it's time you ACTUALLY learn how to make money with A.I.

Right team, welcome back to the channel. My name is Tommy Youngs. This is the Whiteboard of Truth. And in today's video, we're talking about how to actually make money and help your clients make money with AI. And yes, I need you to forget everything that you think you know about how to do that already because the truth is most people who are doing this have got it completely wrong. They don't even understand what AI is. So kick it all in the bin, get rid of whatever you think you know because this is episode three of the Supercoach series. We're actually, we did S last week and we did the overview of the week before. So we're on number three. This is all about how you need an unfair advantage moving into the new era of online education because it has officially begun. This is the 2027 game plan because in 2026 and before, you could get away with being a business coach or a mentor who guided people based off vibes or voodoo. But no longer my friends, no longer. We now have to move into an era where in 2027 and beyond, it's all about data and decisions. This is what we're talking about today. So the reason why I feel like I'm somewhat qualified to talk about this is because I've literally been doing this stuff for the last couple of years. 18 months ago, I made the dedicated effort to move at the forefront of AI or move at the frontier of AI and add it into my consulting business scale 20. I've done some pretty cool things in the last couple of years because of all this stuff. I put in 1.9 million in consulting deals for my business, zero sales calls, not a single sales team member, not a setter, not a closer, nada. And I think that is pretty turbocharged, to be honest. I've had over 7.5 million wins inside scale 20. We've helped a bunch of online business coaches get to their first 100k month or their first 7 figure year, which is also incredible. Now here's the real key. We have collected right now about 199,000 data submissions, unique data submissions over the past couple of years from multi six and seven figure online business coaches. This is everything that we're talking about today. But first, let's crack a fresh bottle of South Downs natural mineral sparkling water, that is the sponsor of this episode. If you have magically ended up on this video for the first time, you've never seen my face before. Hey, I'm Tom. I run scale 20. We are an advisory in consulting business for six and multi six and seven figure online experts. The goal is to get them to their first 100k month, all the way to their first 7 figure year, all the way to 100k net profit months, different kettle of fish. I have a gift for you. If you click the link down below, it'll take you to a free seven day mini course, six to seven figures in seven days. Everything that I've done to build all this stuff in the past couple of years, to your inbox for the next seven days, it is incredible. Go click the link down below. Let's get started. Okay. I apologize because today I'm about to use some long words and I'm going to go against what all of these nerds talk about, which is speak to people in a fifth grade language or below. And I don't know about you, but I left fifth grade a long fucking time ago. I mean, I'm 38 right now. Fifth grade. I was probably just picking my nose, eating burgers. I'm beyond fifth grade. Okay. So we're going to use some long words today. I'm this my apology, some long words today. Here's the first two. Probabilistic and deterministic. I'm going to explain what those words mean today, but this is fundamentally what AI is and what AI isn't. AI, as you use it, as an LLM, as we're going to get into, is on this side. It is probabilistic. What we need to be as guides, mentors, coaches, we need to be deterministic. Okay. We're on this side. What this is about is that guessing does not grow businesses. And guess what? Probabilistic because it's based on probabilities is guesswork. So we're going to move away from guessing and we're going to move towards knowing. Okay. That's the difference. It's the equivalent of you getting on to calls with your coach. If you've had a mentor before, you know, my biggest bug bear was the first mentor that I ever invested in. We'd get on, he'd have three Q&A coaching calls or coaching calls per week Monday, Wednesday, Friday. It would be two hours of him just sat there and it would be bang, bang, bang, bang, bang, someone's question, someone's question, someone's question. Every single person had a different offer, different business, different set of problems. And I'd have to sit there for hours every week listening to other people's bullshit. Now, that, my friends, is not coaching. That is not guiding. It is not mentoring. It is fucking bullshit. It is definitely not education. So we need to do things differently now. And the cool thing is we have technology that allows us to do this. So in 2026, you could get away with that 2026 and before you could get away with coaching with just vibes and voodoo, which is I think you should do XYZ. I think you should do, blah, blah, blah. Because what that is, is that is you tapping into your experience that's floating around in your head in this little like big rock that you've got on the top of your