Latest / Rusty's Garage / The Motorsport Brief | How AI is changing Motorsport

Transcript

The Motorsport Brief | How AI is changing Motorsport · Rusty's Garage

• 0:02
  A listener production.
• 0:05
  How is AI going to change motorsport and how is it already being used?
• 0:09
  It's a panel discussion on this ep of the Motorsport Brief.
• 0:15
  G'day everyone, Rusty here in the garage.
• 0:17
  Shortly to our guests, and rather than interview them separately for this edition, we'll have them both on at the same time and we can bounce around some conversation on today's topic.
• 0:26
  First, a little plug for our newest feature episode, which dropped in recent days.
• 0:31
  Larican motorcycle racer Robbie Phyllis joined us.
• 0:34
  We recorded that at a workshop in Albury, surrounded by some cool old race bikes of his.
• 0:39
  He was ahead of his time in some respects, ahead of the world's superbike growth curve 10 years too early in some ways.
• 0:46
  Winning at Spa in Belgium with some ballsy moves, his friendship with Peter Brock, how the late Johnny Harvey ran strategy for him at the endurance motorcycle events and the chaos that he caused in Japan at the go-kart track at Sugo and hot-rodding some golf carts.
• 1:01
  You'll enjoy that.
• 1:02
  Some good laughs in that one as he rose through the Australian scene in the 70s and 80s.
• 1:06
  before showing some true grit as he raced on the world stage.
• 1:10
  Plus, a special Norton that he helped to develop for Daytona and the 200-plus career crashes that he had along the way and how it affected him.
• 1:21
  Two parts.
• 1:22
  Over two hours for you to enjoy.
• 1:24
  Let's get to today's guests.
• 1:25
  And I do not profess to know much about AI at all.
• 1:30
  It's here.
• 1:31
  It's expanding its presence in our lives at a rate that is almost impossible to fathom.
• 1:36
  Matt Oxley actually wrote a really good piece on this in 2024 for Motorsport Magazine on how AI and machine learning is helping MotoGP and Ducati especially.
• 1:46
  So what about in our part of the world?
• 1:48
  What does it have...
• 1:50
  in motorsport what kind of potential is there and so on first up an old colleague trevor long who i worked with in radio what seems like a lifetime ago now since then he has taken his love of tech and gadgets to become a leading commentator in this space on the today show his own eftm platform um the long-running two blokes talking tech podcast with his great mate stephen fenwick
• 2:13
  We're also joined by Tom Moore, respected supercars engineer at Dick Johnson Racing these days.
• 2:18
  He's also worked at Erebus, Melbourne Performance Centre in their GT program and more.
• 2:24
  Trev, Tom, thank you both very much for coming on.
• 2:27
  G'day, Rusty.
• 2:27
  All good.
• 2:28
  No doubtless.
• 2:28
  Thanks for having us.
• 2:29
  It's great to see you both and to chat with you both.
• 2:31
  Trev, can I start with you?
• 2:33
  We probably don't need to define artificial intelligence.
• 2:35
  I think most people know about it.
• 2:37
  Some are nervous about it.
• 2:38
  How significant is this technology and how much will our lives change in the next few years because of it?
• 2:45
  It's the most significant technology we've seen in a long time.
• 2:48
  We think of eras in the world of tech.
• 2:50
  So we had the internet era, I mean, the mobile era, and this is the AI era.
• 2:55
  This is going to change pretty much everything we do, whether it's happening to you now or whether it happens in the next few years.
• 3:01
  It will have an impact from just simple things that we do day to day through to our jobs using AI, perhaps in some ways that we don't even know just because the software and the tools that we use are embedded with AI.
• 3:12
  So sometimes it's a gimmick of a word,
• 3:15
  But a lot of the time, it's really supercharging the tools that we use.
• 3:19
  And that's exciting for people and jobs.
• 3:22
  Tom, engineers like you are curious types by nature.
• 3:26
  You typically embrace new tech, particularly if there's gains to be had in it.
• 3:31
  When did you first kind of hear about it, begin experimenting with it, learning about it and so on?
• 3:35
  Probably wasn't until maybe only 12 months, two years ago, where it was probably more the chat bot sort of generative technology.
• 3:44
  AI, you know, that would speed up a lot of processes we were doing at work.
