215: Can AI Actually Help You Lose Weight?

Scot: Hey guys, I've been using AI to help me structure and maintain a nutrition plan because it's time. It's time again to work on losing some body fat. And I thought it'd be kind of cool to talk about how I've been using AI in this latest of many throughout my life times that I've decided I need to lose some body fat.

So quick question. First of all, Mitch, I know you use AI extensively, right? You're pretty . . .

Mitch: Oh, yeah.

Scot: Thunder, do you use AI?

Thunder: Yeah, I use AI for a variety of applications. Some of it related to my work, and some of it just completely off the deep end, like how to set up my ski bindings on my race skis. Obscure things.

Scot: Yeah. How about you, John? Do you have AI skills?

John: I wouldn't call them skills. I'm learning how to prompt well. I mean, I think that's 90% of the battle, but I use it often.

Scot: That's awesome. So I think it's worth mentioning if AI is something you haven't used, this might feel a little weird. It might be a little tricky because AI has a few quirks, and I'm going to talk about what some of those quirks are with this program. But for the most part, it's been really, really good. So I can't wait to get into it.

This is "Who Cares About Men's Health," information, inspiration, and a different interpretation of men's health. On the show, we've got myself. I am Scot. I bring the BS. The MD to my BS is urologist Dr. John Smith.

John: Good afternoon.

Scot: Thunder Jalili, he's got a PhD in nutrition, and he also specializes in biochemistry and molecular biology. He's in the Department of Nutrition and Integrated Physiology at University of Utah. So he's bringing the nutrition information. We're going to see if ChatGPT is feeding me a line or if it's actually some decent advice. You ready?

Thunder: Yeah, I'm ready.

Scot: All right. And we're also going to find out if it can replace Thunder, if I need to have Thunder on the show anymore.

Mitch: No.

Thunder: That's right.

Scot: And also on the show, he's playing a little injured today. He's not feeling great, a little under the weather, but he's still here. He promised us a B-minus. So Producer Mitch, I'm glad that you're here.

Mitch: We're going to give you B-minus, a C-plus. I don't know. I'm here.

Scot: All right. So it all started when, again, I got to the point where I am heavier than I want. The clothes aren't fitting the way I want. I'm not digging the number on the scale. And is it a vanity thing? Yes, to some extent. I don't want to go out and buy new clothes, so it's a financial thing as well. But it's also a health thing because we've talked a lot about how that extra body fat really can change how your metabolism works. It can lead to metabolic disease. It can lead to heart disease. So keeping your body fat in a reasonable place is a good idea. So I'm like, "All right."

But here was the thing. I'm like, "I don't want to have to look up my food." So every time I want to eat, I've got to look up, "How many calories is this food? How many calories is that food?" And then I've got to log it and I've got to weigh it. I need to get out of weighing it. And that's when I thought, "What if I use AI?"

And I'm wondering too, as a guy that kind of knows how to do this because we've had a lot of episodes on it, I wonder how much ChatGPT could help just a layperson, even though I'm not, lose weight.

So you ready for my starting prompt? Starting prompt: "I'm trying to lose body fat. I'm strength training twice a week, but I'm going to increase my amount of cardio to a minimum of 210 minutes a week. That's 30 minutes a day. But I need to short myself calories. I plan on eating a diet with protein and whole foods, but I want to severely limit my calories as much as possible while still getting these nutrients. I'm a 55-year-old male. What's the fewest number of calories I can eat and still get the nutrients I need, and what supplements should I take in my journey?"

I'm sure Thunder picked out a word in my prompt right away. You want to go ahead and tell me what word stood out to you, slapped you upside the face until you were red in the cheeks?

Thunder: The emphasis on reducing calories as little as possible.

Scot: Yeah, severely, right? So I will say ChatGPT talked me out of that. It actually . . .

Thunder: Good. Good on ChatGPT.

Scot: Yeah. It actually said, "Hold up, that's not the best idea. You shouldn't severely cut calories like that because then what's going to happen is you're going to lose muscle mass along with the fat mass. It's not healthy to begin with. So you're going to lose weight, but you're really not going to have lost really that much fat." We had a little conversation back and forth about that, so I thought that was awesome.

