Stop rolling the dice with AI. Every second you waste is the second someone else is making the right move and winning. Because the difference between anyone and the people making money with AI is not brain, it's moves. They make the same two to three plays every single time and they don't guess. Because you see, using AI is literally like chess. A beginner moves one piece at a time and takes ages to think. A grandmaster sees the whole board and plans three step ahead. And today I'm giving you those moves so that you can win the AI game. So what does that mean? I am literally going to break down the three prompts that have made the biggest impact in my business and in the businesses of so many entrepreneurs that I've worked with. And for each one, I am going to give you an actual example that you can screenshot and try today so that you can start using AI like the top 1%. And by the end of this, you will have a toolkit that makes AI feel less like guesswork and more like strategy so that you can get faster, better results no matter what you're working on. Number one, accuracy and truth prompt. because most people treat AI like a slot machine. You know, you type a question, you pull the lever and then take whatever comes out. But the problem is that AI is confident even when it's wrong. And if you're making important or maybe even business decisions off guesses that are dressed up as facts, then you're playing a dangerous game. And that is why the first prompt, which I call the accuracy and truth prompt, is about turning AI from a yesmen into your factual accuracy partner. And this way you make sure that it delivers actual, verified, and balanced responses while being transparent about uncertainty or missing knowledge. And I'm sure you also have found that large language models are confident liars when they're unguided. So by teaching AI to doublech checkck itself and site credible sources and expose blind uh spots, you instantly increase output uh trustworthiness, I guess, because truth compounds every accurate decision today is going to save you from 10 bad ones tomorrow. And the move that I recommend is a mix of the CCC framework layered with my self-verification structure. So let me explain. The CCC framework obviously consists in three C's. Number one, context. You want to set the scene before you ask. So you want to define the background and the intended use. Number two, constraints. You want to define source type or date ranges or scope limits. And then number three, criteria. You want to specify what makes the answer good, like separating facts from inferences and citing sources and giving a confidence score. I will explain a bit more about this one. And then what you want to do is layer it with my self-verification structure so that AI checks its own work before you read it. Because you never want to be in a situation where you ask AI for help, maybe to draft the client proposal, for example, which sounds flawless until the client points out that one key fact is wrong because that mistake could cost you credibility but also the deal. So, the self-verification structure is like giving your AI a built-in quality control department. And instead of you having to read and answer and wonder, oh, is this actually true, the AI is forced to audit itself before you even see the final verdict. So, basically, it's going to check whether your starting question is sound and then separate hard facts from assumptions and then site credible sources with dates and rate its own confidence and offer counterpoints and flag potential biases. So this is going to improve accuracy and it is going to make the thinking process super transparent because you will also see how the answer was formed, not just the polished output. And when you're making business or strategic or financial decisions, that transparency is becoming priceless because it's the difference between being handed a mystery box and being given the blueprint for exactly what is inside. So here is my prompt. Be my factual accuracy and truth check partner. If you're missing information from my side, please ask me. Then you go into context and you give it your goal, your audience, your time frame, your constraints. So tell it only use these sources or dated after this year or later. And if you don't find any, just say so. Then criteria, you can tell it to separate facts from inferences to site sources with publication date and to give confidence ratings. And then use this structure. One, direct answer. Two, premise check. Three, fact versus inference. Four, evidence and sources. Five, confidence from 1 to 10 and why. Six, counterpoints. Seven, bias check. Eight, final verdict. So, we're going to put here on the screen an example of an actual prompt that we could use or have used. And then you can screenshot this one and use it as insight or as an example for you to work from. Now, this is a long one, right? So, it's a full prompt and I think it's gold standard. it is your grandmaster move, but sometimes you don't need the entire chessboard. You just need a quick tactical play. So, I'm going to give you shorter versions like quick variations. So, um you can use these short, easy to remember, perfect for when you're working fast, but you still want accuracy. So, for example, um you can ask for factchecking the answer. Maybe you've already gotten an AI response, I don't know, from your Claude or Chad GPT. What you can do is paste it back and say, "Please review this answer for factual accuracy." highlight any areas that you're unsure about and suggest where I should fact check further. And this is going to instantly add a second layer of scrutiny that you don't have to do yourself. Now, second one is asking for multiple perspectives. I do that a lot because if you suspect that there is more than one valid take to something. You can try prompts like give me at least two credible perspectives on whatever topic citing reliable sources for each. And this is going to be great for any market analysis or strategy debates that you have. Then third um this one is um to rate their confidence because sometimes you just need to know how sure are you. Okay. So after the answer you can ask your AI to rate their confidence from 1 to 10 and explain to you why because this forces the AI to surface any uncertainty instead of