The Biggest AI Mistake People Make With Fat Loss Over 40 | Ep 481

You have access to more fitness information than any previous generation. ChatGPT, Google, YouTube, social media, and of course this podcast can answer almost any question about losing fat or building muscle in seconds. So why are so few people over 40 actually getting better results?

The problem was never a lack of information. Most people are using AI to solve the wrong problem entirely. The generic plans you ask for rarely change your body, and this is explained in the behavior research. 

Knowing what to do is not the same as doing it... adherence has always mattered more than the plan you choose.

AI becomes useful when you treat it like a coach instead of a search engine. Feed it your own training logs, nutrition data, and biofeedback, and it helps you make better decisions instead of collecting more advice.

Later in the episode, I also share a simple prompt that gets honest feedback from any AI tool instead of agreement.

By the end you will know which jobs to hand AI, which to keep for yourself, and how to use it for execution instead of just more information.

Try Fitness Lab, the AI coaching app that reads your data, coaches you on your patterns, and keeps you consistent for adults over 40 who want to lose fat and build strength. Summer special is 20% off through July 3:
https://witsandweights.com/app

Wits & Weights is the evidence-based podcast for strength training over 40, body recomposition, fat loss over 40, metabolism recovery, and healthy aging. Hosted by Philip Pape, creator of Eat More Lift Heavy and Fitness Lab.

Timestamps:

0:00 - Why more AI isn't giving you better results
4:22 - The real bottleneck is behavior, not information
8:13 - The gap between intending and doing
11:06 - Where AI actually helps you change
12:42 - Spotting patterns in your own data
14:30 - Accountability, stakes, and the human factor
19:58 - Coaching that works with your data
23:40 - Bad AI prompts vs. better coaching questions
32:36 - The one prompt that makes AI stop being a yes-man
35:20 - The real reason most people stay stuck

• Transcript

  Why more AI isn't giving you better results

  Philip Pape 0:00

  You now have access to more fitness information than any human in history. Every answer about how to lose fat, how to build muscle, absolutely free just in seconds using AI. But ask yourself how that's working for you. You're probably not more consistent than you were a year ago, two years ago. You're not any leaner. And today I want to show you why AI, as you're probably using it, is not really much of an upgrade over a Google search. It might even make things worse. How to actually use AI so it's helpful and effective for you, and a few specific examples of the right and wrong ways to put it to work for your training and nutrition. Welcome to Wits and Weights, the show that puts a popular piece of fitness advice under the microscope, finds the hidden reason it doesn't work, and gives you the deceptively simple fix that does. I'm your host, Philip Abe, and this is an episode I've wanted to do for a while. I think the timing is perfect because everybody and their mother is now using AI for their health and fitness, from Chat GPT to Claude to just random Google searches and everything in between. There's Buco tools and software being generated that use AI. You're asking it, how do I eat? How do I train? Is my plan any good? And I think it's really, I think it's awesome that we have technology to play with like this. And depending on who you are, if you're over 40, if you're navigating perimenopause or menopause, if you're lifting weights, you're trying to lose fat, trying to hold on to muscle, you've probably used AI by this point and asked it a question like, hey, what is the best workout for me given XYZ? Or hey, build me a meal plan given my macros. Am I am I right? Am I right? And if not, you're probably like, oh, I can do that. Let me go try it out. But I bet most of you have done that. And I want to be totally careful and walk the line on this episode because I am definitely not anti-AI. Not only do I use it all the time in my business as a power assist to save tons of time, especially on tedious type of work, but I built an AI coaching app. So I use these tools every single day. It this is not the, you know, robots are coming. This is Skynet Terminator taking over the world episode. And it's also not, hey, I, AI is the best and most amazing thing that will solve everything. Oh, and the last thing, it's not. I'm not gonna get into like how much water it sucks up and how expensive microchips are going to be and how the billionaires are gonna take over the planet. I'm not getting into any of that stuff. I'm gonna try to keep the focus narrow. I think AI is a useful tool. That's it. Like any other tool, it's tool. It's probably one of the most incredibly game-changing advances that we've seen in such a short period of time. And that can be highly dangerous in many of our minds, of course, when there's not appropriate, you know, boundaries on it. We saw this recently when Claude had to walk back the release of their latest mythos class model. You don't have to know what I'm talking about, but basically they said, yeah, this is too dangerous. Like people can create bioweapons or whatnot and you know, hack into banks and if they use it. So we're gonna turn it off. So I totally get all of that. Like I'm pretty educated in what's going on there. I love tech. I'm into all of that. But what this episode today is is about the gap between what AI can do and what most people use it for and how they use it. And we're gonna kind of explore that gap so you can take advantage of the tool for yourself because ultimately I want you to be able to use the tool in an educated way to help you out. That's it. I want you to also stick around to the end of the episode because I'm going to give you a prompt. It's a one-line prompt that you can paste into any AI chatbot that actually helps it help you better by pressure testing your plant rather than just totally agreeing with you. A concept called sycophancy, AI sycophancy, which, you know, South Park made a really funny episode about that a while back. You should look that up. But a lot of people, a lot of people aren't aware that they can they can train or they can update settings in their AI to help it, help it do better for you. And it just takes a few seconds to set that up. So stick around for that. But here's where we're going today. I'm gonna show you the bottleneck that has nothing to do with information. I'm gonna show you four things that AI is genuinely good at that can move the needle for you. And I'm gonna show you the difference between using it well and using it more like a search

