AI vs. Coach: The Future of Programming in the Fitness Industry

Can an algorithm write a better workout than a seasoned coach? In this eye-opening episode of Strong Principles, Rob, Larry, and Zach dive deep into the world of AI programming and its surprising role in modern fitness. Rob reveals how he fed two years of his gym's training data into an AI model, and the stunning results of its ability to identify the programming styles of different coaches.

But while AI is a game-changing tool, they argue it's nowhere near ready to replace the human element. Discover why the most academically correct workout is often the most boring one, and why great programming must be "written in pencil." Whether you're a coach worried about AI taking your job or a client looking for the most effective results, this discussion on intuition, motivation, and the limits of data is essential listening.

Tune in to understand:

• How top-tier coaches are using AI as a validation tool and a "super-reference."

• The critical difference between an academically proper workout and an intuitive, effective one.

• The "human element"—why things like sleep, stress, and motivation can change a program in the moment.

• Why embracing new tools, from AI to peptides, is the only way to stay relevant and enhance your coaching.

Rob (00:00)
All right, guys. We're back to Strong Principles.

Larry (00:04)
Welcome to another episode.

Rob (00:06)
And I'm Rob Dela Cruz.

Larry (00:07)
I'm Larry Medina.

Zach (00:08)
And I'm Zack Bragg.

Zach (00:10)
All right, today, I think we're going to talk about AI, right? Ai versus a trainer.

Rob (00:15)
Yeah, a coach trainer on programming and all things AI, basically, when it comes to in the fitness industry. I've been using it quite a bit.

Zach (00:24)
Yeah, and that's surprising, right? You would think that maybe that's not something that a coach should use or would use. Let me let you talk to the benefits of that.

Rob (00:33)
Yeah. I started getting into a while ago. When I first did it, it was a couple of iterations ago, like chat, whatever, one or two, and it was not very good. And a lot of stuff was making mistakes. But the most recent ones has been a lot better. For me, what it is for programming, strength, or even an endurance work, like cardio conditioning, is that it's a big reference. Before when I would program, especially Especially when I'm doing certain training phases, when we write them up, I'm like, references from books and stuff just to make sure, not just off of memory, but hey, this is the stimulus that I really want to hit. I'm making sure I'm within certain parameters. But what it does is now it's like having everything in right in front of me.

Zach (01:20)
Now, has it surprised you in any way, shape, or form of it being wrong?

Rob (01:25)
Yes. Yeah?

Larry (01:27)
In the beginning, more so.

Rob (01:28)
Yeah, I think still So I'll get into it. I fed it two years' worth of our program, everything we did, things we say about it, personality-wise, everything. So it knows me really. The more it knows you, the better it is. I'll tell you one thing, this is going to surprise, and I don't know if I told you. So I went ahead and I fed it four weeks of our programming. Told it, I wrote two weeks, and my other two coaches wrote two weeks. Now, remember, when we write programming, it goes through certain hands and eyes. So if I write it, when he takes a look at it, we correct things that I might have missed or whatever, and then that's what it goes. It would say Zack writes it, I take a look at it, when he takes a look at it, and then we correct things. We don't change the bulk of it, but we'll still navigate because everyone forgets when you're doing it. Programming should be written in pencil, not in pen, because you want to change things, and you don't I want to be so rigid about it. I asked it, Can you pick out what workouts I did and what the other coaches?

Rob (02:38)
If it got it, boom.

Zach (02:40)
Really? Yes.

Rob (02:41)
And then I had it do another training phase. It picked it out again, boom. That's really cool. Yeah. So now that being said, I've done hundreds of workouts with it and asked it to write hundreds of workouts or to go ahead and progress a phase or something I'm doing. I never use one of its workouts.

Zach (02:59)
You never use one? No.

Rob (03:00)
How come? They're not good. Now, it's given me ideas to possibly change mine, right? But it was never good enough. Sometimes when I wrote it, I'm like, Did you pay attention to what I'm doing? Okay, I'll read it. Yeah, it's not that good. It's great. So you cannot have it dictate you. You dictate it.

Zach (03:24)
Have you used it, Zack?

Larry (03:25)
I have. I've used it mainly for conditioning stuff, but with me, mostly personal training, what it doesn't recognize, and I think like when Rob poses the question is AI better than a coach? There's certain nuances of dealing with people that it doesn't recognize. So it's going to give you a pretty simple, basic program, especially for strength work. That would need to be, you need to change a lot of things, and especially because you get a person who comes in. Maybe they slept bad the day before. Maybe they were really busy with their kids the day before. There's certain things that may affect them for that workout. Yeah. Maybe it's coming off the weekend, whatever it is that you have to communicate with the client on. And then, like Rob said, all good programs are written in pencil. So in that moment, you might have to change things. So it's a good reference tool. If we're working on a certain phase, and I think we'll do another episode talking about, like intensification. But if we're working on a certain phase, it's a good reference tool to give me some ideas for that phase. And then you end up changing a lot.

