo Account, and you can slice and dice it 10 different ways.
46:21So there's going to be a bunch of different ways it'll go.
46:24I think what AI will do is it'll bring more people into the sport if we deploy those things,
46:29press the easy button for the new young kids.
46:32I think what we need to do is focus down market on the sport to get more younger people,
46:37and I'm not talking about in their 20s.
46:38We've lost them at that age.
46:39We need to get the kids that show up to the track, that see the babies in the falls on
46:43the farms,
46:44that will go up and do those things that Keeneland does every,
46:47I think it's every Saturday they bring people in and let the kids come out there in the mornings,
46:52what David O'Rourke does up at Naira and takes families that have never been in the racetrack
46:56and takes them into the paddock and takes them into the winner's circle.
47:00In some cases, those kids are hooked.
47:02In 20 years, those kids might have a lot of money.
47:04They might be your next breeder.
47:05They might be your next Mike Propoli.
47:06They might be your next, you know, Gary Barber or whomever.
47:09I mean, that's what you want to see, right?
47:12Definitely.
47:13Yeah.
47:13And AI with that easy button.
47:15I mean, that's easy for a kid to understand.
47:17I mean, there's applications that are out there now that I've seen where,
47:21think about the horse profiles for, you know,
47:24basically a horse profile or a Facebook-like profile that you can let kids that come through those gates.
47:29You know, look, we've never actually done anything with the kids at these racetracks
47:33because we couldn't sell them a product because gambling is illegal for under 18.
47:36There's nothing wrong with giving that kid a phone number or an email address
47:40where you give them a free profile of the horse.
47:43Let them pick their top five favorite horses, you know.
47:45And from those top five favorite horses, you know, they can follow them along.
47:49If they get a workout, they get an update from, you know, Equibase.
47:52They get pictures if there's uploaded.
47:54They get a race replay from Twitter.
47:56You know, all different types of things.
47:57That kid's hooked.
47:58He's it.
47:58Definitely.
47:59And can you talk to just the advantages of using Equitrack, especially on if you're a farm owner
48:05or a horse owner or a veterinarian, you said, I know billing and getting that capability
48:10and process of getting your, you know, statements out monthly.
48:13What other tools that Equitrack possess to help participants?
48:17Yeah, no, I mean, the main one point blank is I get 70% of my time back.
48:22Period.
48:24Look, if an owner adopts it, he's going to be able to get a real P&L.
48:27He'll know to move a horse on faster.
48:29He can go and say, you know what, it's going to take six months to earn that back on the
48:32track.
48:32Let's just go breed her, you know, or move her on, whatever, which way they want to do.
48:36From a vet perspective, if they go and send us directly the vet bills, we send those to
48:40our clients that are out there.
48:41They're going to get paid faster.
48:42So you're not going to have net 45 days.
48:44You know, we calculated the flow of funds that this really will with the EquiWallet will allow
48:48somebody to purchase an actual extra half horse.
48:51I know you can't purchase a half horse, but the flow of money will help people get more
48:55horses and grow the sport.
48:57From the breeding side of the house, no different on that, on stallion contracts and those things.
49:04There's automated stuff like that in there.
49:06So just think the administrative back office of having to have five, 10 people, or if a trainer
49:11to have three people to go and move your digital certificate around, to go mess with
49:16HISA, to do your Coggins, to do all this stuff.
49:19It's all just seamless and streamlined, right?
49:22You're going to be able to do more with less, right?
49:25And that's the only way the sport grows.
49:26I mean, the other side of it is bookkeeping.
49:28Bookkeeping is done.
49:30AI will eliminate all bookkeepers' jobs in the next two years.
49:34It can probably do it.
49:34It can do it now.
49:35I know it can.
49:36The difference is the bookkeeper that deals with their finances and accounting right now
49:39ends up being more of the bedside manner of somebody that's talking to their doctor
49:43or somebody who's dealing with a dying parent.
49:44I'm not going to trust.
49:46You're not going to trust, sure as heck, the horse people aren't going to trust AI doing
49:50their bookkeeping.
49:51They're still going to want to talk to Betty Sue or Johnny, who's been doing their books
49:55for years.
49:55We can help that person do more with less, and they can get a more real-time P&L, but
50:01they start talking to them that they're comfortable before the taxes, the paperwork, and the files
50:05get sent over to an accounting firm for the tax thing.
50:08So those are the best ways of efficiency.
50:11I mean, if you go to every horseman and say, I'm going to give you four hours back in your
50:13day, or not your day, a week, what are you going to do?
50:17Some are going to go work more.
50:18Some are going to go golf.
50:19Some are going to spend time with their kids.
50:21Some are going to go look at horses.
50:22Some bloodstock agents are going to go look at more horses.
50:25Four hours back in somebody's work week is a tremendous amount of time, and that's the
50:29minimum you're going to get back if you have 10 horses or more.
50:33Definitely.
50:33Well, I appreciate the time, Billy, and it sounds like AI and, of course, Ecotrack can be
50:37used to forward that easy button access for everyone.
50:41Yeah, no, it can't, and I'll leave you with one more thing that I just got off the phone
50:44with, with somebody who's the actual rocket scientist in horse racing that's at a breeding
50:48operation in Lexington.
50:50You're going to see in the next year, you're going to be able to see AI that can ingest race
50:55replays, walking videos, that will see the cadence and the way that those bloodstock agents look
51:00at how those horses come down, that will be able to put some kind of scoring on, can that
51:05horse go and be a group one winner or a grade one winner?
51:09You know, there's going to be a lot of skeptics on that.
51:11I'll be the first to believe.
51:12I've been to a hundred horse sales.
51:14I can go look and tell, you can tell me that the thing is crooked.
51:17You can tell me that it walks or it's over or whatever.
51:19And I, I just don't see it.
51:20That's why I'm not the one that picks our horses in that case.
51:23But when those types of tools come down the line, layered on with the AI tools and the
51:28return on investment and the bloodlines and that easy button, the sport's going to end up
51:32growing very rapidly over the next 10 years.
51:34And I really believe that.
51:35Definitely.
51:36Like you said, money ball for horse racing.
51:37I'd be on board with that.
51:38Money ball for horse racing.
51:40Definitely.
51:41Thanks, Billy.
51:42Hey, thank you for having me.
51:43I appreciate it.
51:45Thank you for joining us.
51:46And we hope you learned more about artificial intelligence and how its impact will affect
51:50thoroughbred racing today and in the future.
51:53Thank you to our guests, Josh Isner of Delta Squad Racing, Billy Painter of Ocean Reef Racing
51:59and Equitrack AI, and Joe Applebaum of Wager.
52:02We'll look forward to seeing you next time on the TDN Business Hour.
52:09For more information on the TDN Business好像 in 1010,시죠.
Comments