body. And depending on the day, the mood you're in, how many hours you've had sleep the night before, what your Woot Band says your recovery was today, you might give a different answer to the same question yesterday than you do today. Then you do tomorrow. That is what probabilistic means. It is guesswork. So guess what? We don't want any guesswork. Can we get clear on this? No guesswork. We're actually going to fucking say you need to do this. That's what this is about. So moving forward in 2027 and beyond, we want to move away from vibes and voodoo and we're moving towards data and decisions, okay? Data and decisions. What this looks like is instead of saying I think you should do XYZ, it is the next gate is X and you need to do Y to get there, okay? The next gate is X and you need to do Y to get there. And I explained today what I mean by gates and how this all operates and fits together as you helping your clients. Now to the nuts and bolts of what actually is AI as you use it. So I mean, I feel like I'm a turbo, not a turbo user, but like I'm at the high end of like a user of AI as you know, Claude, ChatGPT, Gemini if you haven't got any friends. You know, interestingly, Gemini has the best model that is currently is not released, but let me explain. So probabilistic is what large language models are. So the LLMs that you use every single day that you've used for the last three years, the ChatGPTs, the Claude's, like I said, the Gemini's if you're you ain't got any friends. These large language models, their performance is based on two parameters. The parameters are hallucination and accuracy. So we have hallucination going up and down and we have accuracy going left to right, okay? And now you can see here I've got a couple of dots, a green dot and an orange dot. The orange dot is today's Claude Fable 5.1 model. This is as the today is filmed. That's what this is. It's this specific model. And then this is ChatGPTs. I don't use ChatGPT that much. So I'm a Claude person. This is Claude's, this is ChatGPTs sole 5.2 I think. So let me explain why they're at on this chart. You've heard of the idea of hallucination. You've heard of hallucination. What this is. I want you to imagine a classroom of 30 kids and they've got two options to answer a question that the teacher has asked them. They could either say that they don't know the answer, that's one option, or they could say they can make a guess. Even if they don't know the answer, they're going to make a guess. Hallucination, as it pertains to large language models, is all about how much do these models guess. How much do they guess? And here's the fucking nuts thing. They guess an absolute shit ton. Claude, as of today, so Claude's 5.1 Fable model, currently guesses 72% of the time. 72% of the time it is guessing an answer. chat GPT Seoul is. 51% of the time it is guessing, so it's quite significantly lower than the fable model. Now when it comes to the other side of this, you have accuracy from left to right. We clearly want our models to have high accuracy. What this is, is this is a standardized test across 42 different categories where these models get asked a series of questions, and however accurate their answers are, depends on how accurate they are. So the more accurate the better, right? So we want them to be really accurate. So right now the Claude model, fable 5.1, it is 67% accurate, accurate 67% of the time. ChatGPT is accurate 62% of the time. Now interestingly, the new like Gemini argon model, which hasn't been or argon model, but that hasn't been released, it's actually down here. So it's actually got really low hallucination, which is good. So we want low hallucination and we want high accuracy, it's actually got really low hallucination and high accuracy. So it's actually technically theoretically the best, but it hasn't been released yet. So imagine you're in that classroom, you've got 30 kids there, and everyone just feels bullsy enough and cheeky enough just to guess 70% of the time they're just going to be like, they're going to blur out and answer even if they don't have a clue what the answer is, they're just going to blur something out. That is currently the large language models that we're dealing with. So now when most business coaches are putting all of their like playbooks and SOPs and shit and all of their video transcripts into a into a knowledge base somewhere, and then they're just like giving that to their members, like which is frankly transparently what I did two years ago or a year and a half ago, actually two years ago, I gave my first version of Tom AI, which was just a chat GPT with my knowledge base and I gave it to them. What that is is like you give them all this all this knowledge, but then it's just going to guess like most of the time it's going to guess. So what this does is an increase in hallucination, which is what we're doing if we're just using LLMs like that, we're doing educated guesswork. So when business is simply a numbers game, business is sell shit minus how much it costs to sell the shit and how much it costs to deliver the shit and what you're left with is profit. So there's business in a sentence, okay, and it's a math game, it's a maths game. And so it really is like super important not to just be guessing. So