  The real bottleneck is behavior, not information

  Philip Pape 4:22

  engine, like a lot of people use it. Okay. So hopefully this is a great episode for you. And as always, I love to start with hey, what is the conventional wisdom? What do people talk about? What do people claim that is true and maybe it's not? And the information's been out there when it comes to health and fitness and all of this stuff. It's been out there for over a decade. I mean, I got into podcasting in my own personal journey because I was able to get access to great podcasts and books and videos. And this was all before AI. And if you think about it, everything you need to know, to know, okay, information to lose fat, to get strong, to do all the things we talk about, to improve your metabolic health, has been sitting on the internet for free, even since before some of you started your fitness journey, even going back to like the bodybuilding forums and going back to the early days of the internet. How to set up protein and get enough, how to structure your lifting, progressive overload, energy balance, like all that stuff. And if anything, we have more so much information now that it's confusing, right? Like it's who who's telling the truth. We have influencers, we have social media. None of it is secret, I'll say. And if it is purported to be a secret, they're probably trying to sell you something. So, so none of it is gated behind a paywall. You can find all the studies, you can find all the information. You could have learned every relevant principle even 10 years ago without spending a dime, even though the science continues to get refined, is pretty darn close to what you would need to do. And of course, we know this because people have been successful improving their health and fitness for much longer than we've had the internet. And of course, we've now gotten just better and better at it. So the question that should really bother you here is okay, if information is the thing standing between me and what I want, the body I want to be lean, to be strong, to be healthy, to stop binge eating and to get rid of food noise, whatever it is, you'd already have it, right? You'd already have it. You you'd be done. Like you would have solved the problem, you'd have your plan, you'd execute, you'd be good. Everybody who has a smartphone, which is like the entire world, would have what they want. We'd be lean and strong. They'd they'd be millionaires, billionaires, whatever. Right. And you, I know you've heard this argument plenty of times, right? Everybody has access to the same information today. Information is as cheap as it's ever been. It's more democratized than it's ever been. You know, started with the printing press and it just went from there. And now we have the internet and now we have AI. But that doesn't matter, apparently, because many of us are still struggling. That's why you're listening to this podcast, because you're trying to figure out, okay, what is the secret or what is the thing I need to do? So information was never the bottleneck. And that's important because AI, the way most people use it, ChatGPT and Cloud and all of these, is a machine, a highly intelligent, you know, neural network machine learning trained, you know, whatever tech behind it machine for producing more information. Now you might not like that, that's the case, because that's really what it's doing. It's taking all of the knowledge that we have in human existence that it has access to and it's processing through that using very intelligent algorithms that are self-learning in many cases. And I know that scares a lot of us, but it's faster, it's, I'll say, cleaner, it uses better grammar, it's personalized to your exact question. I mean, now when you go to Google and you search, Google's smart about it. They use Gemini to just to give you the answer at the top. In fact, a lot of a lot of businesses don't like it because now there's sponsored results or their links at the top, people don't go to them anymore. They just get the answer right from Google. But it's still more information. And it's still, it's like Wikipedia, it's kind of like what Wikipedia has done in a way, where you guys, you know this. You go to Wikipedia and you search up something and then you can follow through with all the the references. So if information was never the problem, then more information, even if it's better delivered, is not the solution, is it? I mean, you tell me. It has it solved things for you? Or is it a marginal or maybe even large upgrade on something that was never really the problem?