• 3:46
  You sort of, there was a lot of things that we knew were possible, but it's like, oh, that's a software engineer's job for six months to make that happen.
• 3:52
  And now it's a data engineer's job for a week and you can sort of do it.
• 3:55
  So there's sort of, yeah, that's really what I think where we first started implementing into the sport.
• 4:01
  And then as it's becoming more widely used,
• 4:04
  Many companies are producing software or off-the-shelf products that are then being advertised to us that can do things that we couldn't do before.
• 4:11
  So there's two sort of facets to it.
• 4:12
  One that's made it a lot more efficient to do work that someone could already do, but it makes it available to more people to achieve the same amount of work.
• 4:18
  And then there's features that are now possible because of the machine learning in particular and those sorts of things.
• 4:24
  Okay, we'll come to that in a bit more detail shortly, perhaps how it's helping you in a racing sense and what it's doing perhaps from an industry perspective, which you've alluded to there, right?
• 4:32
  Trev, there is kind of talk of high net worth individuals investing in things like festivals, events, because human contact will – or connection, I should say – will become even more important as we go forward here.
• 4:45
  We'll get to the how, but kind of broadly speaking, is sport, is motorsport kind of well placed to benefit from AI in that sense and why?
• 4:55
  Yeah, I think so.
• 4:57
  There's a lot of – because the concern there and the reason people are looking at those in-person things is about the fakery that's existing online.
• 5:04
  So videos, we scroll for hours on social media platforms, and we're getting to a point where you're like – you see a video and you go, is that real?
• 5:11
  Can't tell.
• 5:12
  You can't tell.
• 5:12
  It's so weird or it's so outrageous.
• 5:15
  And it used to be really easy to pick.
• 5:16
  And I'm talking six months ago.
• 5:18
  So it's advancing so fast that it's very hard to pick.
• 5:21
  And so I think that we're going to get to a point where credibility from personality.
• 5:26
  So think about our V8 supercar drivers, right?
• 5:28
  Supercar drivers have personalities.
• 5:30
  The cars, the brands have followings.
• 5:33
  And so we're going to get to a point where people are going to be drawn to the things that they have a passion for and they love.
• 5:39
  And so they will want to go and see the real thing as opposed to just scrolling through it, which is great for in-person events and in-person attendance.
• 5:48
  But it also kind of validates the official channels of like teams and sports that become the credible place to find that thing.
• 5:55
  So if you see...
• 5:57
  a pass or an incident or something, then you go, hang on, is it real?
• 6:00
  And you might search up DJR and go, oh, they've posted it.
• 6:03
  It's like that really did happen.
• 6:04
  That validity of real content is going to be critical over the next few years.
• 6:10
  Okay.
• 6:11
  Tom, motorsports, a game of closely guarded secrets.
• 6:14
  So in exploring a bit more of what you said there before, I don't want you necessarily to
• 6:18
  share proprietary things from DJR there.
• 6:21
  Maybe you can talk more about it in a broad paddock sense.
• 6:24
  Tell us ways that AI is perhaps already being used in the pit lane and then more broadly in the industry.
• 6:30
  Yeah, so we probably, like I said before, there's
• 6:34
  We're using it sometimes on a day-to-day basis as far as, you know, our world is full of spreadsheets and tools of software and that sort of thing of, you know, we have endless amounts of data that we're always trying to process and draw trends from and things like that.
• 6:46
  So, if AI is really allowed, you know, you would often think, oh, there'd be a way to do this, but I'm not, you know, unless you're a software engineer or, you know, a real, you know, computer science, something like that.
• 6:58
  I know it's possible, I don't really know how to do it.
• 6:59
  Now you can go straight to your Claude, your co-pilot, your ChatGPT, that sort of thing and go, make me a tool that does this assist to this data set.
• 7:07
  And it goes, yep, there it is.
• 7:08
  And you can finesse it a little bit, troubleshoot it a little bit, and all of a sudden you've got a tool that used to take weeks, if not months, to make in a matter of hours or a couple of days, depending on how big it is.
• 7:19
  So that's probably where it's been used in a more day-to-day sense.
• 7:23
  And then there's also things like, particularly with the machine learning, things that...
• 7:30
  we use like pit stop video for an example.