And then at one point I asked it, "All right. Well, how many calories do I need? What can I do reasonably?" So it figured out what my BMR was, which is the amount of calories that you need to consume if you just were a slug on the couch all day and didn't move. And then based on my activity that I told it, two strength trainings, 210 minutes of Zone 2 aerobic training . . . Which is another thing it counseled me on. It said, "You want to stay in the low end of Zone 2."

But it came up with 1,600 calories a day just to lay on my butt, and 2,400 to 2,500 maintenance calories. So that would be maintenance calories where I wouldn't gain nor lose weight if I was doing the workouts and the cardio.

So Thunder, first question. Does that sound reasonable to you?

Thunder: In broad strokes, it sounds reasonable. You're targeting the amount of calories to meet what you're expending with your resting metabolic rate and your daily activities added on top.

Scot: And that number seems reasonable? Sixteen hundred to be a slug, and 2,400 to 2,500 based on the activity that I told you I was going to do.

Thunder: Yeah, it seems reasonable. I would be surprised if it was higher than 2,500, but yeah. I mean, most people don't quite need as many calories as they think.

Scot: Right. I thought that seemed high. I thought 2,400 to 2,500 seemed high.

I do also need to say I had some data that I was able to give ChatGPT. I have my body weight, my height. We all have that, right? But I also had a general idea of my percentage of lean mass, and that gave me the ability to tell it a general idea of how much body fat I had right now. I've done these BOD PODs in the past. I've been at this number before. I've seen this guy in the mirror before. I kind of have a pretty good idea where I'm at. So I think that was useful.

John: So you talk about those things of giving ChatGPT that information. Your prompt that you gave us in the beginning, did you give it more information after it came back with what it told you to do so that it could refine that a little bit further? Is that what you did?

Scot: Yes, absolutely. And I wish there was a way for me to re-create that whole conversation, but it was a back and forth. There was a lot of back and forth between me and ChatGPT. I'd be like, "Wow, 2,400 to 2,500 calories sounds kind of high. Are you sure about that?"

One of the little weird things about ChatGPT is sometimes it doesn't get stuff right. And if you call it on it, it will correct itself. So I did, and it went back through the math again. It said, "Based on the fact that you're this tall, weigh this much, have this much lean mass allegedly, this much fat, that much activity, for somebody that's 55 years old, that should be kind of where your beginning range would be."

So yeah, I think that's probably Lesson 1. Ask questions and push back. Use this as a learning experience as well. Again, I have a little bit of experience in all aspects of this from ChatGPT to calories and exercise and whatnot.

Thunder: Hey, Scot, can I add one other addendum to Lesson 1?

Scot: Yes.

Thunder: Be really careful about the data you feed in. You fed in exercise data, like, "This much cardio, this much strength training, etc." Strength training is a pretty ambiguous term. What does that mean? How intense are you lifting? How much are you lifting? And there can be a lot of calorie variation in that.

Scot: That's great. And I'm glad you brought that up. That is something I did. I told it I have two full-body sessions per week, 30 minutes each, one set per body part. So I mean, I gave it that information. But that's a great point for somebody that wouldn't know to give it that information.

Mitch: Scot, I do want to pause for a second with this discussion about AI. You are, one, already kind of healthy and know what you're looking for. And two, you've used AI quite a bit. You know how to ask it follow-up questions, etc.

I worry sometimes that there are people out there that maybe are going to ask it like, "Oh my god, I want to lose weight. Would eating a grapefruit three times a day be good?" And it would be like, "Yes, user, 1,000%. That would be the best diet ever."

And so, if I may, there are three main things that AI can get wrong that I train my students on these days that not everyone is super-duper aware of. Would you indulge me?

Scot: Indulged.

Mitch: Perfect. So number one is hallucination. We've heard this one before. It is where an AI or a large language model completely makes up something. It does not have the information readily available, and so it just whips up a study or whips up some idea, right? We're pretty good at spotting that because we are aware of it.

But the other two, one is called confabulation. This is where the AI basically runs out of resources and basically picks pretty good information to fill in the blanks. The problem is that that "pretty good" information does not engage any of its reasoning circuits.