putting it under the cover or the carpet. And finally bias detector because every answer has a perspective that's baked in even if it's not obvious. So what you want to do is always ask your LLM to identify any possible bias in your reasoning and how the answer might change from another culture or political or industry perspective. And this is going to be perfect for anything that you create that is public facing. So literally you can think of these like your opening moves in chess because they're quick, they're efficient, they're easy to execute when you don't have time for the full game plan. Now look, if you've been following along and you like what the CCC method does for accuracy, you are going to want to go much deeper with this. HubSpot has put together a free resource that is called advanced chat GPT prompt engineering and it is basically a playbook for making your prompts sharper, your outputs more accurate and your AI more reliable in high stakes situations. And trust me, it does not stop at the CCC method. In fact, you are going to find so many more frameworks there. Also, they have token economics and token optimization techniques so that you can get the maximum depth and detail in your responses without wasting processing power. You're going to learn, for example, about the format framework, which is a simple but super super powerful structure that makes your prompts more error resistant and easier for the AI to follow. Exactly. And my favorite is the modular prompt architecture, which is a way to design flexible prompts like building blocks. And then you can mix and match them for different use cases without starting from scratch. Look, we have been using these techniques in our business for market validation, for risk assessment, even for competitive analysis. And the difference is night and day. And the good part is that this is completely free and it'll make your AI work like truth first biasaware research partner instead of a smooth talking guess machine. So, hit the link below to grab your own advanced chat GPT prompt engineering guide from HubSpot. And also big thanks to them for sponsoring this video and for making this resource available to all of us. Now, prompt number two that I want to share with you today is about your creativity and brainstorming because most people open chat GPT and type give me some ideas and then get exactly what everyone else gets because they ask the same question that everybody else is asking. The problem is that AI defaults to safe and predictable and average in all honesty and average doesn't get attention. It doesn't sell products and it doesn't build an audience. What you want is an idea factory because creativity isn't magic. It's the byproduct of seeing the same thing as everyone else, but noticing what others are ignoring. And if you want unique answers, you have to feed AI unique perspectives and constraints and role reversals and maybe unlikely combinations because I think the real difference is training your brain as well as AI to think like a lateral thinker. Read the book if you haven't. You want to connect the dots that no one else is connecting. And Edward Dono mentions in the six thinking hats that the magic happens when you deliberately force yourself and the AI to see the problem from multiple angles. The wild, the cautious, the analytical, the visionary. Because innovation is, I think, pattern recognition applied in a new domain. The more patterns you collect, whether it's from art or science or business, even hobbies, the more raw material you have to remix into something new, because originality is a habit. If you can't think differently, you will compete with everyone who can. And what I found is that a good prompt doesn't need to tell AI what to do. It needs to show it how to think. And the highest leverage isn't in the output. It's in training the AI, but also yourself to see the problem from angles that no one else considers. That's when you you basically stop competing and you start creating. So here is my prompt structure. You would start with the role. So you assign AI a specific creative identity and then give it constraints. You want to set unusual boundaries that force it to think different and deeper. And then give it contrast. You want to mash your problem with something that's completely unrelated like open territory. Okay. So for example, you can say you are, I don't know, a marketing specialist tasked with generating creative solutions for this problem. I don't know reducing ad spend. Here are your rules. Each idea must combine and give it industry and concept with an unrelated field. Include one wildcard idea that seems impossible but could work. And for each idea, describe it in two sentences and explain the unique twist. Now, you can also change the role in constraints and refresh the output and you're going to get very different, wildly surprising, sometimes answers. And we're going to put here on the screen an example of this prompt that you could take as inspiration. You can screenshot it. You can add it to your library of prompts and come back to it, but also ideally make it into a habit, use it frequently, try it out, see how it turns out for you, because you will start using it a lot better, a lot more when you will have had your practice time with it. Now, if you don't have time for this entire prompt, you can try these quick variations. For idea stretch, you can say, "Give me 10 ideas for this topic and then push past the obvious by giving me five more that are unconventional or surprising or counterintuitive and explain why they might work." Or you could do a roll swap. You can say, "Pretend you are this famous innovator or fictional character. How would you approach this challenge or this problem differently than most people?" Or do a mashup. This is my favorite. I learned it from J. Abraham. You want to combine two unrelated industries or concepts, maybe industry A and industry B to generate X unique ideas that could be applied in whatever situation you're in. And then number four, you can try constraintdriven creativity. So you can say something like, give me five innovative solutions for this problem that cost under $500 and can be done in a week or a day or a month. Now constraints are not limitations. Um they are going to fuel the creativity. So just