  The gap between intending and doing

  Philip Pape 8:13

  So I want to talk about the actual problem. Okay. There's a researcher named Pascal Sheeran who did a big review of health behavior studies. And across studies on exercise, on screening, on all kinds of health behaviors, he found that of the people who had a genuine positive intention to do the thing, to do the thing they wanted to do, they wanted to do it passionately. They planned to do it, whatever that thing is, the median proportion who then did not follow through was 47%. 47%. So almost half of the people who intend to act don't. And so I want you to sit with that because there's a lot of judgment today about, well, you just gotta do it, right? You're just, I don't even want to say people get accused of being lazy, but we think of obesity and things like that. Even if behavior is a part of the equation, okay, okay, let's say it is. Are you able to follow through on that behavior? Right. And and as soon as you judge somebody else, like put point the finger at yourself and think of anything that you failed to follow through on in your entire life. There's probably millions of things. Okay. The constraint was never knowing what to do. It's that vast canyon between intending to do it and actually doing it. Not only that, but consistently, especially on the days you don't feel like it, because that's how you get consistent. The days you feel like it, you're gonna do it. But there are probably more days you don't feel like it than there are that you do. So knowing and doing are two completely different problems. And AI, especially when it's used as a fact retrieval machine, only really solves that first problem, the knowing part. And that's not really the problem that you have. Does that make sense? So, and okay, by the way, we have good data on the doing side as well, like actually following through. There's a famous diet study by Danzinger, where they put people on four completely different popular diets: Atkins, Ornish, Zone, and Weight Watchers. Now we're going back in the time machine, aren't we? Okay, let's see. I've tried two of those, Atkins and Zone. And so they're completely different diets, different mechanisms, different macros, whatever. And what they found is no meaningful difference between the diets. What predicted whether someone lost weight was not the diet they were on, it was whether they stuck to the diet. Adherence. Okay. Broken record, guys, if you ever, if you've already been following my podcast, you know this is a tenet of what we talk about is adherence. That is the key. It's not the mechanism, it's not the plan, it's the doing, the doing, the doing. Okay, do it, just do it. Okay. So if you've been treating AI, I don't know why I'm in a good mood today, but if you've been treating AI as the thing that finally gives you the answer, I want to gently suggest to you that you're solving a problem you don't even have and you're ignoring the problem that you do have. So we're gonna get into that now. All right, so if information is not the bottleneck, behavior change is the bottleneck. Let's

  Where AI actually helps you change

  Philip Pape 11:06

  talk about why this matters for how you use AI, because this is where I think it gets interesting. AI is really, really good at some of the behavior change stuff, but it's not what a lot of people use it for. All right, and I'm gonna piss off a lot of coaches, a lot of behavior change specialists who think you have to have a human in the mix all the time to help people change their behavior. That's absolutely not true. Okay, well, follow, follow me on the logic. There's a study out of out of a team at Google. Now take that with a grain of salt, a team at Google where they took a large language model, so that would be like a chat GPT type software, and they they directed it at activity coaching. So activity coaching, coaching for activities. And they tested on things like is the advice actionable? Is it realistic? Is it empathetic? Does it motivate? And what's so telling is that the model itself got significantly better at being actionable and realistic when they set it up well. Okay. In fact, big jumps in those two areas, but on empathy and motivation, the gains were not even statistically significant. So the machine got better at the practical translation stuff, and it stayed pretty weak as a model or as a tool at the human stuff. And so from that point, we can say, okay, so there are things that need to be kept with a human. Like I just said, empathy and motivation. There are now, having said that, there's a way to marry the two and kind of amplify the two. We're gonna get into that, but first we need to understand the AI side of the equation.