• 7:32
  There's now machine learning products that can watch the video of a pit stop and it will pick up, it's been trained to know when the gun reaches the nut, when the wheel comes off the car, when the wheel goes back on the car, when the car arrives in the box, when it leaves, when the spike goes in, when the spike goes out.
• 7:47
  And then it can process that data in an instant and generate your report, clip the footage and that's done.
• 7:52
  So that used to be, that could have been a data or performance engineer's job for two, three hours on a Saturday night.
• 7:57
  Now it can happen instantaneously.
• 8:00
  that's sort of where it's being put in more of a commercial sense as a product.
• 8:04
  But, you know, we couldn't make ourselves in short amount of time, but there's guys that, you know, your computer science guys and that sort of stuff that now go, this is another really good tool I have.
• 8:12
  My strength is not necessarily like your computer science guys are not necessarily, their strength is just in writing code.
• 8:18
  Now it's utilizing AI, the machine learning, that sort of stuff to create products that, yeah, were simply not possible a matter of years ago.
• 8:24
  So, yeah.
• 8:25
  I'm glad you bring up machine learning.
• 8:27
  That was something that in that article I alluded to at the top that Matt Oxley did around MotoGP that he talks about too.
• 8:31
  So Trev, probably worth telling people a little bit more about that.
• 8:35
  You know, if I'm Googling here, it fundamentally is a subset of AI, is it not?
• 8:39
  How is it?
• 8:40
  What does it do?
• 8:41
  And how does it sort of work in the manner that Tom's just talked about?
• 8:44
  So, I mean, traditional computing is lines of code that says, if this happens, do this.
• 8:49
  And then if this happens, do that.
• 8:50
  And it's really a complex set of commands that tells a computer how to operate.
• 8:55
  And you think about computer games, you know, if you point your mouse this way, the head turns that way, right?
• 8:59
  So that's computer programming of old.
• 9:02
  But the modern world of computer programming, and machine learning has been around for a while, long before we used the term AI, is about your device or the chipset being able to understand more and actually inherently make a decision based on its knowledge.
• 9:17
  So I guess the simplest way to put this would be if you feed a machine enough information.
• 9:23
  I'll give you an example.
• 9:23
  I remember being with Oracle in Melbourne at a Red Bull event, and they were talking about the billions of lines of data that they get from a Formula One race and the number of potential strategy options they could come up with, you know, in a computer set.
• 9:37
  It would run through the cloud and do all this stuff.
• 9:38
  But now, if you have machine learning in there,
• 9:42
  not only can it ingest all that information, but that it can in real time look at what's happening and be asked a question about that and not need to go away and process for days and hours, but potentially draw upon recent iterations of that same thing or scenarios that it's learnt and make decisions or offer up opinions based on that.
• 10:03
  I think we're probably already seeing at the absolute pinnacles of motorsporting.
• 10:06
  possibly even in supercars that Tom and others can't talk about, but we're probably already seeing some unbelievable use of machine learning in things like strategy and basic decisions that are made day-to-day on a race weekend.
• 10:21
  Tom, people are listening, they can't necessarily see you.
• 10:23
  I can.
• 10:24
  You're nodding a little bit there.
• 10:26
  Just go to that if you can around how it more broadly may assist with things like strategy, perhaps performance of the machine or perhaps even the athlete.
• 10:38
  Yeah, definitely.
• 10:40
  It's definitely, we're always looking at ways to fast track our decision making or even automate our decision making.
• 10:45
  So a huge part of being a race engineer is that you need to make a decision in a split second that needs to be correct.
• 10:50
  And you've got more often always more data than you can ingest at that point in time.
• 10:55
  So every team has, for example, their timing software that they run in front of them during a race or a session, which is deciphering a huge feed of timing software.
• 11:06
  timing loop data and that you know makes it legible for us to be able to sort of understand so if there's another layer that we can put on top of that via machine learning ai what have you that could then go okay i know what you're thinking or what you're trying to achieve this is actually what you should best do so that if there's a safety car comes and you go all right i should be pitting now i shouldn't be putting now if i pit now i'm going to come out here what effect does that have on the cars in front behind me etc etc so you're always trying to make the best decision in the shortest amount of time possible so if there's any way that we can speed up that
• 11:36
  it becomes, yeah, really, really beneficial and crucial to, yeah, performance at the end of the day.
