Copilot coaching: Can AI train me to run the London Marathon better?
Ed Warner has swapped relentless pavement-pounding for AI prompts in the hope of training smarter – and less – for Sunday’s London Marathon.
I’m getting my excuses in early. What began as my AI-generated, pared-back London Marathon training programme has – due to entirely unforeseeable circumstances – become the skimpiest of prep for this weekend’s race.
And yet Copilot will insist on channelling relentless positivity my way. Not sure about following the trend du jour and falling in love with a chatbot; mine is heading for an early divorce.
Einstein may (or may not) have defined insanity as doing the same thing over and over again while expecting a different result, but up to now I’ve been content to repeat the same marathon training programme each year and to embrace roughly the same outcome. After all, I’m one year older each time, so the same finish time surely means a better achievement.
This year, though, tired of the drudgery of the weekend long slow run, I scratched an itch that has nagged away ever since my first marathon.
Back in 2002 I read of a runner of a certain age (very close to mine now) whose training programme consisted of racing as often as possible but whose every other run was simply a flat out five miles. It flew in the face of all conventional marathoner wisdom but, lured by the thought of lazy Sunday mornings, I turned to AI to construct just such a plan for me.
Like many mothers, mine was fond of asking whether I’d stick my head in an oven or jump off a cliff if my friends told me to. All with the intention of drumming some common sense into my young self.
Goodness knows what she would have made of AI’s propensity to encourage the most ridiculous schemes and ambitions, including the belief that muscle memory and will power are valid substitutes for hours of pavement pounding in the hope of a decent marathon experience.
Having tapped my past marathons, upcoming race entries, and recent finishing times into Copilot, I asked it whether my minimalist approach to training would work. The reply: “This is not a conventional plan – but it can work for someone with your background, discipline and tolerance for quality work.”
Background? Discipline? Tolerance for quality work? Who was it and/or I kidding? Nevertheless, I seized on the 16-week plan that it spewed out, wrapped as it was in further cheesy encouragements.
The trouble is that while I’ve run all the shorter races I intended – including last weekend my fastest 5k since 2009 – my midweek running has been close to negligible. Just to give you a yardstick, one close (albeit much younger) friend’s mileage has been three times mine in his training. So I’ve given my Copilot coach an easy out, having barely tested its plan.
With just a few days to go, and so few miles on my clock, I returned to my virtual coach and asked it whether I should even travel to the start line this weekend. To be fair, it did scrape together the obvious caveats and health warnings from the sum of digital knowledge and, in short, warned me of a grim final few miles. “The final 10k will be a grind”, to be precise. No kidding!
Whether I’m prepared to aim for a one hour slower time than in recent years, as per Copilot’s warning, is another matter, but its advice to “treat it as a celebratory run, not a race” is clearly wise.
I wonder just how bad the data I inputted would have to be before my AI coach simply told me not to bother. Instead it has given me a pacing strategy that means: “You finish because you didn’t burn the matchbook early.”
I think I know what that means. Not sure Copilot really gets my bloody-mindedness though, or whether the two of us have done enough to warrant maintaining our relationship for the long term.
I’m not convinced either that the relentless “you got this” positivity that pervades all sport today, from elite to grassroots, is necessarily a good thing. This general mindset is reinforced by AI’s tendency to tell you what you want to hear – assuming of course that you’re a glass-half-full person.
Does this encourage people to have a crack at things that they might otherwise swerve or believe beyond their reach, or does it dupe them into grossly underestimating just what it takes to succeed, whatever one’s level of ambition?
Me, I’ll start on Sunday and see how I feel. I probably will indeed find those final 10km grim. I’ll be wearing my purple Arunners club shirt, bib number 45032. Give me a shout if you see me.
I’m not running for charity, but please do sponsor anyone you know who is, especially if they are taking part in their first marathon and have really done the training miles. Or make a donation to the cancer charity of your choice. Hopefully I’ll be in a fit enough state to report back next week.
Robot love
You’ve probably seen the video clips of the robots running (and some falling) in the Beijing Half Marathon. If not, check out this BBC report here.
It may not yet mimic human biomechanics very accurately, but the winning humanoid was certainly swift. If you’d like my take on the impending robot invasion of sport and the commercial potential of man v machine, read this.
Living in the past
If retro rather than futuristic running is your thing, I can highly recommend the Seasoned Strides podcast. The latest edition is a London Marathon Special, including legend of the race and last British male winner Eamonn Martin. Listen here.
Making your mind up
Two and a half years’ work to modernise the governance structure for rugby in England and it all comes down to this: a Special General Meeting of the RFU tomorrow at which member clubs are invited to approve a set of reform proposals.
I’m told, in classic election night fashion, that it’s too close to call whether the requisite 66 per cent of those voting will tick the necessary “yes” boxes.
Regular readers will know that I’m more than an interested observer here. Rather, I’ve been a disinterested participant in the review – one of the steering panel’s independent members – that has resulted in the proposals.
It’s fair to say that emotions have run high in some quarters throughout the process at the thought of possible altered roles and shifted responsibilities. My hope is that all interested parties approach the SGM with suitably dispassionate perspectives and vote through changes which the recently appointed chair of the RFU, James Wates, correctly describes as “overdue”.
“My very strong and clear view is that standing still is not a sustainable option.”
James Wates
If the reforms fail to garner the necessary backing tomorrow, English rugby will yet again have passed up the opportunity to craft a democratic and effective structure for the RFU capable of giving the sport the best chance of thriving in this increasingly chaotic and competitive world.
Ed Warner is chair of GB Wheelchair Rugby and writes his sport column at sportinc.substack.com