  Spotting patterns in your own data

  Philip Pape 12:42

  So what's it what's AI good at? I think there are four things. Four things. And a lot of you aren't even using it for any of these things. Number one, and I think this is the big one, is it can recognize patterns in data. Makes sense, right? It can take a lot of the more data you give it, the more it can act on it. It's really powerful for that. So instead of when we think of food, not what should I eat, it's here's what I actually ate. So here's my food log, macrofactor, whatever. And I've given it to you for the last three weeks of data. What patterns do you see in my eating to help me decide what to eat? Now that's a very different question, isn't it? Because now it is working with your reality instead of giving you a generic answer. So it's using data, which makes a lot of sense. And this can be used for a lot of things, like your lifting, right? You can put in your workout logs, you can put in your biofeedback data, you could you could do everything that I have clients do as a coach, everything that they track, body measurements, weight, like you name it, you can give it that data. In many cases, these tools are now integrated with Apple Health and whatnot, and it makes it even easier to do that. And a lot of tools are being developed to take advantage of that as well, right? Same thing with something very specific like your protein intake and looking at the patterns and saying, oh my goodness, it looks like my protein falls off here and here, and here's why. Because if you're just tracking it in a food log, like let's say Macrofactory, you can kind of see that visually and figure it out on your own. But what if there are things you can't really see as easily? And I think the AI is great for that. It can catch it in you know milliseconds or seconds. So finding a signal out of the noise, as we say, in a pile of data. And the more, and the bigger the bigger the pile of data, the better. So that's what's great is you can feed it so much data, lab work. I mean, just dump it in there. All right, number two is accountability

  Accountability, stakes, and the human factor

  Philip Pape 14:30

  and a check-in structure. Now you might say, well, isn't that the human part? Well, no, again, we're talking about activities. We're not talking about like empathy or motivation per se, although I'm going to argue that a lot of our motivation is intrinsic motivation that does come from checking in and accountability. So let me get a little more precise on this part, on the accountability part, because it there's nuance here. AI can scaffold a check-in, let's say. Let's use the word scaffold, right? Like a scaffolding around a building. It can ask you the right questions on the right cadence. What it cannot do, it doesn't care whether you lied to it, it doesn't know. There are no stakes, there's no aspect of like empathy to it. It's just cold calculated. And it's like, if this is the check-in you want to do it. And let me let me elaborate on this because there's there's a a data set in the research where they had 65,000 people using a coaching app and they compare people who had AI coaching alone versus people who had AI plus a human coach. The group with a human with the AI lost over 74% more weight than the group that just had the AI coaching. And it was the same app. The only difference there was a human in the loop. And the mechanism here is that the people with a human coach, they actually logged their food and their weight more often. And it's not like the human had secret knowledge, the human just created stakes that made them actually do the thing. So that's that's what's kind of interesting. That's a motivation piece. That's pretty cool, right? That's why humans are gonna be important for a long time, hopefully for the rest of eternity. But I wanted to talk about that difference, right? So, in other words, AI can be incredibly useful, humans are incredibly useful, and men put them together and you have a very powerful combination. Number three is a customized application of a general principle to your specific situation. Anyone take physics, right? Or, you know, you've heard of Einstein, where we talk about general relativity versus special relativity. So think about that. All the principles you read about or hear about on podcasts or this show, they're general principles. At least I hope they're talking to you in terms of principles and not telling you this is the answer, the one answer. Because there is no one answer. There are principles, and then you have your specific situation. So, like the principle of eating around 0.7 to one gram of protein per pound of body weight, everybody knows that. But what's your body weight? What's your goal? What is your schedule? What are you capable of eating? Are you vegan, vegetarian, blah, blah, blah? You know, are you dairy-free? Like, how do you hit that number? You know, all those things are the you parts of it. Are you in fat loss or not? This and that. That's where a real person gets stuck. That's where people get stuck. Okay. The translation, the translating that principle to your messy, chaotic real life, AI is great at. It's actually great at because it's not just retrieving a fact, it is taking all the data that you've told it, finding the patterns like we just talked about, and then translating it to you. And you actually don't need a human for that per se. Okay. Now, that's like there's a caveat on that because obviously if the AI is not asking the right questions or you're not feeding it all the data it needs, it can give you a garbage answer. Whereas a human might say, Whoa, whoa, wait a minute, I need to know this and this as well. But that doesn't mean AI can't do it. Okay. Understand the difference. It has to be good enough to do it and you have to use it the right way. And then number four is AI is good at asking the questions that you don't know to ask because you don't know what you don't know. So, as an engineer, I come from the world of requirements for a product, and there's things that you know, there's things you know that you don't know, and there's things you don't know that you don't know. So instead of like, hey, what's the best exercise to get bigger glutes? Because I want a big butt, you say, Hey, what am I not considering about my training or my protein as I go through menopause and given my program right now because I want a big butt, right? Or something like that. Like giving it that context and then have it give you what the blind spots are. You know, notice what I just said. What am I not considering about dot dot dot? Great prompt right there. What am I not thinking about? Right. And so that would turn AI kind of like into a coach that asks you questions that you wouldn't think to ask of yourself. So those are the four things pattern recognition in your data, an accountability structure, applying principles to your specific situation, and then uh kind of surfacing up these blind spots that you're not thinking of. Now, what does every single one of those have in common? Well, they all require you to bring something to the equation. That is your data, your plan, your situation, your context. I mean, until the day that AI is like hardwired into our brain, God help us all, you have to bring that stuff. Now, I just I mentioned before, sometimes it's integrated with Apple Health or Aura rings or whatever. And granted, that is kind of an automated way to do that. I'm not saying it doesn't have to be automated, just you have to bring that stuff. And so the good uses of AI, in my opinion, are a conversation about your actual life and getting some help on the blind spots and structure accountability, all that stuff. Whereas a bad way to use AI would just be a generic search, a generic question, using it as kind of like an advanced Google. Does that make sense? This

