Dr. Mike T. Nelson discusses how AI is shaping the coaching industry — where it falls short, where it shines, and why human coaches still matter. Plus: the Flex Diet Certification is open for one week only.
In this solo episode, Dr. Mike T. Nelson dives into the intersection of AI and coaching. Using his quest to pick up the 175-pound Inch dumbbell as an example, he shows how AI struggles with niche topics where quality training data is scarce — and why that matters for fitness advice.
Topics covered:
Flex Diet Certification — now open through Monday, June 22nd, 2026 at midnight PST. Eight interventions (protein, fats, carbs, exercise, sleep, micronutrition, fasting, and more) built on coaching leverage and metabolic flexibility. Includes direct email access to Dr. Nelson for enrolled students.
Enroll now: https://miket.me/fdc
FD Podcast - The Flex Diet Cert is OPEN and AI vs Coaching Solocast
[00:00:00] Dr Mike T Nelson: Welcome back to the Flex Diet Podcast. I'm your host, Dr. Mike T. Nelson. On this podcast, we talk about all things to increase muscle, improve performance, improve body composition, do all of it within a flexible framework without destroying your health. Today on the podcast, it's just me with a short, uh, solo cast.
We have another guest episode with my good friend, Dr. Sarah Campbell, coming up later this week, all about gut health, gut testing, and much more. But today, I wanted to talk about AI and coaching, and then also remind you that the Flex Diet Cert is now open. So if you're listening to this as of Monday, June 15th, it is open for only one week through Monday, June 22nd, 2026 at midnight Pacific Standard Time.
This is eight [00:01:00] different interventions to maximize body composition and performance. It's primarily designed for coaches to give them a system of how to do this. Right now, we've had 60% are coaches or trainers in some capacity, and 40% are actually just fitness enthusiasts who just really want to understand the material and learn more about how to do it.
[00:01:25] Speaker 2: We have eight different interventions, including proteins, fats, carbohydrates, exercise, sleep, micronutrition, fasting, and a whole lot more. The overall concept is based on the theory of metabolic flexibility meets, meets flexible dieting. This is not necessarily a counting macros program. It's primarily designed by habits using what's called coaching leverage.
So coaching leverage is the physiologic response times a client's [00:02:00] ability to change. So the example I use is protein, pretty high physiologic change. If you eat more protein, it helps with recovery, satiety, muscle growth, a whole bunch of good things. And most people, when you explain to them that they can actually eat more of a macronutrient and get leaner, they're kinda surprised and interested.
So the ability of the client to do it is pretty high, so probably a nine for both categories to give them a combined score of nine times nine is eighty-one. If we compare this to sleep, sleep has very high physiologic response, probably somewhere maybe upwards of a ten on a ten scale. However, trying to get clients to sleep more, no matter how much you optimize your environment, which you definitely should do, if you're only sleeping six hours a night, it is gonna be very hard to get enough sleep, period.
So sleep ends up being dead last. High physiologic, ten, maybe a [00:03:00] client's ability to do it is a two. So aggregate score is around twenty. So it ends up being number eight. Therefore, you would start with protein. You wanna rig the system in your client's favor, do the thing that physiologically is gonna move the needle, and is actually easier for them to implement.
So that's how it is set up, and then we've got one hour on each intervention. So I have one hour to talk all about the research and everything around protein, and one hour to talk about micronutrition, one hour to talk about sleep, et cetera. So that was a constraint on purpose so that you're not having to sit through eight hours of a lecture on protein.
As fun as I would like to do that, I wanted to get it condensed down into just the essentials, and all that is based on research. I don't have the exact number in front of me here, but I think I ended up pulling over, it was like three hundred and some research studies. Uh, not all of them made it into the, the cert, but there's literally [00:04:00] hundreds of research, uh, studies in the cert that are cited and referenced.
Uh, but we explain it to you in language that you can understand. You don't need a thesaurus or anything to get through all of the material. And then you'll learn about the big picture, and then you'll learn exactly what to do. So each intervention has five explicit action items that you'll learn how to apply in each individual case.
So it is customized to each particular client because everyone's gonna be a little bit different. You have five different options for protein that tells you exactly what to do. My pet peeve with some other certs was they may have really good information, but you're kinda left to your own devices on how to apply it.