So in your particular thing that you sent us, when it is looking at, say, dosages that would be good for you, etc., if it hits a token limit and it says, "Hey, I'm done thinking about this," it will just fill it in with whatever it has best available to it, right? And that can cause some real problems because it sounds like real information. They are giving us real stats, but who knows if we just asked it a hard enough question that it filled in the blanks.

Scot: Thunder, sounds like you're hanging around.

Thunder: Okay. I guess I have job security for a few more months.

Scot: That's right.

Mitch: And the third one is what I think Thunder is really good about, and it's called sycophantic drift. And it is where the AI wants to please you, right?

And so you prompted it with saying, "Hey, I want to cut calories." Luckily, this particular model either has interacted with you enough or knows you well enough that it kind of pushed back. But if you were to really push it, if you were to say, "No, I think I really do want to eat 100 calories a day," your robot will tell you that is a good idea.

So much so that the more you question it . . . what we find is the further recall and response that goes back and forth with an AI, the more likely it is to start to reshape to your ideals and values, which, when it comes to medical questions, can be problematic because as you ask more and more questions, it's more and more likely to give you bad answers.

So those are the three things to look out for: hallucination, confabulation, and sycophantic drift.

Scot: That's good because I guess the information back is only as good as the person . . . Because I do have a little bit of experience with this, I'm less likely to put it in a situation where it's going to do some of those things, it sounds like.

Mitch: Yeah. And you know who won't? It's Thunder. Thunder won't do any of those.

Thunder: Nope. I'm an open vessel. Just tell me what I should do.

Scot: Oh, no, you're not. Not at all.

All right. So it figured based on that, if I wanted to lose fat, my target for fat loss should be 1,900 to 2,100 calories a day. And it told me this creates a deficit of 300 to 600 calories a day, which supports gradual fat loss while preserving muscle and maintaining training performance. So 1,900 to 2,100. What do you think of that for a number, Thunder?

Thunder: It definitely sounds like it would be a calorie deficit. There are two ways to look at that. There's the total number of calories you're eating, and there's the food sources that are providing those calories. For someone who eats largely healthy, I could see that, "Okay, let's just reduce a little bit of what I'm eating. I'm largely eating healthy."

But for someone who is maybe not eating largely healthy, instead of worrying about reducing calories, they should try to reduce or eliminate the things that they know are blatantly unhealthy. Start there. Don't just eat less of the other stuff, if that makes sense.

Scot: And I will say, in my instance, I don't know that I was eating mostly healthy, and I personally just feel better if I have the numbers. If I'm weighing the food, if I'm logging that stuff, I personally feel more hope than if I'm just kind of blindly going. And I guess that's kind of a personal thing probably.

I will say even though I've done this before and I feel like I have a pretty good grasp of food, I discovered a couple surprises about nutrition along the way. So this was the first thing that I had a back and forth with, and I even said at one point, "My friend Thunder is going to disagree with this."

Thunder: Did ChatGPT say, "Don't listen to that idiot"?

Scot: No. The daily protein target, 140 to 160 grams per day.

Thunder: Wow.

Scot: Which is 1.6 to 1.8 grams per kilogram of body weight. And we've done shows in the past where Thunder's going to say that is pretty ridiculously high.

Thunder: Well, it's definitely a solid number. It's not necessarily high. It just comes back to your exercise load. What does your strength training look like? How much muscle damage are you inducing in your resistance workouts? So it could be fine. It's just something you have to kind of be honest and, again, try to give that really good data to ChatGPT so it's nailing the protein target.

Now, having said all that, if you're eating 20% extra protein, you're not going to drop dead. It's just that from a practical standpoint, sometimes people have trouble figuring out how to get that much protein in their diet.

Scot: And it is a challenge, as Mitch can attest, because Mitch has done this before.

Mitch: Oh, yeah. Where you're eating a chicken breast or two each and every meal and you're just choking it down by the end of the day. Yeah, I've done that.

Scot: Right. For example, my breakfast right now . . . I'm in a calorie deficit, but this is my breakfast. Two eggs, a cup of egg whites. I put just a few vegetables in there, and then I have a cup of oatmeal, and then I have three-quarters cup of yogurt. So you've got yogurt and eggs for the protein. And that's what I have to do. It's insane.