most people most business coaches and mentors are just stuck right now guessing because they're using these large, large language models like a monkey. And so what we want to do instead is we want to start using these large language models in a deterministic fashion, but we have to use them differently and the stuff over here that I'm going to talk about in a second is all of the stuff that you need to do to be able to use them in a deterministic fashion and for you to be able to actually make loads of money with them, help your members and clients move faster than ever before. So what we want to do is we want to move from probabilistic, which is educated guesswork to deterministic, which is conditional logic. So what does conditional logic mean? So where is on the left hand side here, probabilistic systems are large language models, which are educated guesswork, deterministic systems are like calculators. So for example, you know that when you add on any, you can grab a calculator anywhere in the world and you put two plus two and you know for a damn sure that that calculator is going to say equals four, okay. You know you can go to any calculator and you could put three times three and that's going to be equal 12. You could go do this, right? And make sure you this is the difference between probabilistic and deterministic. For those of you listening by the way, that was a gag. I do know that three times three is nine, not 12, but there you go. So what this is about is conditional logic. You may have heard of this thing called like this or this then that. If this then do that. It is how calculators work. So if this happens, do this next. And I want to give an example here, it looks a little bit scrappy, I thought this is going to look a little bit better when I try it out, but it looks like shit. If you are in a situation and you want to achieve like a result, if you want to achieve a result, you may go one of multiple routes based on your situation. So let me give this example here. So I gave a playbook to my members the other day called the furnace. So it's a turbocharged playbook that I've been running for the last three years, which is it takes leads from being ice cold to pre-sold within seven days and it's insane. So it's based on whether you already have a core sales asset that I helped my members inside scale 20 build. And so when I was delivering this playbook the other day, I said to them, I was like, okay, you are either going to be in camp A or camp B. Camp A is you have already made this asset that you need or you haven't made this asset that you need. You are lit, it's, it's A will be. There's no middle ground. It is literally you're, you're here or you're here. You need to do it. If they, if they're in the process of building it, they're over here. They haven't built it yet, okay? So there's, there's A or B. So I said to them, if you are in camp A, you are going to do X, Y, Z in this sequence. If you are in camp B, you are going to do X, Y, Z in this sequence in a different sequence. So this is, this is what conditional logic is, if this, then that. And that's the thing that's missing with large language models with probabilistic systems and large language models, they are not, they do not have and contain conditional logic. They are guesswork. It is just guessing an answer based on like what is, what is, what is most likely to come up next. And so this is what we want to do, we want to use conditional logic if this, then that. Now I now have built Tom AI, we've had it for two years and we've gone through evolutions of it. And I'll explain at the end of this video, like what the current version looks like and how you can build it yourself. The first version when it was just a knowledge base, Tom AI wrapped in a chat GPT knowledge base, custom GPT, given to my members, now our Tom AI's are completely personalized. Every single Tom AI has context to the individual person. So we build these things called context rooms, just like how businesses would create a data room when they're looking to sell a company or get acquired. They would create a data room where they literally put all their data into a Google Drive or a Dropbox folder. That's how this worked. And I've created the same thing within my business intelligence platform. So when we onboard a new member in scale 20, we get them to do this very comprehensive context room dump. It's got numbers, context about them, where they've been before, what their pricing changes have been over the last couple of years, all of their numbers for the last 12 months, their team, who's in their team, what they're paying every single team member, all this context. And we build this context room for them. Their Tom AI's also have access. Every time they submit monthly totals to us, have access to their monthly numbers, all of their numbers in real time. We've also connected them to their content, so it has all their social media numbers. It has all of this context about them. And then also has access to our number, our full database, so the database of all the numbers they were getting from all of our members. Then when you layer on top, the conditional logic that we have inside our playbooks, that's when these things, these LLMs, become turbocharged, they turn