They may have really good application, but if it wasn't covered exactly in the cert, you didn't really learn anything about the theory. So you couldn't figure out what would be the next thing to do because it's impossible to cover absolutely every single question that you would get. And if you get stuck at any [00:05:00] point, at least for now, I'm not sure how long I'll be able to keep doing it depending upon volume and schedule and everything else, uh, but for this enrollment, you can email me Once you're enrolled, if you have any questions or you're working with a client or yourself and you get stuck on one of the modules, there's a concept you didn't understand, you can literally get my private email and just email me, and I will get back to you and help you out, which I know is different than a lot of certs.
So a lot of certs, they make it very difficult to email the person who did the material, or if you can, you have to jump through 18 other hoops in order to do it. Uh, once you enroll, you literally get my private email. It comes to my inbox, and I will answer you probably within 24 to 48 hours, you know, unless it's a weekend or something like that.
So that is the Flex Diet cert, which is open until Monday June 22nd at midnight Pacific Standard Time, and we'll have a link down below where you can get all the information. If you are gonna be [00:06:00] at the International Society of Sports Nutrition Conference, I'll be there this week. This is recorded on Monday.
I fly out tomorrow, Tuesday. I'll be giving a talk there. I think my talk is on Thursday, so I'll be there the whole time. So if you're listening to this please find me, come up and say hi. Would love to chat with you. I'm super excited for all the presentations and seeing old friends again, learning all about research, kinda what they're up to and what's going on, and meeting new people.
Which brings me to the topic of the day, which is, how is AI going to affect coaching? If you played around with AI, it, in my opinion, is both extremely fascinating and also can be very maddening at the same time. The example I gave the other day was on Instagram. So one of my goals is to pick up the hundred and seventy-five pound Inch dumbbell.
So if you've ever seen the kinda [00:07:00] old-school-looking solid cast, uh, dumbbells, that's what it is. Except for the handle is two and a half inches thick, so the handle's about the diameter of a pop can. So one-handed deadlift of a hundred and seventy-five pounds for people who lift isn't really that heavy.
Um, but because of the handle's shape and because of the size and it's all one cast material, it's opposing the rotation and being able to hold it in that position that makes it extremely difficult. So for a while, uh, the original guy was a guy named Thomas Inch. So he did vaudeville acts many, many years ago, and this was promoted as the unliftable dumbbell.
He would have people, and even some strongmen, come out of the audience and try to pick it up, and pretty much no one was able to do it. Even now, if you attend, uh, sometimes the different booths, such as Sorinex, uh, will do this very often at the Arnold or other events. They'll have one there, and vast majority of people are not able to pick it [00:08:00] up.
So it is a very difficult task, and every once in a while, I'll put into AI Hey, how should I train for this?" And it'll give me some crazy answer, which isn't really close at all. Uh, although it has been getting better over time. Uh, the other day, I did this, and instead of saying, uh, six to eight weeks it would take me to pick it up, it did say six to eight months.
So it is a better estimation And then I would try to put in more specifics of hey, here's my current lift. Like I have a replica of this dumbbell. It's a hundred and thirty-five pounds that on a good day I can pick up for a single. My actual double overhand deadlift as of this recording is two hundred and ninety pounds.
And I put in specific lifts that, you know, should have some transfer to it. It didn't really seem to change the output at all. And I was kind of scratching my head and then I realized that if we think back to how AI is getting information, and this applies to all information, even if your goal is not to [00:09:00] pick up heavy-ass dumbbells, it's scraping the internet and scraping forums, it's scraping published stuff, it's scraping all sorts of things.
And I was talking to my coach, Adam Glass, about this, and he made me realize that, oh yeah, because this is such a weird niche type of lift, there isn't much that's been published on the web about it. And a lot of the people who are able to do it, they might have it in a paid product, which may or may not be scraped.
Um, but a lot of the information about it, there just isn't a lot. And there's some early forums of a bunch of people talking about it. And a lot of that information from the early forum was just really bad and was from people who hadn't really done it or weren't even practicing doing it. And so the information that the AI was basing it on was primarily information that wasn't very accurate.
And this goes for all sorts of stuff. Like up until-- I haven't ran this check [00:10:00] either, but looking at protein and kidney damage Uh, last time I ran this was, oh, six months ago in a couple different models. It said that, yes, protein will damage your kidneys and you should avoid it. Now, again, that may change.
You know, the models are always being updated, and they're trying to make them more accurate. But the point of this is that there still is a fair amount of information that is not accurate on AI. So if you are starting with information that is inaccurate, it is definitely not gonna help you get to your goals.
