Many triathletes will ask ChatGPT to write them a triathlon training plan, but are these plans any good? What are the advantages relative to purchasing a plan or working with a coach?
This can feel incredibly appealing. It’s essentially a free personalised training plan, right? But how good is the work it does? And are we placing a bit too much faith in artificial intelligence to prepare for such an important goal?
How to create a ChatGPT triathlon plan
One of the big misconceptions about AI is that you simply input what you want and it spits out something perfect. The truth is that the output is only as good as the prompts the user gives it. As the old saying goes, garbage in, garbage out.
When you ask ChatGPT to write you a plan, you need to give it a LOT of information. You need to tell it how much time you have each day, what your schedule is, which equipment you have, your goals, experience, and any other relevant information. This is far from a click and collect service.
This is the first hurdle, which is that the plan is only as good as the prompts. If you don’t know much about triathlon, you won’t know the right questions to ask, or information to share. The risk here is that athletes new to the sport will get themselves into trouble by having unrealistic expectations of what is achievable, or by focusing on the wrong things.
The sycophant trap
ChatGPT is well known to be a sycophant. It will always big you up and tell you what to hear. How many times has it ever told you that you’re wrong, or need to completely re-think something from the ground up? If you ask it for help training for a sub 12 IRONMAN in six months with no experience in any of the disciplines, is it going to give you a reality check, or is it going to write you a wildly ambitious plan?
OpenAI is a business. It wants happy customers, and for people to pay for its premium services. While I’m hoping it would never outright lie to you, it is going to be overly ambitious and positive rather than realistic.
The result of this is ChatGPT triathlon training plans are often overly optimistic, with larger increases in volume and intensity than many can handle. You can of course use prompts to correct this, as long as you know enough about yourself and the sport to be able to think critically and spot issues with the plan it creates.
How much money do you save with a ChatGPT triathlon plan?
The first thing to note here is that if you are writing a ChatGPT triathlon plan, you will need the ‘plus’ subscription at a minimum. The amount of prompts you’ll need to use to create a plan reflective of your ability/experience combined with the advanced reasoning are essential. I highly recommend against creating a triathlon plan using the basic version of ChatGPT.
So this means we’re immediately looking at a £20 a month subscription. That’s not terrible, most people could stretch to that. However, when you compare it to the cost of purchasing a plan outright, you’ll probably end up spending more through ChatGPT than you would from buying a triathlon plan outright.
If you already have a ChatGPT subscription that’s a huge tick in that box. But if your primary motivation is to save money, this quickly becomes a false economy.
We also have to look at the bigger picture. Let’s say you’ve entered an IRONMAN event. The entry fee will probably have cost you north of £500. You will have purchased a wetsuit, a bike, several pair of running shoes, spent on gym/pool fees, and invested hundreds of hours of your precious time.
Yet to save spending less than £100 on a plan written by a highly experienced professional, you instead decide to spend £20 a month on a subscription to an AI service. That’s not really good economics is it?
For some athletes, my all singing, all dancing bespoke triathlon coaching can work out at less than 10% than the overall cost of someone’s triathlon budget. A training plan could cost them less than 1% of the overall investment.
If you fail to finish your event because the plan left you burned out or injured (which comes with substantial physio rehab costs) and you lose the four figures you have invested, you’re going to be kicking yourself.
This is where you need to ask yourself why you’re using ChatGPT. Is it because you want to save money, or is it because it’s a novelty and you want to have a play around with it? If you’re an experienced triathlete, well versed in AI prompts who wants to have a play around, this is very different to a new athlete putting their training into the hands of AI because they don’t want to pay for a training plan.
CharGPT is not a replacement for a triathlon coach
There is a big misconception over what triathlon coaching actually is. Many think that a coaching is all about the training plan. It’s not. A great training plan written by a professional is obviously incredibly valuable, but also over-hyped at the same time.