Larry (04:26)
So it doesn't even look like the same workout, maybe a couple of the same movements. But It's close, but it's not there yet, I don't think. I think it's still got some time to go. It's maybe not even close. It needs some time.

Rob (04:36)
Well, there's... It was so hilarious. When it comes to programming and coaching and training and everything else in being a trainer, you have Academically proper written workouts. An AI can do that, a very academically proper. Yeah. Nobody's going to want to do it. It's going to be boring as hell. It's going to work.

Zach (04:55)
It's going to work. When you say boring as hell, what do you mean?

Rob (04:57)
It's three sets of 10, five sets of five. You move up week one, we start at this percentage. Week two, we're at this percentage. It's not intuitive. A really good programmer is really intuitive. On not only the people they're working for, how things are going to progress, what they see, what exercise pair works really well, what exercise pair doesn't, what's going to bother the lower back, what's going to use shoulders too much, bother knees, all that. It doesn't see that.

Zach (05:29)
So As far as the AI, how did you go about learning the prompts and everything you need to do? Because I'll be honest with you, I'm a techie. Knowing that you went in and uploaded last two years' worth of programming, that's pretty intuitive of you and your team to do that, because that's not something that, I'll be honest with you, the layman person would know. Even if you're a techie, you're in a tech company, You're going to upload your code into it. And you're going to do that because you may need it to help you write more code. And so in order for it to know what the heck you're doing, you're going to upload all your data. So what made you upload your content up there?

Larry (06:20)
Didn't Joel Jamieson do that?

Rob (06:22)
Well, Joel Jamieson has some books in online, so it already had that set in. So I knew that we didn't have a ton that was out there for it to grab from. So I wanted to go ahead and feed it a lot so that it knew not our methods, but our principles that we use, and then our trains of thought of how and what we think when we're doing it, and for it to see a pattern of how we program. I knew that in the beginning because what I was seeing with it, it was given when I was going back and forth with it with the prompts. Even I was using good prompts, your PhD in exercise science and whatever you have to do prompt. And it was coming back very academically, correct, but very boring and not very intuitive.

Larry (07:09)
If you followed a program on it, if you had it build you a program, let's say we wanted to work on a certain phase and it built you that program, you'd probably see results from that program. But like Rob said, how involved is the client going to be? Are they going to want to keep coming in and doing the same thing over and over again? Because oftentimes that's what the program looks like, whereas we can change the nuances of it and keep the client more involved and make It makes it more interesting.

Zach (07:30)
And that's the human element. Yeah, that's the human element. How do you keep somebody motivated?

Larry (07:33)
Exactly. Yeah.

Rob (07:35)
What exercise are you going to change? What are you going to keep the same? Or intuitive enough to go ahead and know that this is what I want to keep. This is what I'm going to change. This is what's not going to make it boring to individuals or general audience.

Zach (07:53)
Got you. Yeah, that's interesting. Because I'll be honest with you, I never even thought about the human element of the motivation piece. And obviously, you have a class. This gym is based on a class, right?

Rob (08:06)
Half it is based on class group, and then the other half is based on personal training and small group.

Zach (08:12)
Yeah. And so how do you keep them motivated? Keep coming in?

Rob (08:15)
Yeah, that's very individualized. The personal training is very individualized. Everything is written for their gold and the specifics that they need to a certain extent. The group is written for a general, like the bell curve, like the majority of the people, that's what it's written for, the cover of that 75% of people.

Larry (08:32)
And I think that's where it's behind in both those places is that with personals, it's so individualized that it's going to write you a program that is not going to be ready for that individual. And then also in the group classes, if you Especially with conditioning, you have to run group classes a certain way. You can only run seven or eight people through something. And it gives you a conditioning program where that would work great for one person because maybe you're repeating a bike over and over again. But then it's like, all right, we have a group of 21. We got to cycle them through this workout. So we need them separated so where they can all get through the workout and they're not going to get bunched up in one spot. I think it was behind there, too. Yeah.

Rob (09:09)
And what I really like using it for, too, is it's referencing. So, hey, is this the style? And I put a specific style. Does this meet the criteria? I was like, yes, you're in it. And it'll give you a couple of suggestions of what could make it better, what not. Works really well with conditioning and cardiovascular methods, because a lot A lot of methods use a lot of intervals. For instance, there's something called aerobic power intervals, and it's a certain amount of time. A lot of intervals are the time you're doing them to the time of rest periods. And then that all changes. You either shrink the time you're doing it, so you make it more intense, and you raise the rest time, or you shrink the rest time, and you lengthen the interval. You just move that around. But you ask it like, Hey, is this meet the criteria of this specific interval? Yes, or it does, but this would make it even closer to it. And then I ask, Hey, give me a couple of different ideas.

Larry (10:09)
So it's a good validation tool, too. Yep.

Rob (10:11)
That's a good point. That's actually what I use it the most. It's a validation of like, Okay, I'm right on spot.

Zach (10:18)
Okay, that's interesting. And to the person that's anti-AI, what would you say to them?