  Coaching that works with your data

  Philip Pape 19:58

  is honestly right now the perfect place to mention the thing that I've built that's out there that I think takes advantage of all of these in a very thoughtful way, where AI is the application layer. It's not an information fire hose. It just dumps information at you. It really is the this intelligence or interpretation on top of your data. And that's called Fitness Lab. This is my app. I'm giving you a shameless plug for it because it does exactly what I've thought through AI can do well, that can help in the fitness and training world, that also is pre-trained with lots and lots and lots of good evidence-based data, as well as my entire podcast, all my guides, all that fun stuff. It's all trained into there. Exercises, you name it. What I've found is that people are kind of sick of just having more trackers, more tracking apps, right? For training and nutrition, what do we need? Well, we need a way to track our workouts, a way to track our nutrition. And then we probably want to track some other measurements, right? Our weight, our measurements, our photos, things like that, body fat. And most people are tracking with a few of these apps. Some people have app burnout. That's what I find a lot. But even when you're doing all that, you don't have a way to bring it all together unless you have a coach. And we've talked before about human coaching, right? It's expensive, it's hard to get a good coach. It's getting more and more where it's, it's, you know, it's a tough industry, let's just say, for on both sides. Okay. And yet if I said, hey, I just want to use another yet another app, here's another tracker. You're just tracking, tracking, tracking, tracking. Even if you're tracking in a notebook, how are you using that information? Even if you're doing something like starting strength and you're using a notebook and tracking your lifts, do you truly know what decision to make at your next session based on that data? That's what Fitness Lab does. It is not trying to be your tracker, it can definitely track lots of things. It integrates with Apple Health, all that good stuff. We tried to reduce as much friction as possible on that side. But if you still are tracking in other apps like MacroFactor or a workout app, it can work alongside it in a really cool way. And it does what we talked about before. The more data you have, the better it can actually use that data to help you do what you need to do. So, what it does is the four things I talked about. It look at your patterns, it looks at your biofeedback, like sleep and hunger, et cetera, at things like your protein. If you've been feeding it meals, not like a tracker, but more like an analysis tool. Your training, it tells you what's going on. It tells you here's what, here's the pattern, and here's the questions I want to ask you to make sure you do the right thing next so that you don't have to stress about it. So basically gives you daily coaching. It gives you activities each day. It adapts to you. You can talk to it anytime and it can adjust any of those activities based on what you need. If you're like, man, I'm burnt out. I don't want to track, it'll say, okay, let's just limit it to one or two activities a day and we're still going to make progress. All right. Right now, we're running a summer special from June 23rd through Friday, July 3rd. If you're hearing this afterward, I want you to reach out to me. I can probably get you a secret discount. But during the promotion, it's 20% off going to our link witsandweights.com slash app. Go to wits and weights.com slash app. As of this episode dropping, you've got about five days left. And if you've been listening to all of this, you're like, okay, I get it. I need that application layer to do something productive with the data. I don't need more information. I want to lose fat. I want to get rid of that belly fat. I want to know how to train. I want to progressively overload and get stronger. I want to deal with my sleep issues. I want to deal with my food noise. All of that, it's going to help you with that. Go to witsaweeits.com slash app and grab that 20% off discount. I would recommend signing up for the year subscription because you'll save even more doing it that way versus the shorter one. Go to witzawaits.com slash app for fitness lab today.