Real coaching is about data analysis, feedback, motivation, asking the right question at the right time, identifying fatigue, creating race and nutrition strategies, being on the end of the phone when needed, and helping the athlete self reflect. It’s not about a perfect plan which will guarantee you success.
I had a chat with a previous client yesterday who has come back to me after a few months with another coach. He went for a bike fit, and the fitter was also a coach. My client said “He sold me the world, but failed to deliver”. His goal is to quality for the world championships. The other coach gave him a spiel about how many athletes he’d helped qualify, his experience, qualifications and the like.
However, the other coach didn’t take into consideration his dodgy hip, and set him entirely inappropriate workouts. He felt the coach wasn’t listening to him, and was giving him a copy and paste programme, despite the fact he was paying a premium for a bespoke plan. When he was working with me we had several conversations a month on the phone, and I edited the plan as we went along. For what I believe was less money than he was paying the other coach.
The client has now signed back up with me because regardless of the other coach’s qualifications, experience and knowledge, he wasn’t really being coached. ChatGPT cannot replicate the experience of being coached, even if it had a world class level of triathlon knowledge, which it doesn’t.
Is it good enough for your needs?
Many people won’t want to pay me to coach them, they’re not racing at that level, so why would they? They just want a basic plan, and think ChatGPT’s triathlon plans will be good enough.
There is a chance it will be. Especially if you are taking on a short race for the first time. However, I’d like you to do me a favour. Ask ChatGPT to create an infographic around a subject you consider yourself a genuine expert in, and have a look at the first draft.
Chances are, it will be riddled with errors. The likelihood is that these errors aren’t totally glaring and obvious, it will take a trained eye to spot them. Those same errors will be present in a triathlon training plan it creates. You may not be able to spot them, but someone such as myself would.
Accountability and ease of use
When ChatGPT writes you a workout, you will then need to create it within your training software to export the workout to your watch, turbo training software or cycle computer. You’ll need to do this for every workout, which is a complete time sink.
You’ll also (naturally) be second guessing the plan a bit. You know you ain’t following the world’s greatest plan, so your adherence to it will be patchy. Will you try to tell yourself that it’s fine to skip or shorten today’s workout, as it’s likely a mistake by ChatGPT?
These may not sound like a big deal as you’re reading this, but these are the kind of things which will become a big deal as you get further into your training.
The future of AI
Many people will have you believe that refusing to use AI is like refusing to use the internet in the early 00s. Anyone who questions its brilliance will be left behind in the revolution.
However, OpenAI (the company behind ChatGPT) loses a predicted 15 million US dollars a day. The ability of generative AI to actually drive profit for companies is coming into question. The number of AI tokens generated by users each day is down. Unless OpenAI can start to make a large profit in the very near future, it is going to have to increase its prices significantly to avoid going under. The free tier of ChatGPT may be watered down, or disappear completely. Most people (including OpenAI themselves) expect the bubble to burst in 2026.
What the effect of this will be is hard to say. On the one hand it all feels too big to fail. The chances of AI disappearing completely are zero, but that’s not to say that the price won’t skyrocket, or that it won’t become watered down by the government, limiting its power.
To cut a long story, and speculation short, this current model for ChatGPT is completely unsustainable. The version of AI that you use to write your plan today may well disappear, or be prohibitively more expensive by the time race day comes around. That £20 a month you’re planning to spend could become £50 a month or more. At this point, it is no longer the cheap option. You may be able to get a plan written by a professional for less than a month of ChatGPT in the not too distant future.
Conclusion
Can ChatGPT write you a triathlon training plan? Yes, absolutely. However, it will be flawed, will involve extra work to transcribe into a format you can follow on your devices, and require a high level of knowledge of the sport, and your body, to really work.
My recommendation? Just suck it up and pay for a plan written by a human. It will be of higher quality, and may actually be cheaper in the long run. You are investing far too much time and money into this adventure to pinch pennies in the wrong places.
I have a growing selection of training plans for sale which you can browse here. Or if you have the budget, you can check out my bespoke coaching here.
