How Fitness Coaches Can Use AI to Save Time and Coach Better

Absolutely.

So moving down one level from conceptual to practical, and then we'll go to tactical towards the end.

So practically, what's something that you did this week using assistant coach that solved that conceptual problem?

I reworked is actually that same client.

I reworked his program.

He's very busy.

trains in a really hot climate.

He's getting ground down.

And I said, Well, we need to rework the program.

So I had a conversation with assistant coach.

And these are the constraints.

Here's his equipment.

Let's talk through a kind of a program.

and it would give me proposals and then I would respond to them.

And then we iterated on that.

I say, we, I mean, as if it's another person, but it's just person shaped, you know? Uh, but it, and I iterate and I think we came up with a pretty good program, uh, you know, much quicker than it would have been me just writing something out on paper, uh, because it would push back, you know, you want these things to push back, which leads to some tactics that you can do, uh, So if you know how an LLM works.

Double click on that one thing.

One thing you said there that I have personally experienced, not just with the assistant coach, but just working with AI in general, that I think is the differentiator between quality and slop.

You know, the AI slop is the buzzword right now.

And often why people have a bad experience with AI is that if you sit down to work with an agent and expect it to work like a Google search or some other deterministic computer-like code, or not computer-like, but traditional software, where you expect to be able to type something in and get results, an expected output that is the answer, like the answer in the back of the book, you will either generate slop or have a bad time with AI.

But if you think about it as a tool that helps you think and that you are iterating with an editing eye where you read all the output, you think about all the output and you push back and it pushes back and you have this kind of back and forth interaction with it where you're your output is evolving towards something that you can then actually use or maybe you put the final polish on it then you're going to generate high quality output significantly faster than if you were to try to sit down and do it yourself

yeah i think so i think uh so if you think about what these things do well someone said to me it's like a batting cage but this is a good way to use it so you're not facing live pitching there's not a game on the line uh you're standing in there trying to hit a fastball but you don't want it to just lot it into you want it to be you know a 90 mile an hour fastball see if you can even touch it uh and The difficulty is what makes you, the whole goal is for you to do the thing.

And so we want, I think it is a cheat code in these prompts to make it adversarial.

In other words, tell me where I'm wrong.

You know, try that.