But here's what ChatGPT had to say about that, because I did question it. And I also tend to agree with Thunder, that a lot of times we over-protein ourselves. But I liked its answers. First of all, it said, "You're resistance training." And I did give a little pushback, that this is only one time a week. It's not like I'm an Olympic-level resistance trainer. But it also said, "When you're in a calorie deficit, it's a good idea to increase your protein to help preserve muscle mass."

Thunder, what can you say to that? Is there research to support that?

Thunder: Yeah, that's true. It depends on how much of a calorie deficit you're in. If you're in a mild calorie deficit, it's not as critical. If you're cutting your calories in half . . . I think you said 2,500 was your recommended. If you go to 1,200, then yeah, the protein becomes a lot more important.

Scot: And then the third thing . . . and I actually read an article a long time ago in the "New York Times" about this, supported by research. When we get older . . . and I'm not sure what the age cutoff is, but it likes to remind me, "As a 55-year-old male who's trying to . . ." It loves to say my age. The older we get, the less we are able to efficiently process protein.

So it's saying that muscle protein synthesis decreases with age, meaning that you should have slightly higher protein intake in order to maintain muscle tissue in older adults. Have you heard this, Thunder?

Thunder: Yeah, that is true. When we're young, we're wildly efficient at stimulating muscle protein synthesis. Insulin and amino acids are super potent and stimulating. As you get older, that effect is blunted, and getting your kind of maximal usable amount of protein in a given meal is definitely key.

But the amount that you need for maximum muscle protein synthesis when you're older is not as much as you think. People have done studies on this with young people and old people. When you get above 25 or 30 grams of protein, you don't really increase muscle protein synthesis anymore. It kind of maxes out. There's not a lot of efficacy pounding 3 chicken breasts to get 90 grams of protein or whatever in a meal, because you're not going to use that much for muscle protein synthesis. Now, you will use it for other stuff, like for calories. And we can talk about that later if we want to get into that.

Scot: Yeah. I'm eating about 50 grams per meal, give or take, right?

Thunder: Yeah. You're using half of that for just your calorie intake. It's not going to protein. But that's all right because what's happening if you're in a calorie deficit is your body is looking for other things to fuel it, especially if it goes more than a couple days. Now you're pretty much using your fat stores to make ketones, and you may be tapping into your muscle stores to strip away those amino acids and use those as carbon skeletons to make glucose and stuff.

So it's not a bad idea to have some extra protein in that meal. You can just use it for calories. But it's not going to stimulate your muscle protein synthesis any anymore.

Scot: And I think ultimately, I settled with, "Okay, I'll go along with this," because of the reason that you said, that if I get another 10, 20 grams above kind of what you normally would have, there's not really a downside to it.

Thunder: No.

Scot: All right. So Thunder, next thing here is . . . here are the targets that it ended up setting up for me based on the calorie goal. Let me know if this ratio sounds good. So protein, 140 to 160 grams. Carbs, 170 to 210 grams. Fat, 50 to 65 grams. Now based on stuff I've done in the past, that seems pretty reasonable.

Thunder: Yeah. I'm not going to push back too much on those general guidelines.

Scot: But what you do want to talk about is the types of foods, it sounds like. If you remember, my prompt said I want to eat mostly whole foods. And based on what Mitch said earlier, that could have steered it in a direction that maybe it might not go if somebody else would've put the prompt in and left that out. It recommended stuff like steel-cut oats, potatoes, sweet potatoes, and squash. By the way, I've grown to love those three things, Thunder.

Thunder: There you go. Welcome to the club.

Scot: I've always loved potatoes, but I guess I didn't realize how really kind of low in calories they are for the amount of food you get. I always thought potatoes were kind of a dense food that you didn't want to have a lot of, but they actually have a lot of volume for the calories you get.

Thunder: Yeah. Especially sweet potatoes.

Scot: Yeah. Fruit, vegetables, whole grain bread, those were some of the carbs that it recommended. And then it gave me some daily targets of vegetables of three to five cups a day. One to three servings a day of fruit. Does that sound reasonable?