• 11:41
  This is fascinating to you both.
• 11:43
  Can I get you to hold there for just a moment?
• 11:44
  Our conversation around AI machine learning in motorsport continues right after this.
• 11:53
  We may not get to every benefit or aspect of how AI will impact motorsport in this part of the world during today's episode of the Shortcast, but it is good to kind of start a conversation around it.
• 12:02
  I'm with Trevor Long from Two Blokes Talking Tech and The Today Show, as well as Tom Moore from Shell v Power Racing.
• 12:08
  Let's launch back into the chat now.
• 12:11
  Can we talk, Trev, about the broader benefits, say, for a fan at the event or maybe even watching the coverage?
• 12:18
  What impact will AI have in that regard?
• 12:21
  I think it's super exciting to think about both those aspects.
• 12:24
  So from a broadcast perspective, again, go to the pinnacle and look at Formula One.
• 12:28
  You already see the use of data, and I would argue they're already empowering that with AI.
• 12:34
  Through things like AWS is a big sponsor of the Formula One.
• 12:36
  That's just Amazon's cloud services.
• 12:38
  But you get this thing where, you know, how many laps until Oscar catches Lando?
• 12:42
  Or who's going to be in front at the end of this pit stop?
• 12:44
  that's using data.
• 12:46
  So data from in-laps, out-laps, fast-laps.
• 12:50
  And while you could always do that by going, well, he is six-tenths faster every lap, it's going to be this.
• 12:55
  The difference now is it can bring in things like with the knowledge of six other drivers have come in for softs and their out-lap was this much faster, it can bring all that information in and give us a far more accurate potential assessment of what's going to happen in the race.
• 13:10
  So I think data and visualising data for people
• 13:14
  is so much more exciting.
• 13:15
  It's funny to watch races that we used to watch as kids, Rusty, and you see there was just this, you know, simple little timing chart on the bottom of the screen.
• 13:22
  And now we've got so much data, we want more.
• 13:25
  And, you know, on the Supercars app, you can look at, you know, real-time data from cars.
• 13:28
  Like, it's unbelievable.
• 13:30
  So I think that we will start seeing things like predictions and estimations of things.
• 13:35
  I think you'll even get to a point where if I'm Supercars, I'm allowing people to play strategy.
• 13:40
  You know, you're at the 1,000.
• 13:45
  I think we should pit now and what would happen if we did that?
• 13:48
  It's completely possible that the user in an app as a kind of while watching the experience could be doing that.
• 13:53
  Another one that's happening in broadcast, KO, for example, they have a new app where they are automatically clipping events.
• 14:01
  And it happens in supercars as well because the broadcast vision is quite wide, whereas you're watching it on a phone.
• 14:06
  So it actually uses AI to go follow the car in this clip and video.
• 14:11
  And I think at an actual event, so if I think about Albert Park, that is a busy place.
• 14:17
  I don't know that you'll see AI in daily use, but imagine, and this is going to be a ridiculous example, but you know the queue for the toilets are ridiculous over the other side of the golf course?
• 14:27
  Let's introduce into the AGP app some AI that helps us understand when the queues are going to be the less.
• 14:33
  Not because of time of day, but because actually there's a race on it this time.
• 14:38
  That's going to happen at this time.
• 14:39
  There's rain predicted now.
• 14:41
  This is your best window to go to the bathroom.
• 14:42
  I mean, you could use AI for that, totally.
• 14:43
  Totally.
• 14:46
  This is probably more of a question for the rule makers, but I'll put it to you both.
• 14:49
  What about the umpire, race control, perhaps even policing ways that we do or don't use AI?
• 14:57
  My question probably sounds a little bit like a student here and reminding kids either at university or school that they can or can't use so much AI in assignments and so on.
• 15:05
  Could you see that level of involvement?
• 15:08
  I think so, yeah.
• 15:09
  I mean, it's sort of had his argument about sport and, you know, the...
• 15:14
  ball tracking and cricket and that sort of thing, and do we trust that the half width of the ball has to be hitting the stumps and that sort of stuff.
• 15:19
  You always need some sort of human adjudication over it, and I don't envy Craig Baird whatsoever.
• 15:24
  He's got the hardest job in pit lane as far as deciding who was at fault and the intentions of each other at the time.