Rob (10:24)
Well, okay. So I took an AI course on the last strength condition form that we were at, the seminars and stuff. And the way he put it, and I 100% believe it, once I started using it, I use it a lot. And it's that you can either use it and get used to it and you help it, it's helping me, or you could not use it and go up against someone that's using it. Now, what it does, it picks everybody up one level. So it takes an individual that has some experience, and it brings them up another level, and it gives them a pretty good academic understanding. It takes a really good programmer and coach, and it brings them up one more level, and it takes an expert programmer, and you're done, you're gone.

Zach (11:07)
Yeah, and that makes sense. It's very much like the tech world.

Rob (11:10)
You got all the references you ever need right in front of you.

Larry (11:13)
Right in front of you, yeah.

Zach (11:14)
And you have to spot check it, right? Because it can't be wrong.

Rob (11:17)
It can be wrong. It can be wrong. It can be wrong, but when you reference this against a book or a method that's already out there, it's going to spit back that exact. It's not going to be wrong on that.

Larry (11:32)
So it'll be academically correct a lot of the time. Correct. Yes. Yeah.

Zach (11:35)
I guess I'm saying that becauseNot intuitive. It's not intuitive yet, right? But also, I do feel very strongly that it's just like every other industry. If you are not embracing the AI piece, and you are, right? But the other person isn't, the other person that isn't using it is going to get left behind.

Larry (11:56)
They're going to fall behind.

Rob (11:57)
100%.

Larry (11:57)
It's a good tool. You have to be able to utilize.

Rob (12:00)
It's going to get rid of the half-ass programmers that are just throwing stuff out there and selling just templates and stuff. And it'll get rid of some of the trainers that are just not there.

Larry (12:16)
We've always been a fan of leaning into anything new that comes out. You get a lot of trainers who get scared of it. Oh, it's going to take away my job. And I'm sure it's like any other industry. You got peptides that come out that make people lose weight. You got some peptides that are coming out that we see it's going to improve how much easy it is to build muscle. And then trainers are like, oh, well, let's just make it seem like those are really bad for you and don't use those. But no, we can work with those. And it's the very same with AI. We can work with AI. It's not going to take the job away. We know how far behind it is right now. It's going to enhance us. It's going to enhance us. Yeah. And make it better. And you can be very clear in communicating that with your clients and with your members is like, we use these as tools and it's making us better. And a lot of people appreciate that clear communication.

Zach (12:59)
Why Why the hell would you stay away from a peptide that helps you gain muscle? Because if anything, I come in for a month and all of a sudden, Rob Zack, I am Jack, man. You know what I mean? I see the difference in a month. It makes you better. It makes you better. And honestly, it just enhances.

Rob (13:18)
Well, everything's tools. You got to learn how to use those tools. I love it. I think once I started really understanding, getting really deep into it, I'm like, Oh, I'm gone. I use it every day for referencing back and forth with But like I said, I have asked it to progress tons of training phases for me and workouts, and I've never once used it. But I've used ideas that it's given me.

Zach (13:41)
Sure. And that makes sense. I do the same thing for programming or anything else. Now, the other question I have before we wrap up is, you may be partial to a certain AI platform. Do you spot check it against other AI platforms?

Rob (13:56)
Yeah. So chat compared to Gronk. One's a little more cynical, basically, and it's not going to build you up as much. But the problem is I really worked with Chad a lot, and so it's so far in-depth with us. I'd have to do the same thing with Gronk. And the little bit I use Gronk, it's supposed to be really good, but it was so far behind. I just kept on going back to chat. Yeah, and that makes sense. And I probably should move into it a little deeper. Because I'm going to pay for the professional Because I'm going to use it nonstop, and I want to be able to put in research, and I want to download all the stuff. And then I have two bills going on at the same time.

Zach (14:37)
There's also Gemini.

Rob (14:39)
Yeah, I haven't touched that, besides using it on the phone.

Zach (14:42)
Got it. Okay. It's okay. But chat team seems to be the best right now.

Rob (14:48)
Well, once you feed it a lot, like how much you work, the more you work with it, the better it is. Because in the beginning, it was not... Even the newest version, you got to feed it a lot. But if you're willing to put the time in and you feed it, and you want it to understand your personnel and give it as much information from you as it can or your coaches and understand what's your principles and your values then. Okay.

Zach (15:10)
Yeah, I like it.

Larry (15:11)
So I think the biggest takeaway is if you're a coach, let's say you're a beginner coach, you need to continue to gain experience and gain knowledge in other ways. You don't want to become reliant on that. If you're an expert coach, you don't want to stay away from using it because it's a tool that you can use, but you should understand it needs be changed quite often. And if you're a client, then you're just typing in prompts for it to write you up a strength workout or a conditioning workout. You're still going to need to be cross-referencing that with a coach or if you have access to other forms of programming, be cross-referencing it with other things. It's not going to be the main tool, but I think everybody should be using it. It's just knowing how to use it.

Rob (15:52)
100%.

Zach (15:53)
It's like everything else, right? Yeah. All right.

Larry (15:56)
Awesome. It's been strong principles.