Thunder: Yeah. I mean, it's a little on the low side.

Scot: Oh, really?

Thunder: A serving is usually less than what you think it is. But I mean, it's not crazy. You're eating fruit.

Scot: Right. So three to five cups of veggies, one to three of fruit gives you four to seven total fruits and veggies. What would you like to see that at?

Thunder: I mean, personally, I think fruits and vegetables are a great low-calorie way of making yourself full and getting tons of nutrients for not that many calories. So if I'm counseling someone to lose weight, I would pick those as things to emphasize in the diet.

And I would say if you're wanting to lose weight, let's also jumpstart it by reducing your bread intake or not eating that much bread or pasta, but get those carbs from fruits and legumes and vegetables and things like that.

Scot: I will say sometimes the portion sizes seem a little crazy. So I think it's trying to get me to get more than that in serving sizes. One time it recommended to have 300 grams of Brussels sprouts, which is 10 ounces of Brussels sprouts, which is three-quarters of a pound of Brussels sprouts.

Thunder: Dude, you'd be living on the toilet with that much.

Scot: I guess that explains the gas.

Thunder: Yeah. Exactly.

Scot: It also, I think, was telling me . . . So I also like to eat broccoli and cauliflower and bell pepper, and I put it in just about everything, salads, my breakfast omelet in the morning. I think at one point it told me to have 300 grams cooked of that out of the air fryer. And I remember pouring that onto my plate, and I'm like, "You've got to be kidding me."

Thunder: Choke it down.

Mitch: That's too much.

Scot: Yeah. And then on top of that, it's like, "Oh, do 250 grams, 300 grams of potato." And my plate was . . .

Thunder: You're already up to a pound and a half of food.

Scot: Yeah, the first few days. So I've since told it, "I can't eat this much. We need to figure something else out."

Thunder: Yeah. Well, AI has no frame of reference because it doesn't eat.

John: Do you feel like the structure has helped you in preparation and getting to the gym? What's the benefit been for you in using AI? For someone who may not have done this before, what's your true benefit that you've seen? Obviously, this helps you to kind of structure and plan things, but what overall have you seen is the biggest benefit for you?

Scot: Yeah. So I would say definitely the structure and planning, the fact I don't have to log it, and then the fact that I eat my breakfast, it knows what I had, and then I'm in the kitchen for lunch, and I'm like, "All right. You know what I had for breakfast. Go ahead and summarize that for me." Getting back to the idea that sometimes it just makes stuff up. So I can visually confirm it's got the right information.

And I'm like, "All right. This is what I've got. What can I put together to hit my macros and calories for lunch?" And it'll be like, "Oh, you've got a can of tuna. Great. Why don't you go ahead and take that and put some Dave's Killer Bread on there and make 300 grams of Brussels sprouts." And I'm like, "No." "How about 100 grams of asparagus? Would that work for you?" "Oh, that would be great." And then it logs that, right? So I don't have to worry about logging things.

And then after I get done eating, I can go, "All right. Where am I at? What do I need to eat for my next meal?" One of my big challenges is eating enough food, actually. And when you're eating whole foods like this, you're eating a lot and for a long time when you sit down. You sit down and chew a lot.

Thunder: It's an event.

Scot: It is an event. It's exhausting, is what it is. So it's easy for me to just be full all the time. So that was another benefit. I'd be like, "Gosh, I still feel full from lunch. Here's what I've got. What can we do?" And it will give me some suggestions that might be a little lighter. Maybe that's where the bread comes in, right? Because It's not as filling. It still has some calories, but it's not a pile of sweet potatoes. So that's been super useful for that.

The other part that's been super useful is a week into this, I went to visit my mom, and my mom was like, "Hey, when you get here, let's have pizza night, and then I want to have hot dogs the next night." She's 95 years old, and she's on a hot dog kick.

John: My girl. I love your mom already.

Mitch: That's what I was about to say. I'm into it.

Scot: So, anyway, I'm like, "Well, this is going to derail the whole thing. Not only that, I'm going to be on the road for eight hours." So I'm like, "All right. Well, tomorrow, I'm leaving on a road trip. I can eat a normal breakfast, but I'm going to be on the road. And then when I arrive, it's pizza night. So how can I make this work?" And it helped me plan around that. It took the event . . .