• 15:29
  But, yeah, there was – I'm not sure if they actually still use it, but they used to use Hawkeye technology for curb hops and that sort of thing at street circuits, and there's – yeah, any –
• 15:40
  any technological advantage to make those things easier, potentially more repeatable, more consistent, because that's obviously like a regular argument is, you know, is the rules the same, every track for everyone, that sort of thing, and, you know, all we want is consistency, blah, blah, blah.
• 15:54
  So anything that, yeah, assists those decision-making, yeah, it's invaluable.
• 15:58
  I think it's twofold, Rusty.
• 15:59
  I think you'll find people like Craig Baird and others can, you know, when they release information about why they made a decision, they could potentially demonstrate it and go, well, for four laps consistently, this was the...
• 16:09
  The breach, whatever.
• 16:12
  And whatever.
• 16:12
  And this was a different – he did it differently this time, and that's why we did the breach.
• 16:17
  But I also think that if you look at it from a – could AI give a team a massive advantage?
• 16:23
  And you've got to be real here.
• 16:24
  AI isn't cheap.
• 16:26
  It's fine for all of us to muck around with ChatGPT for free, but when companies like race teams want to throw data at it, those companies are going to want money, right?
• 16:34
  So does the top end of the grid –
• 16:37
  have the ability to use AI more than the bottom end?
• 16:40
  You know, do we need to even that out in some way?
• 16:42
  Do we need to be worried about that, you know, in a rule sense?
• 16:45
  I don't think you'd be able to, you wouldn't be able to monitor it.
• 16:47
  It'd be impossible because they're just computers and clouds.
• 16:49
  So we kind of need to normalize it.
• 16:52
  And we need to hope that people that come up with those systems that benefit teams are able to make it available to as many teams as possible.
• 16:58
  So that perhaps the little team at the end of the grid
• 17:02
  will spend more on AI and actually reap some advantages on the racetrack through strategy and other decisions that they make.
• 17:09
  You never know.
• 17:09
  You're talking, both of you, about it at the moment as though it's a very valued assistant, that it's in collaboration with what you're doing, so to speak.
• 17:19
  Can you see it getting to a point relatively soon?
• 17:21
  I mean, they're talking about
• 17:23
  big changes in the world by that stage?
• 17:25
  Could it get to the point where it's perhaps taking over jobs, maybe creating new ones in motorsport that we don't know about yet and so on?
• 17:33
  Funnily enough, we actually were approached by a company in the States recently that are trying to develop kind of exactly that.
• 17:41
  And it's an AI agent to dummy a race engineer effectively.
• 17:45
  And they've started at a data slash performance engineer level.
• 17:48
  So what they've got is a product with a
• 17:52
  they've got a bit of software that sits on your data stand, for example, and every time you download the car, your log data goes to that bit of software.
• 17:59
  It gets uploaded to their cloud.
• 18:01
  Every time you download footage, it goes to their cloud.
• 18:03
  And they've got an agent that then clips up the data
• 18:07
  clips up the footage for each corner and then overlays two sets of data and then writes notes on each corner.
• 18:14
  So something that typically a data or performance engineer would do on a Friday, Saturday night at an event and go, oh, these are the two fast laps between the drivers.
• 18:21
  Guys, here's a little blurb on each corner.
• 18:23
  It now does that in the time that by the time the driver gets out of the car and gets back to the truck, it's in an email on their laptop already.
• 18:30
  So it's...
• 18:32
  There's obviously still things you might not replace someone entirely, but you look at the value of that and go, well, how much time am I freeing up for my data or performance engineer by having this tool?
• 18:41
  Is it then worthwhile?
• 18:43
  And quite often, even having it...
• 18:46
  faster is not actually possible with the amount of personnel we have at the track and that sort of thing.
• 18:50
  You might not normally get it till Friday or Saturday night, but if you can have it in between a qualifying run or straight after practice and that sort of thing, it becomes really valuable in that point of view.
• 18:59
  So yeah, there are definitely companies that are trying to look at what are the tasks that these guys complete?
• 19:03
  Can we automate them?
• 19:04
  Can we make them faster?
• 19:05
  So it's definitely on its way.
• 19:07
  I don't see it costing jobs, Rusty.
• 19:09
  I think that supercars is already so lean.
