30 Days of (practical) AI – Day 2: Falmouth Road Race Training Plan

Like many others worldwide, over the pandemic I picked up running as a hobby. I started from humble beginnings barely being able to finish a mile without stopping while continuing to tack on the miles. After moving to Boston in 2021, I signed up for my first official 5K race and have not looked back since. Each year I’ve continued to push my upper distance limit by signing up for a race in the future that is just out of my comfort zone and working backwards to try and figure out how I’ll end up completing it.

Historically, figuring out a training plan would involve something like 30-45 minutes of tedious work. First, I would have to peel through different websites I found on Google offering “Moderate runner half marathon training plans”. From my experience, over half of the links would be behind a paywall and we’d be back to square one. After eventually finding a plan that suited my schedule, I would pull up my personal calendar and go through the monotonous process of creating an event for each individual training session. With 4-5 sessions a week for 6-8 weeks, this took up unnecessary time and had me wondering if there was a better process.

I’ve just recently begun my training for the Falmouth Road Race this summer and couldn’t help but wonder if GPT 4 could help.

My first prompt was as follows:

The output was a decent looking schedule, although I’m a calendar person and it would be much easier to read if I had it in my preferred format. Let’s try to see if Chat GPT can edit the presentation for us:

We are on the right track now, this information is now presented in a calendar format and is clear and easy to follow. The preferred calendar for my personal life is Google Calendar, let’s try and see if GPT can put this schedule into a format that will be easily imported into my preferred application, rather than having to manually create 40+ events:

Okay, not exactly what we wanted. The goal here is for me to do as little manual work as possible. Let’s see if GPT will structure the individual events in a file that I can then import:

Progress is being made. Now I’ll request that GPT creates the .csv file on our behalf:

This is what I like to see – GPT making baby steps towards the solution. Often it seems like a smart/lazy kid you’ll come across in undergrad, always has the solution but realllllllllllly doesn’t complete the work unless you ask him directly with instructions leaving no room for ambiguity:

As a final step, let’s get step-by-step instructions on how to import that file into Google Calendar:

Below is the end result of our import. Events in purple I manually created one by one a week ago and took around 30 minutes to do so. The training events with no color were part of the GPT-created schedule that we just put together in the span of about 10 minutes. As you can see, it will often take a bit of back and forth communication with these AI tools before arriving at the correct answer.

Chipping away at problems by leveraging these AI tools can help carve out time in our day to do things worth living – like going for a 5K run. I’m looking forward to finding another (practical) application of AI tomorrow, and until then keep on running.

Leave a Comment