And regardless of what the event is now, if it's eating out with friends, if it's wings with my friend, I will tell it, "This is the thing I've got today. How can we build the rest of my day around this so I can still do that thing?"

And so for pizza night, it's like, "All right. Well, you're going to probably have an unknown amount of fat in your pizza and high carbs. So let's still have your carbs in your breakfast, but instead of two eggs, let's just go with some more egg whites instead. So we're going to reduce your fat throughout the rest of the day because you're going to be hitting that fat at night." Same with the chicken wings.

It's been super useful to work around those little things that used to just derail you.

Thunder: I have a question. So after hearing what you've been talking about with AI and the tracking and all that, the million-dollar question is have you lost weight doing this? And how much?

Scot: Do we need to wait until the . . . That feels like an end-of-the-episode question, doesn't it?

Thunder: It is, but you've gone so much with the description that . . . And we can keep talking about it, but I've got to know, has this been working?

Scot: So I'm three weeks in, Thunder. Have I lost weight? I have.

Thunder: Yay.

Scot: I haven't lost a ton. Now let me tell you some of the confounding factors in this, in my opinion. Before I did this, I didn't really eat that much. I was almost kind of starving myself. So now I have more food in my internals consistently throughout the day, right?

Thunder: Yeah. Fibrous foods, which take longer to go through.

Scot: Right. So immediately I noticed the scale go up because now I'm eating again and I've got food that I'm digesting. It's going through the system. Number two, I started taking creatine three weeks ago.

Thunder: Oh, that'll retain water.

Scot: One time I started taking creatine, I think I put on eight pounds in two weeks. So that also was a struggle, right? I know creatine is really good, the research is good on it, but that's not the direction I want to go and it's not very good for morale.

Thunder: Are you still taking it?

Scot: Yep, I am. So three weeks in, I have lost three pounds of body weight. When I weigh myself first thing in the morning, empty stomach after I've gone to the bathroom, and I've lost an inch and a half off around my waist.

Thunder: Well, you've lost fat in the right place. That's good.

Scot: I have. So it is going in the right direction, although it doesn't feel like it. If I wasn't doing those measurements, I don't know that I'd feel as though that I've made any progress.

Thunder: Scot, you know what would be much more interesting than the scale? If you had done a body composition before and done one now three weeks in.

Scot: Yeah, that would have been interesting. But have you ever been to the BOD POD place? They've got these cute girls that are there, and you have to wear spandex shorts.

Thunder: They're all judging you.

Scot: Yes, they are. They're looking at the cautionary tale of the 55-year-old man that they never want to see anybody in their life become.

Mitch: That's not true.

Thunder: But the thing with weight loss is you can be losing some fat using this approach, and you probably are, but if you're exercising, you're doing your creatine, you're working out, you could actually have gained some muscle mass. So that's why the body composition is really important, because the percent fat is really the important metric here, not what the scale says.

Scot: Yeah. One hundred percent. And that's what I'm shooting for too, right?

Thunder: Yeah. And your pants test, the smaller . . . Do you remember the punishment pants from an episode we had a long time ago?

Scot: Yes, I do. I need to go revisit the punishment pants and see how they fit.

Thunder: Yeah, the punishment pants are probably going to fit great. But that's a good indication of the fat loss because you're not building muscles in your abs, right?

Scot: I have noticed that the pants that I wear are fitting better three weeks in. Definitely. I've also learned that, a lot of times, that doesn't take a lot to do. A pound or two sometimes can make a huge difference in how your pants fit.

The other thing I'll say, I have a body comp scale that does do muscle and fat, although I've never trusted it because when I've compared it to the BOD POD, it's just not even close. I'm hoping maybe the trend might show up even though it might not be accurate in numbers.

Thunder: Yeah. Exactly. It doesn't have to be the same numbers as the BOD POD. It's just relative to its own baseline. So relative to the baseline you had a month ago, is the number now different on your scale? If it is, then yes, you're losing fat. The number itself almost doesn't matter. It's more of a qualitative test.