into deterministic systems or at least get us really close to it. And I'll be honest, like I've been using this shit a lot over the last like a few months and it is crazy how good the recommendations are. We now have a recommendations engine for our members when they start with us. They get a custom game plan for the first 90 days. We put all their numbers and all their context into our engine. We build this customized roadmap for them and it's incredible. We now do this also for QBR, so we take all of our members data for the last like three months, compare it to where they've been on their entire journey working with us and then help them understand which two key projects they need to focus on for the next quarter. One to make their business bigger, one to make their business better. And you know, at the end of the day, what our job is here, our job is to help port clarity, confidence and certainty into our clients' brains. And that's the cool thing about what we've built here. And one of the awesome, also cool things is Slack questions, so questions to Slack have dropped 56% year-on-year from Q3 last year to Q3 this year. We've had like Tom AI is now dealing with 40 times more volume than Slack is. We've had like in Q3's 3000 questions, absolutely incredible. So that's the foundation, right? So you've got to understand actually what AI is and what the limitations are. the limitation. of an AI model as you just get them out of the box. The chat GBT's, the fate, you know, the chords, the germanize if you haven't got any friends. You know, those things are limited by the fact that they are just educated guessworks. We need to make the shift from educator guesswork to conditional logic. Let's get into it. Okay. Let's talk about project moneyball. This is all about benchmarks, beat guesswork. So you may or may not be familiar with the film moneyball. This is a film probably like 15 years ago now. I remember seeing it when I was like on holiday in Seattle, when I just dipped in and watched it. And it was a film with Brad Pitt who played this character called Billy Bean who was a scout for a baseball team. Or he may be a manager, but I think he was like, I think he was a maybe a manager, yeah. And he was a manager for the Oakland athletics, Oakland days, who the fuck knows. And then his little side kick that he had was Guy was Jonah Hill, his character. And Jonah Hill was an analyst. And this guy Billy Bean was like losing, losing, losing, losing, losing. And he had a small budget compared to all of the other teams. So I think his budget for the Oakland days, this is a true story by the way. He had like a total budget of like $40 million to spend on players, everything. But all the big teams had like $120 million. It's $120 million. So like three, four times more. And he, this Jonah Hill came to him one day and was like, dude, you don't need to just buy the most expensive players. You need to buy these players because these players have the certain skills based on their statistics. That if you stack all of these players together with all of their different skills, you will win games. And the whole idea was that they used, all of this data that they had to go find all of these players that were, they weren't the top players, but they had like little like, they had statistics that were extremely favorable that when you stack to all of them together and you put everyone in a team, they then went on this unbeaten run. Like I think the biggest unbeaten run in like 100 plus years with a tiny budget. And what instead of what was going on, which was like, they, the scouts who were sat these old dudes who were there, like literally just trying to find new players. But all they would say is like, yeah, this guy's, this guy swings a well. This guy, he's got a good face. Yeah, this guy's, you know, he's got a hot girlfriend. So he's probably a big hitter. They would literally go off just like vibes and voodoo. And then Jonah Hill came, comes along and he speaks to Brad Pitt's character and was like, hey, dude, I think you can do it a different way. And he's a super nerdy character and then Billy's like, well, fuck it. I don't, I don't, I don't have any other option here 'cause we're getting like new to every single game. And so they switched the question they were asking. Instead of him and the scouts who are looking for these players asking the question, who is the best player? Because they couldn't ask, they couldn't answer that question because they didn't have the money to buy the best player. So instead of asking who is the best player, they asked the question, what actually wins games? What actually wins games? And that's when they dove into all of the statistics and realized that like there was this gate, this level, this line that above this line with different statistics, they could hire, they could basically hire these people at below the cost that would usually cost them to buy a player who had all of these statistics, they could buy these players and they ended up winning a load of games. And so at the start of Q3 this year, so a few 12 weeks ago or 16 weeks ago whenever we, you know, 60 weeks ago now, we started a project called Project Moneyball and this was my project actually to make the business better. We have projects