It has also prompted many industries to evaluate the impact that advancement in AI technology will have on them in the next few years – both in terms of how it can be successfully exploited, and where it could also replace humans.
Formula 1 doesn’t escape this, with teams already using AI across a wide range of its activities – including car setups, development directions and resource deployment.
It’s playing a part in race strategy planning too, and that has opened some debate about whether or not the robots can take over the pitwall entirely.
For after a 2022 campaign that proved once again how critical strategy is to victory in F1, and how high the penalty can be when the pitwall gets it wrong, it is clear what the attraction could be for a successful AI model.
After all, it shouldn’t crack under pressure. It should in theory deliver the right answers based on evaluating a data set far wider than a human could ever get through, and it would never be worried about what the press writes about it the next day.
But what sounds simple in practice is much harder to deliver in the real world.
The opportunities and challenges are something that McLaren has come to understand quite well through recent developments to improve its own strategy and software.
Interestingly, it is work that the Woking-based squad has been undertaking with technical partner Splunk for its Esports team that has enabled it to begin experimenting a bit more with new strategy ideas – and the potential use of AI.
Since joining up with McLaren at the start of 2020, Splunk has helped the team through delivering software that is able to search, monitor and analyse machine-generated data in various aspects of the racing operation.
One area where it has excelled is in its simulation systems, and especially delivering live tools – like the famous race trace – that are now essential for understanding the rhythm of a grand prix and plotting the best pit calls to win.
Last year, as its McLaren Shadow Esports team ramped up its effort to win the F1 Esports Series, Splunk adapted a version of its F1 strategy tools to help.
And it was that ability to have ready access to the kind of data and software that the team gets on the pitwall during F1 races – like tyre degradation analysis, and undercut/overcut predictions – that proved to be critical in its ambitions.
Lucas Blakeley duly went on to win the championship for the first time, and was not ashamed to say that the Splunk strategy software was so important to his success.
“That extra level of depth that we had thanks to Splunk, and I’ll keep saying it, is just one of the coolest things that we've got to use,” he said.
“It just gave us that extra perspective and extra layer. An extra arrow to your bow if you want.”
From McLaren’s perspective, the value was clear. As Ed Green, head of commercial technology at McLaren Racing, explains: “It was an absolute game changer.
“I think it was pretty instrumental in getting the number of wins we did, leading up to the championship.”
As the Splunk system became a core part of the Esports team, it opened up avenues for some experiments in trying new things, which would not be possible in the environment of a normal grand prix weekend.
James Hodge, GVP & Chief Strategy Advisor, Splunk, said: “There's more you can do with Esports, in terms of rapid development.
“The stakes are very different. If you touch anything in real F1, you are touching an almost mission critical system. If an F1 car isn't receiving telemetry back into the McLaren garage, they can't turn the engine on.
“On the Esports side, it is less of an implication: if you don't get the telemetry and that doesn't work, you know Lucas can still go and drive.
“Plus there's less complexity. You're not moving an IT rig around 20-something different locations around the world.
“So it allows you to try to rapid prototype things. We've been able to try things on the gaming side that we might have taken a year to get into production on the real side, just because you've got a big team that needs to change the way of working to adapt to that new dashboard analytics.”
Green adds: “What Splunk is allowing us to do is kind of iterate really, really fast, without needing a roomful of super expertise.
“And I think the speed at which we're developing with Splunk, some of that is feeding back into the F1 team.
“They're seeing a little bit of what we're doing and going, ‘okay, that's pretty quick.’”
The ability to experiment so much on the Esports side, with car performance parameters so similar to the real world, has inevitably opened the door to see what role AI can have in strategy calls.
Green cites the AlphaGo computer programme, developed by Google subsidiary DeepMind, that was able to beat a human at the game Go, as an example of where the right kind of AI can outsmart the human brain.
“It’s interesting what happened in Go,” he said. “How far can we push things like that in to sport?
“It could provide some real inspiration and, on a personal level, I think I'd love to see an AI-led strategy one day.”
For now, Splunk thinks that the technology is only at the stage where it can offer help in the decision-making process, rather than making the final call itself.
“What we're doing is looking at what is the probability someone's going to pit based on their performance,” added Hodge.
“You can start to see maybe a degradation in lap times, so this is probably going to be their pit window.
“And that's where we're at: we're not quite in the ‘run me a race and predict everything from it.' But we are definitely helping with the decision-tree making.”
Green agrees that letting AI run the show is something that is not realistic right now.
“Are we there today? No. Do I think we'd ever have AI making the decision of when we get in the pit? Well, there are times when you can see it’s time to stop, as there's that dotted yellow line [on the race trace] that shows us the pit window.
“But there's so many parameters. You've got 20 drivers on the grid, you've got all the different variations, you've got people's driving styles, and people don't execute and behave as you sometimes expect.
“Drivers are very good, but they're going to change their lines, there's a competition element to it. So I think we've got to understand more about all the parameters we can collect, to really, really understand it.
“And, if AI is telling you, you're winning the race but four laps before the end we intend to change tyres, would you trust it enough to follow it? Who knows? But I don’t think we're there yet.”
There is also another critical factor at play here: that for entertainment purposes, F1 should remain more a sport that an all-out technical exercise.
It’s why drivers must drive the car ‘alone and unaided’, and the use of automatic systems to help them has been consistently clamped down on.
Perhaps such restrictions need to be implemented for pitwall decisions too, because part of F1’s team work attraction is that sometimes the humans get it wrong – and that helps make things unpredictable.
As Hodge said: “I play racing games. I'm not very good at it, but I'm happy.
“If there is an AI that I'm racing against, you never quite feel satisfied. It doesn't have that drama: I've beaten a computer.
“I find it much more fun to play races against 19 other people I've never met, because it's got that human element. It's got a sport behind it.
“That’s why I don't think we'll ever get it to the full AI. It's one of the reasons in F1 why you still need Lando [Norris] or Oscar [Piastri] to press buttons.
“You still need the sporting elements and some of the skill to actually bring the drama, theatrics or heroes and villains to it.”
And what does ChatGPT think? This writer asked it if it could write a code for an F1 race strategy plan?
It's response: "Unfortunately, writing a code for a race strategy in Formula 1 would be a complex task as it involves various factors such as weather conditions, track conditions, tyre choices, performance of the car, competition, etc."
That sounds like a no.