  Bad AI prompts vs. better coaching questions

  Philip Pape 23:40

  Okay, let's get back to it and talk about okay, how do you use AI? I think the fastest way to fix this is with a few examples, right? A bad use versus good use of AI. So a bad use of AI is opening it up. And again, any AI I'm talking about. Heck, even my own app, if you just ask it this way and you say, like, what's the best workout for a woman over 40? Or weight loss for women in perimenopause is a super common search. Well, you're gonna get a perfectly reasonable, totally generic plan that you will not end up doing. That's that is the honest answer. You'll just see it, you'll say, like, okay, that's nice. Kind of like an article in the newspaper. That's nice. You see this, honey? Okay, moving on. I'm gonna go have my uh oatmeal now. All right, now why aren't you gonna do it? It's because it has nothing to do with you, your lifestyle, your equipment, your dieting history, your particular issues, right? We've we've kind of nailed this or hammered this to death, but you know what I mean. It's an answer to a question that unfortunately has a thousand correct answers, kind of like the infinite universes, the multiverse, right? Spider-Man, the multiverse. Anyone watching Nick Cage and Spider-Noir, awesome show on Prime. Uh, when do I find time to watch TV? I I'm only one episode in. So anyway, it's a thousand permutations of an answer, and it's not going to help you. That's the information fire hose. That's more plan, but no more progress. We don't want that. A good use of that would be pasting in your current program, the one that you're actually running right now. So it could be a screenshot from your workout app. You can type it in, you can take a photo. There's lots of ways depending on what you're using. Like if you use Fitness Lab, you can go into the chat and you could do audio or what you could do video or files or text or whatever. And you say, okay, here are my constraints. I train three days a week. I only have dumbbells. I don't have any micro plates, my knees are cranky. Like, what's the weak link here? That's a great way to question, ask it. Like, what are the blocks? What am I missing here in my plan? I mean, you could say, hey, just give me a better plan, but I love to start from finding the weak spots. Right. So now that helps you learn where the plan doesn't stack up and improve what you're following for the things you just told it are your constraints. And that's super powerful. And you could have all sorts of constraints, right? The days of the week that you can do it, like anything, you know, you hate doing this exercise, whatever. So, do you see the difference between those two? In the first one, you get a generic answer based on information, but the second one is diagnosing you and giving you the blind spots and how to improve. And by the way, you're gonna learn in the process as a human. You're not hopefully not just blindly following it and moving on, but it's like, okay, now I understand why it's working. And again, my app, Fitness Lab, is is kind of cool because it it talks to you like I would, if you will, for better or worse. And it kind of tries to explain it in an accessible way. Here's another, here's another bad example, right? Or bad versus good. You say, is creatine good for women in menopause? I don't know why I keep hammering on the women in menopause, but it's probably because I hear a million questions in about that. So is creatine good, right? And and honestly, that's a question I get, I get that question over Instagram or people had just found the podcast. Hey, is creatine, should I take creatine? I'm like, really? That like that's your question. Sorry, I don't mean to judge people, but like we need to, we need to put more effort into our questions, but I get it, I get it, because maybe you've never heard of creatine you just heard about, you're like, what is this creatine thing? Is it good to do? I would say a good use of AI or asking a question would be, hey, I'm I'm thinking of starting taking creatine. Here's my goal, here's my routine, here's my other supplements, how much do I take, or how do I remember to take it every day, or how do I know it's working? Like there's a lot of ways. I'm not telling you that there's one right way to ask the question, but almost pretend like it's a coach in front of you that you're picking their brain. That's really what you're trying to do. All right. You're not just trying to get a generic fact, you're trying to find out how it's gonna work for you. And again, the more data you give it, the better. I hope this episode isn't going too long, but I do have one more. This is this is for all of us here. Why am I not losing weight? Okay, well, it's an interesting question, right? Because some people say, Why is my metabolism too low? Why am I not losing weight? One of the most common questions. And as a coach, I would say, Well, I need to learn a lot about you right now, based on that. In fact, people sign up for coaching calls with me just to dig in and kind of figure the heck out of this stuff. Because last I counted, there are at least 13 specific reasons for a plateau. And I've built them into like the tools that I create to help people understand where your plateau