to make the business bigger, projects to make the business better. This was my project to make the business better. One of my members and my clients to move faster than ever get better results than ever. What I did is I spent probably like 40 hours over the course of the quarter looking at one single spreadsheet and I built the ultimate scorecard. There was a few couple of weeks ago I literally spent probably like 16 hours across two days just working on this one spreadsheet. And what we now have is we have the ultimate scorecard which is 39 key performance indicators. Oh, it's from Speak. So we built the ultimate scorecard with 39 key performance indicators with every single metric has a green, amber and red gate. So a gate is quite simple. It's a, you gotta get through the gate to get to the next thing. So every single metric has a gate in between it. So the idea is obviously you go from red through to amber through to green and you end up in a green. And every single metric now we have has a red benchmark, amber benchmark, green benchmark or a gate. So red amber green gates with the benchmark eventually being the green number. We want people to end up with a green number. I'm gonna give you some examples of all that means. But remember, this is the difference between saying like I think you should do this versus the next gate is x and you need to do y to get there. Okay, it's the difference between, I think you should do this and the next gate is x and you need to do y to get there. Okay, that's the difference. Hopefully you can see this. That's the difference between coaching off vibes and voodoo and data and decisions, okay? So we've got, you understand gates and benchmarks. So how do we strap all this shit together so that you end up with an unfair advantage? Because this is what this is all about. Now the information is valueless and I can get the answer to everything through chat GPT or Claude in three seconds. What carries value now? What carries value now is your numbers. Your data, your data, your numbers is the mode. Fucking write this down. Everyone's got the same AI, but nobody's got your numbers, okay? That is what the game is all about now. And I'm gonna say this to someone who doesn't, I'm not a number's gun, not a spreadsheet's dude by the way, but I understand the importance of it and building a dataset. That's why we've been so trying to be so far ahead of the curve with this. Okay, so now you know to build your unfair advantage, which is the you in the supercoach series that we're on. The first thing you need to do and this goes without saying is you need to collect some fucking data, all right? You need to collect some data. Now this is like the record, the record of all the stuff that's gone on. So when Brad Pitt and Jonah Hill were looking at all these statistics, they were looking at the record of, the records that were all going on like, okay, what were the KPIs? What were the benchmarks? What were the gates that they could build? Now to start off, you know, you can just use like a Google sheet or a notion database. This is what I did a couple of years ago. This is how I first started helping my members. Then we graduated to a table, so a more sophisticated database system. But now we are in pro leagues, we are using superbase, maybe using superbase for the last year. And so that's the first thing you need. You need to collect data and ideally, you need to be doing it in a sophisticated way. The second thing you need to do is you need the gates, you need benchmarks, okay? So what you need is a red, amber and green for every single metric that is a key performance indicator that you know if you help your members and clients get closer towards a benchmark, the green, they will win eventually. It's giving people a clear sight and saying, let's aim here, let's figure out what we need to do to get over here. If you've been following me for a while and you've been watching my content, you will know the of the nearly 200,000 unique data submissions that we've had from multi-six and seven figure online business coaches, there is one metric that stands above all others. It is the king metric of leverage. It is the metric that is most associated with cash in your bank, cash in your pocket. And that metric is time under attention, long form content view hours. We've analyzed over 200,000 data submissions. We've seen it. The clients are members who have the most long form content view hours, they have the most cash in their pockets at the end of the month, okay? So, I bet if I ask you right now, and I asked you how many long form content view hours did you get last month? I bet you wouldn't have a fucking clue. If you do, please indulge me and drop it in the comments. Tell me how many view hours that you got on your channel last month. Now, I keep in mind, I have just told you, The time under attention, long form content view hours is the single most important metric that you could optimize for in your business, and we've got the data to prove that. And it's everything that I've built in my business has been to optimize for that one number. And you probably don't have a clue how many you got last month. Now I want the penny to drop here like are you acting like a fucking professional or not? You're probably