Scot: Yeah. So that's been my journey so far, guys. And I will say it's been a valuable tool for me. I really have appreciated it. I will say it can be frustrating sometimes because it does forget things. I do feel like I have to call it sometimes on stuff. One time I got in a big fight with it because I'm like, "Why do I have 900 calories for dinner? How'd you let this happen?" Completely putting the blame on the AI.

John: "This is not my fault. This is clearly your fault."

Scot: That's right. How are we going to fix this now? Sometimes I guess I just feel like yelling at something.

Thunder: It's a convenient target.

Scot: So after this, more likely, less likely to use AI if you decided to set down this pathway? I'm going to start with Mitch because I think Mitch is the most likely that would be to do this.

Mitch: I don't know. I am always hesitant these days, especially with health decisions when it comes to AI. That being said, the nutritionist I work with has suggested a handful of nutrition-specific AIs, which maybe that is worth checking out. We know these days that a bunch of AIs are good at different things. And if there are some nutrition ones out there that that's all they're trained on, maybe, just maybe, that would be helpful. But otherwise, no, I think I'm going to use what my human told me to do.

Scot: How about you, John?

John: Oh, I wouldn't hesitate to use it. I think the more familiar I become, the better the prompts you give, I think I can get some good information. But then again, like you've said, always push back. I think I'm just going to ask, "What would Thunder do?" at the end of every prompt. And I think it'll probably give me everything I need at that point.

Scot: That's brilliant. Thunder, how about you?

Thunder: Honestly, I would not use AI. Not because I'm against AI, but all the things you talked about, you described, and what you're doing, I already do for myself. I mean, just because this is my game. I'm good at opening up the fridge and looking at what's in there and coming up with something to cook. So what people struggle with, I think, is "What do I eat?" and "Give me some ideas on meal prep."

Scot: "What do I eat and how much of it do I eat?" I think is huge.

Thunder: "I've got these things staring me in the face. How do I make a palatable meal out of it?" AI is good at that, and that's what you're getting help with. So I think it's great. Now, me personally, I can look at that stuff and come up with how to make it. Because again, this is my thing, right?

Scot: Yeah. Is it something you would recommend to somebody if it wasn't their thing?

Thunder: Totally. I would go about it a slightly different way. I would say if you want to use AI to lose weight, keep track of everything you eat for three days, feed it into AI, and say, "This is my usual diet. I want to lose weight. How would you recommend modifying this diet?" And then see if the AI would say, "Hey, you eat vanilla yogurt every day. Skip that and do plain yogurt. You'll save yourself 25 grams of sugar each shot." Something like that.

I would recommend starting with that, and then use the AI to give you meal planning ideas, recipes, things like that. It's super useful.

Scot: Well, thanks for letting me share this. I hope that it's been useful to somebody listening. Again, this is not medical advice. This is just the experience of one person. You should probably consult with a health professional if you are not well versed in this sort of thing and have them offer guidance as well, perhaps a nutritionist. I do think it can be a useful tool. So if it's something you want to experiment with, I hope I gave you some ideas.

Any final thoughts before we call this one, guys?

Thunder: I'm happy we did this episode and you shared your experience with AI, because I think it's a new world with this technology, and this is a great way of using it to promote health.

John: I couldn't agree more. I think getting good information, questionable information, but always be mindful if you're listening or you're putting in a prompt and it seems like, "That seems fishy," always push back.

But I think these tools are going to be here to stay, and I think they're going to get better with time. And so the more we know how to use them . . . If we don't have someone we can call on like Thunder where we have a question and we can get an answer, I think it's good to get us started.

We may need to actually find a human, like Mitch said, at some point. But I think if you're going to get started and it's the only thing you've got available to you where you just need something to help make it easier and kind of give you . . .

It's kind of just automated it for you, Scot. I think sometimes we just need that. I just need someone to tell me, "Hey, go eat this. This will fit in your macros." I think that can be very empowering for some people, and I think it's helpful.

Scot: All right, guys. Well, thank you very much for letting me share my experience. If you have an experience with AI, using it for your health, trying to get a better nutrition plan, a better workout plan, and you'd like to share that, we'd love to hear about it at hello@thescope.com.

Thanks for listening. Thanks for caring about men's health.