is coming from. But then even to get to one of those 13 answers, you need like 1300 pieces of information. I'm I'm I'm exaggerating, but you know, you need data indicate data out. So if you just ask the question generically on Google or AI, you're gonna get all a list of all the things. Now you have to go figure out, okay, which of those lists apply to me? Calories, sleep, stress, right? The usual suspects. Or you're gonna get fear-mongering, like, oh, it's your hormones, it's cortisol, right? It menobellies called by cortisol caused by cortisol or something like that. So it's generic, it's useless. And look, you already know a lot of those, don't you? Yeah, I think you do. Or if you listen to the show, you do. Maybe maybe you don't. That's fine. Basic information's cool, it's good education, but it's not gonna help you. A good use of AI for this, if you're not losing weight, is take three weeks in a row of your data. Now you're like, oh crap, I don't actually have data. Well, listen to my podcast for a while, folks, because you're gonna learn how much I love tracking. And if you feel like tracking is a chore, there are easy ways to do it that are low friction. It's a problem we deal with all the time. Okay, but if you don't track, how the heck can you tell what's going on? You've got to track something, right? You've got to track something. So, and maybe this will motivate you to do it. Knowing that you can get an answer may motivate you to do it. So take three weeks of your food log, your training, your weight trend, your biofeedback, like your sleep, and then you paste it all in. Maybe you have lab work too, but I would keep lab work to the side initially unless you really suspect something. And then ask the AI, hey, what patterns do you see? And what would you change first if my goal right now is to lose weight or to lose fat? Okay, and if if you're using Fitness Lab and you ask it, it might even push back and say, hey, we're not trying to lose weight, actually, we're trying to lose body fat. It's gonna, it's gonna mansplain to you, but actually in a general uh gender neutral way, I guess. But I'm joking, kind of. It will then help you understand. Ah, okay, here's the thing that's the most important for you right now. That's pretty cool. That's pretty amazing. Like if you all you have to do is have data, and I get it takes work and effort to track data, but really not that much, guys. Not that much. Okay, just a few things can be very powerful. So the through line on all of these, I've heard the word through line lately. It's like jargon buzzword I hear everywhere. Maybe it's because of AI. I don't know. The through line on all of this is that every good use of AI or a human coach, but we're talking about AI today, starts with you giving it some good data. Your plan, your goals, your data, your constraints, all of that fun stuff. The the incorrect use of AI is without that. It's just asking questions, but not, and it's lazy. It's a lazy use of AI that's gonna give you a lazy answer that's gonna result in you not getting any progress. Okay. So I am I am not saying that AI is useless for fitness. I clearly don't believe that. I built an entire tool around it, Fitness Lab. We talked about it. I'm not saying information doesn't matter. It doesn't, it does. If you are genuinely like at a zero out of 10, you got to learn stuff. You gotta learn the principles. Go ahead and use AI to learn them. That's fine. Be a little bit cautious and careful on where the data is coming from, right? Again, that's where if you had Fitness Lab, you can learn as part of the process. In fact, in Fitness Lab, every day there's a briefing that teaches you something. That's pretty cool. So that helps you level up on your knowledge. I'm not saying you should trust AI blindly. That absolutely you shouldn't, because the tools tend to want to agree with you and like give you a pat on the back and say, like, hey, great thinking there, even when it's terrible thinking, I'm sorry to say. And there is good research showing that they still have a strong tendency to just go along with what you say, even when you're wrong, because they're built to be that way. So they're not a substitute for an actual human coach who's gonna give you, I'll say, tough love, something that you don't want to hear. Now, caveat on that, we're gonna get back to at the end of the episode. Can you train your AI to give you some tough love? Yes, you can. So hold on. I'm also not saying AI replaces a human. I mentioned that 74% improved results when you had the human with the AI. And I get we can't all afford a human or whatnot. So pick the best combination of tools for the job that you could afford that fits within your life. That's what I'm saying, right? The bottleneck for almost everyone is not the information, especially who listens to this podcast. You got the information. It's applying it and doing it consistently. And AI is a phenomenal application tool, but it's like a mediocre information toy. I mean, it's great for information, but what I mean is it's not going to do anything for you just by having the information. All right, before we wrap up, I did promise you a one-line prompt that can train your AI and turn it from a yes man into something that will help