not like to I want you know the reality is you're probably not but like if you want to scale to a million dollars a year and above, you need to take this shit seriously. Business is a numbers game. You need to know your numbers. So let me score you here. You don't know how many long form content view hours you got last month, but I'll tell you what what good is, okay? So if you've got between 500 to 2000 view hours for a long form content last month, you're doing okay. You're not quite leveraged yet, you're doing okay. If you've got above 2000 long form content view hours, you're doing good. You're in the leverage phase, okay. You're in the leverage payoff. You've got there. You're probably getting inbound DMs every single day, hopefully getting qualified inbound DMs every day. You're probably getting people sending you the message, I'm in, can you send me the payment link? You're probably not having to convince people in the DMs and you may have even been able to stop taking sales calls. So I've just given you a clear goal. Get to 2000 long form content view hours per month and you will have those magical things that I've just said, okay? Now one of my members Richard, so I've been working with him for literally about about two year anniversary and we've helped him get to over 100 grand a month and we've really been focused on staging, phasing his business. So last year was all about short form content and making sure that he got the skills of short form content, he's now absolutely elite at short form content, getting, growing his audience, getting leads, that was the focus for last year. This year it's been more of a focus on educating his existing audience, developing depth, trust, building that authority with long form content. And he'd been struggling because he'd been kind of like stuck in like three, four hundred view jail on YouTube for months and months and I filmed a demo, over the shoulder demo of exactly how I increased my time under attention by, I think it was like 2000 plus, it was like 26 or seven times, I can't remember the exact numbers. But in Q3, I turbocharged my time under attention. I filmed a demo video, showed it to Richard, he installed it the next day. His next video was a 117x outlier, it got, it's got now 46,000 views. And if you just assume that like an average view for that video is four minutes, so most people that watch is going to be the average view duration is four minutes, that is 3000 view hours from one single video. So as you can tell, he's happy because he knows that he's hit that gate for this month. And that, when it comes back to our job is all about giving our members clarity, certainty and confidence, how clear, how certain and how confident do you think Richard feels about where his business has headed? Pretty fucking confident, pretty fucking clear, pretty fucking certain. The final piece of the puzzle here is context. Now I mentioned to you earlier about how when we start working with new members of scale company, we build a context room, but what makes up that context room? There are three things that make up that context room. The first is quantitative data, so this is the numbers. So this is what was the last 12 months of revenue? What were, what offer, what is your offer priced at right now? All of the content, social media performance, how big is your email list? How many, how many subs on your YouTube channel? How many long form content view hours did you get last month? All of the numbers. The second piece of the puzzle is qualitative data. This is the stuff that isn't the numbers. This is the stuff that you can only explain by with using like words. So this is like, do you have X built, yes or no? So it's either yes or no, it's not a number answer. The final piece of the puzzle is we have started to add a psychographic analysis to our members Tom A.I.'s. Now we have a couple of months ago, we brought on a performance and mindset psychologist into scale 20. When our members get to their second month, they get a courtesy one on one call with Jay, who is our psychologist who does a breakdown, a psychographic breakdown of our members. So we understand their limiting beliefs, their core operating system, their default pattern behaviors, and like literally all the bullshit that's going to get in the way for that. And so we get this analysis, we add all of that as well into their Tom A.I.'s. So their Tom A.I.'s have incredible context about them. And then when you add all this stuff together and you build conditional logic in the way that you make recommendations to people with your SOPs and playbooks, this is the most important thing. You have to be thinking every single time now, every time you make an SOP, a training, a demo, a guide, you have to build conditional logic into it. If this, then that. If you've done all these things, that is when you will build an unfair advantage and you will have the opportunity to prosper in this new era. Okay, that was sick. And yeah, okay, so that was number two or three, we're on three. The next one in the series is going to be all about proximity being the premium, okay? Now if that video is out, you can go watch it here next. If it's not, go watch this video next instead. Right, that was sick. We'll see you in the next one.