• 19:12
  You know, it's not like – I think a supercar team would look at a Formula 1 team and go, I wish we had that many people just on the pit wall, let alone in the garage, right?
• 19:20
  So I think supercars is safe because it's only going to empower the staff and make their jobs more efficient and faster at delivering the data, as Tom said.
• 19:29
  But I think at the highest levels of the sport, there would be jobs that change completely.
• 19:33
  And they would be, there might not be a physical person doing that thing now, but there'll be someone else or several other people employed.
• 19:40
  It's the same in general society.
• 19:42
  AI will cost jobs.
• 19:43
  There are jobs that are not going to exist in 10 years from now.
• 19:46
  But there will be three more jobs created in the software engineering world, which just goes to a generational change in employment, really.
• 19:52
  But it's exciting.
• 19:53
  Yeah, I mean, to your point, that means no more Greg Russ commentary.
• 19:55
  We'll have an AI commentator who will get it right.
• 19:57
  That's great.
• 19:59
  Can I come to you just on its deficiencies, perhaps, Trev?
• 20:02
  I mean, you hear people talk about a thing called hallucinate.
• 20:04
  Yes.
• 20:05
  Clearly, they'll get on top of that.
• 20:06
  What is that?
• 20:07
  And how far away is that from being sorted?
• 20:09
  A hallucination is when an AI really wants to please you and it just gives you an answer.
• 20:13
  So if you say, can you tell me what happened on, you know, at this date, it'll just go, this happened.
• 20:17
  But if you actually research it, you might find that it embellished the truth or it just made it up completely.
• 20:22
  So that's a hallucination because a lot of AIs out of the box, as in when you log in, they're built to want to please you.
• 20:29
  They're built to want to give you an answer and actually what
• 20:33
  happen in a corporate world and I would assume in a team world is you need to build your agent to be very truthful and honest and be willing to not give an answer if it can't things like that so it's dangerous it's dangerous to go just out of the box ask a question get an answer so it's dangerous to say if this has happened at every Bathurst 1000 tell me what I should do next because we all know that you can win from the back of the grid and you can win having come last on the first lap it can be done
• 20:59
  So, you know, you can't just assume that because the last 50 races have been this way, that nothing extraordinary is going to happen.
• 21:06
  The fact is that AI can't predict everything, but that's why you need the human eye over those decisions.
• 21:12
  That'll be the fascinating thing if decisions are made based purely on AI and it costs someone a race or a podium.
• 21:20
  That'll be fascinating.
• 21:21
  Okay.
• 21:22
  To get to the almost checkered flag point of this shortcast, what are you both excited about in motorsport in terms of AI, its potential, things that we might see in the not-too-distant future?
• 21:32
  Maybe leaning into, Trev, some of the stuff you've had discussions around Formula One that are already happening.
• 21:38
  I would say as a fan who gets up at 10 o'clock, 11 o'clock to watch Formula One, as a fan who sets an alarm for a three o'clock supercar race, I enjoy watching the race, but I've grown to know that while I want to hear the commentary and I want to know what's happening with the totem, sometimes I want more.
• 21:55
  And we know that, you know, on KO and other things, you can sometimes choose different.
• 21:59
  We know we can get a timing screen.
• 22:01
  But now I'm going to want to ask questions.
• 22:04
  Like, what if this had happened?
• 22:05
  Can you tell me, would he have come out in front?
• 22:08
  So I think as a fan, if I can query the data...
• 22:11
  by saying, well, let's be real.
• 22:13
  If Oscar had done a one-stop, what would have happened?
• 22:16
  I think that's cool.
• 22:18
  And it's just a matter of whether the sports are going to make that data level available to the fans.
• 22:22
  And I think they should because I think true fans will engage more and, frankly, it's probably more revenue.
• 22:27
  I mean, if I can do that in the official app, then I'd probably pay for it.
• 22:32
  Tom, to you, what are you excited about on your side of the pit wall?
• 22:35
  And can I extend that question that I see this as a human game in the crazy sort of far-fetched sci-fi world.
• 22:44
  The thought of a race with far less or almost no human involvement I think would struggle from a fan perspective.
• 22:52
  Call me old school.
• 22:53
  I'm more than happy to be called old school there.
• 22:55
  What are you excited about?
• 22:56
  And can you see it being the kind of thing that still absolutely leans into human contest?