  The one prompt that makes AI stop being a yes-man

  Philip Pape 32:36

  you more. I'm gonna give you that in a second. But if if this episode has surfaced all the things you've been thinking about and wondering about AI, and you're like, yeah, now it makes sense that, you know, I've got all this information, I follow so many people, but I don't have the results. That is the gap Fitness Lab was built to close for you. I am so passionate and supportive of what we did here. I think it's a fantastic tool. People are using it, love it, and they keep renewing because it's helping them. Those who actually use it, it's helping them. It is that layer that looks at your data and it coaches you on those patterns and it helps you keep consistent instead of just handing you more information. And again, we have a summer special right now. Go to wits and weights.com slash app, 20% off. When you go to that link, you take a very short two-minute quiz and it gives you what your plan would look like if you decide to use the app and you don't have to then use the app if it's not right for you. So it's a really great zero risk way to check it out. Go to witsandweights.com slash app. Link will be in the show notes. Okay, here's the one-line trick that changes how AI will talk to you. So if you ask AI, hey, is my plan good? Or what do you think of my plan or spot check my plan, it's basically gonna agree with you. It's kind of the default setting. Depending on how you ask it, it might go a little bit deeper than that, but you have to force it out of that. And this is where prompting is very important. Now, in some tools, depending on your account, what you have access to, like Claude or Chap GP or whatever, you can go into your profile, into your settings, and you can give it instructions that it always looks at. In other words, it's like global instructions all the time. And I actually have some of those. But even if you don't use that and you just want to type it right into the chat, what I would do is, you know, paste your data in and then you add, put your question and then put the following act as a skeptical coach. Argue against this plan. Tell me the three things most likely to fail and why. That's it. Now, if you guys want to know what I use in my instructions, I have a very detailed uh set of instructions on like, hey, don't sugarcoat it, don't tell me what I want to hear. I want you to push back. Like, I have all this stuff because I've learned over time how to make it more and more tough on me so it's a more realistic. But just in the chat, do what I just said, go back and listen. I'll repeat it right now, actually. Act as a skeptical coach, argue against this plan, tell me the three things most likely to fail and why. That's it. That instruction turns it from this sycophant, this cheerleader, into a critic, like in a good way. It's gonna give you helpful critique, and that's where the gold is, right? That's where the gold is. That's what you need to hear.

  The real reason most people stay stuck

  Philip Pape 35:20

  Okay. So I hope that was helpful. I hope this whole episode was helpful. I know I pitched my app in there a lot because I truly believe in it and I think it does what we are talking about. It's why I made it. I hope this tool, this episode helped you in even just day-to-day use of AI. And just go forward and conquer and get those results, please. I'm rooting for you. Until next time, keep using your wits, lifting those weights. And remember the answer was never the thing you were missing. The doing is. I'm Philip Pate, and I will talk to you next time here on the Wits and Weights podcast.