And the way AI is phrasing things, it makes it sound like it is accurate. And even in my case, when I directly challenged it and put in specific lifts and how long it took me to accomplishment and what my limiter and everything was, didn't really change it much at all. So I think one issue is the potential, the accuracy of information, which, you know, has gotten better over time.[00:11:00]
Number two is there will always be people who are not necessarily looking for coaching, just want information, and I think that's fine. That's why there's articles, and I have a newsletter and put out stuff on this podcast, and you can get tons of free information. Now, again, that may not be specific to your situation and your goals.
Again, I think it can be useful, but there will always be people just searching for information. And for that, if you are relatively careful, I think AI can be useful A third thing is if you have larger data sets, I think AI can be useful for some of that. So I've played with, you know, de-identifying people and putting in three-year trends of heart rate variability.
And in some cases, it actually came out with some pretty interesting stuff things that if I were to do by hand in an old school spreadsheet would take me substantially longer. So I do think there [00:12:00] is some benefit to processing existing information. But if you are a coach or trainer, if people are just looking to get information, I would argue they're probably not gonna be your client 'cause working with a client is working with another human, and part of that, I think, is also knowledge and accountability, and the ability to use your knowledge and experience to get them a better result and a faster result.
I think that's where AI is going to struggle for a long time. And I'm also super curious to see if people are even gonna be compliant to a robot. My gut feeling is probably not. There's definitely gonna be programs that will, will try it, but I think at the end of the day, there's still gonna be a market for paying another coach paying them money, so you have skin in the game, to have them customize [00:13:00] your training, your nutrition, your recovery, maybe your stress, lifestyle, sleep, blood work, et cetera.
I think everything is moving more towards a complete solution, looking you as a human organism from multiple different facets. I think the days of just writing a training program in nutrition are coming to an end. I think you definitely have to do that. You definitely have to be very good at doing that.
But if that's the only thing you're good at, I'd say that's just kinda the, the baseline entry point right now uh, always curious what your thoughts are. I think AI is interesting. I do think it will replace some of the stuff we're doing in the coaching industry. I think most coaches, or a lot of them, are trying to optimize air quote AI for all the wrong things.
If you only get three inquiries a week about your coaching, like I don't know why you would have them talk to a robot. [00:14:00] Maybe actually, shocker, be a real human and reply back to them. That might be useful. I get it if you're getting thousands of replies a day or something like that. But, you know, use AI for the things it can be useful for, which right now I think is maybe some backend tasks, scheduling, maybe some production of things you've already written in terms of publishing it on a website or taking the end product and putting it into a newsletter, but you're still the one creating it.
Maybe it'll get to the point where it can do video editing, things of that nature. Data analysis that's more in, in-depth that may take you a while to do. Organizing data. So one thing I played with AI is because looking at blood work, depending upon the labs, you will get them in different orders and different units.
So I did create a little AI program to put them all in one unified [00:15:00] output and making sure the units are correct and then comparing that to the values that I'm using for performance. Um, that saves me a ton of time because in the past, I would have to take it and rekey it in and sometimes change the units if it's from overseas or it's from the US and then put that into the right category and re-put that back into a spreadsheet so that everything is all nicely and neatly organized.
And now it literally takes seconds to do that, which is amazing I think AI can be definitely useful. Um, it'll be interesting to see if people are gonna pay AI as a coach. Definitely think there'll be some lower-end products that will do that, but I don't really think it's gonna change higher-end coaching as of now.
I think people are still gonna wanna know that there's another actual human on the other end of the spectrum that cares about them and is working to get them a better result, [00:16:00] and part of that is gonna be information, part of that is gonna be problem-solving, and then just also accountability by having skin in the game.
So curious what your thoughts are on that, and then if you wanna get better at your nutrition and recovery components, the Flex Diet Cert is now open. It'll be open through this coming Monday, February 22nd, 2026, at midnight Pacific Standard Time. We'll have a link down below. If you have any questions, you can reach out to me through the website or on Instagram.
It's just @drmiketnelson. Website is miketnelson.com. You can hop onto the newsletter for free. Got a ton of information on the Flex Diet Cert going out this week over the newsletter. It's free to join. And as always, thank you so much for listening. I hope to see you in the Flex Diet Cert if you're not already enrolled.
And then if you are at the International Society of Sports Nutrition [00:17:00] this coming week, please come up and say hi. I'm doing a talk on wearable technology on Thursday. I don't remember exactly what time. I think it's in the afternoon and would love to see you there. So thank you so much for listening.
Really appreciate it. Talk to all of you next week.