• 23:01
  I think one of the really cool things is it sort of opens up your imagination as far as what tools can I create to be better at my job.
• 23:08
  Ultimately, we're all extremely competitive.
• 23:09
  That's why we do it.
• 23:10
  And, you know, we are competing against every other driver and engineer in pit lane.
• 23:14
  So previously you'd go, oh, you know,
• 23:17
  I wish I could do that.
• 23:18
  Yeah, in the future, that would be a cool tool to have.
• 23:21
  But now it's just becoming like the growth rate is just exponential as far as things that you can, you dream up a tool and it will be produced in the near future or someone's working on producing it now.
• 23:31
  So thinking of, we're always trying to make, yeah.
• 23:35
  not necessarily our job's easier, but like you say, more efficient, more accurate, make more accurate decisions.
• 23:38
  Because quite often it might not necessarily be predictions, but if it's predictions based on data that we probably would have come to eventually, if we can make that decision faster, then that's the decision we want.
• 23:48
  So, yeah, there's always a level of will it be correct?
• 23:51
  And like Trev said, it's endlessly unpredictable, the sport, and you've got no idea what's going to happen around any corner.
• 23:57
  But if you can...
• 23:58
  you know, work to the odds and work to what's most likely to happen, then, yeah, it's any more tools that are going to make your decisions more accurate, the better.
• 24:06
  It would be interesting to see, you know, I'm sure, you know, the topology optimization as far as designing parts and that sort of thing, like it's not necessarily relevant in supercars anymore, the way that it probably is in Formula One, where they have more freedoms in part design.
• 24:16
  And I can't ever see it being the point of,
• 24:19
  can you set up my race car with these parameters based on this?
• 24:23
  I don't think it'll ever get to that point or not for a long time yet, but it'll be interesting to see where we do get...
• 24:29
  Cool.
• 24:30
  I'm really grateful that you've come on today, Tom, for this conversation.
• 24:33
  The team has had a huge couple of days.
• 24:36
  You're coming off the back of the round at Queensland Raceway.
• 24:38
  I think they're literally doing 200 rides out at the track today.
• 24:42
  You've got the day off.
• 24:42
  Thank God I'm not there.
• 24:45
  You had a massive pre-enduro test, though, yesterday.
• 24:48
  How are you going heading into Talon Bend and Bathurst?
• 24:51
  Yeah, good.
• 24:51
  Yeah, it's always good.
• 24:52
  The enduro tests are exciting.
• 24:54
  You're working with your co-driver a lot more.
• 24:56
  You're working through a lot more pit stop stuff, brake rotor changes, all that sort of thing.
• 25:00
  So it's sort of a lot more things to pack into your plan.
• 25:03
  And it comes at a time when we have one less test day this year than we have had for a number of years before, going from three to two.
• 25:10
  You really don't want to lose, you know, your tuning time.
• 25:12
  And it's interesting having it at the same track straight after an event because you always leave an event with so many questions and going, oh, if I had another practice session, what would I have done?
• 25:19
  So it gives you that luxury.
• 25:20
  But then you go, oh, we need to get this stuff done as well to prep for the enduro.
• 25:23
  And, yeah, busy, busy, busy.
• 25:25
  But, yeah, looking forward to the enduros.
• 25:27
  It's always an exciting time of year.
• 25:29
  Good stuff.
• 25:29
  Go well in South Australia.
• 25:30
  I'll see you at Bathurst.
• 25:31
  No doubt I will see you there too, Trevor.
• 25:33
  Thank you both very much for coming on today.
• 25:35
  Cheers, Rusty.
• 25:36
  All good.
• 25:37
  My thanks to Paul Reardon too and the DJR team for coordinating with Tom to come on during that very busy period that I talked about there a moment ago.
• 25:44
  Nice to get Trev on the garage too.
• 25:45
  He's been a friend and colleague for 30 years in the media.
• 25:48
  What do you think?
• 25:49
  We may not have covered every aspect of AI and machine learning in motorsport.
• 25:54
  Maybe there are other things you'd like to know about when we revisit this topic at some stage.
• 25:58
  Let me know in the comments when we promote this episode on socials.
• 26:01
  That's it for today.
• 26:02
  We'll catch you next time.
• 26:03
  Bye for now.