asking AI for ideas is literally like going to a restaurant and asking for food. You are going to get something but probably not what you really wanted. The magic happens when you frame the challenge with playfulness but also with creativity yourself. Now I want to share a story with you because this entire video was inspired by a conversation that I had a few months ago with my son as he was coming back from a chess camp. So when he came back, I asked him how was it. And he said something that really struck me because he said, "The Grandmaster always wins, but he doesn't always make the same moves." And that lit up the light bulb because every game was different. And the grandmaster saw the whole board and picked the best move for the exact moment. And I think that's exactly what these prompt categories or these prompts are going to give you. They are not scripts that you want to memorize. They are types of moves. There are patterns that you can adapt to in any situation so that you can win no matter what the board looks like. I think this is called specific knowledge if I'm not mistaken in uh Navarikans almanac. It's the kind that you can't just copy from a template. And once you understand these categories, you can combine them in ways that no one else would think of. So here is my third and final one. This is about giving your AI memory because most people treat every AI chat like it's their first date. They start from zero every single time. But the thing is without continuity AI forgets everything you've worked on together. I mean not so much now anymore but it used to quite consistently. And if that happens you basically lose the momentum. You you lose the context and the shared brain effect that makes it feel like that real partner. The thing is AI doesn't have real memory yet. But you can give it memory by helping it a little bit by creating a loop of context that carries forward session to session. This is the illusion of working with a long-term collaborator. And continuity is not about convenience. It's about leverage because leverage is the ability to apply your past effort to future results. And literally, this is what we're doing here. Every recap you create is going to compound over time. It's like building intellectual equity. Each conversation invests in the next one. Instead of starting from scratch, you're starting from a stronger and sharper and more informed version of your AI every time. So, what you want to do here is train both yourself and your AI to think in context. you are not uh looking to give it instructions. You are building a shared playbook. And the more the AI remembers your style or your goals or your preferences, the faster it becomes a catalyst for your amplification. So here is my prompt. You want to start with a recap, right? So at the end of a session, you can have your AI summarize the key goals, the key decisions, the key style cues, and then use that to inject it. So, at the start of a new session, you want to feed that recap back to the AI uh almost like a like an onboarding before asking your next uh task. So, here's the prompt that I use and that you can use as well. Before answering, restate our current project goals and constraints from earlier. Then continue with whatever task you've given it. At the end, update the project summary with any new information we decided on today. And basically, as I said, you can think of these like um your onboarding checklist. It only takes you 30 to 60 seconds to copy paste, but it's going to buy you back hours of context setting and creating the illusion of a partner who never forgets. And if you want quick variations, for example, what I sometimes use as a session recap prompt, I say something like, "Summarize what we worked on today in bullet points, keeping the style, the goals, and the tone we used. Include key decisions, preferences, and examples so I can paste this into our next session for continuity." And then if you want to do that as your context injection, you can use um this at the start of your new session. Here is the recap of our last conversation. And then you paste your recap. Use it as your background. So you can respond in the same style and continue where we left off. And then what I also do a lot uh when it comes to style and voice, I use something along the lines of this is how I like my outputs to sound. And I give it an example for all future answers in this thread. Match this tone, this vocabulary, and this formatting. And then you can also use it as your project tracker uh with something like for this ongoing project create an update a table of tracking goals, deadlines, decisions that we made and the next steps and present it each time we talk about this project. I mean you can carry forward yesterday's clarity into today's actions and you move faster than 99% of people. Context is going to give you massive efficiency but also massive compound intelligence. Look, I have given you my most powerful prompts and I think they will make a huge difference. But these alone won't change your game. You need to turn them into habits. So, pick one and go and practice with it and come back and let me know in the comments how it went for you, how it helped you, how it impacted maybe your business or maybe your life. And I would love to learn as much as possible from your experience. And if you want to share that with a lot more people, you are always welcome to come and join our free community. The link is going to be here as well as down in the description below. and you will be able to be surrounded by over 10,000 people who are on the exact path of learning and growing with AI like yourself. So, you are going to have a lot of fun. We have free AI challenges over there every two weeks. We have calls every two weeks where you can come and join us and we can discuss about our latest discoveries and the latest insights and you can share and ask questions and it's a lot of fun. I love those calls and I'd love to see you there as well. Until next time, like this video if you did. Be sure to subscribe if you haven't done so and also share it with anyone in your circle of friends or family who is passionate about learning more about how to improve their prompting. Until next time, I suggest you go ahead and watch this video over here and I'll see you soon. Bye.