You’re Not Behind (Yet): How to Learn AI in 19 Minutes

In this video, we're going to talk through the five-phase process that you can follow to become fluent, native, super productive using AI in about 3 months. Increasingly, business owners are genuinely making decisions about who to hire, who to fire, and who to promote based on their level of AI fluency. And if you do, in fact, happen to own your own business, there is also a widening gap between businesses that are using AI properly, and businesses that are not. So, this video is split up into five distinct phases. There's timestamps for everything down below. And let's get

into it. Phase one is building your foundations. And this is going to be week number one. Now, before you try and do anything clever with AI, we just need to make sure that we have the habits and tools appropriately set up for us to actually remember to actually use it. I consider these things non-negotiables for everyone in my team. And so, if you skip them, I think your life will be a lot harder than it needs to be. Foundation number one is use AI for everything that you were initially thinking of using Google for. My personal preference is Claude by Anthropic for most things, but you can use Chad GPD, you can use Grock, you can use Claude, you can use Gemini, whatever seems reasonable, whatever you already have access to. Foundation number two is always have that particular website open in a pinned tab because the habit we want to be developing is basically always having some kind of AI chat window open on your screen or whatever with literally everything you're doing. Foundation number three is to use your voice rather than your keyboard. There are also things like Whisper Flow and the built-in dictation on Windows and Mac. And also, there's now built-in dictation into most of these AI tools. And you will find that if you are speaking to the AI, you can generally speak a lot faster and you can generally ramble a lot more compared to if you are trying to type. And honestly, this one habit will absolutely transform how much value you get from AI. So, if you're not using it already, then feel free to get started. Foundation number four is to download the mobile apps. This means you can use AI again as a Google replacement or as a thinking buddy wherever you are in the world because you always have your phone on you. You can use it while walking. commuting. You can use it while on the sofa. If you're on the toilet, you can talk to the AI. If you're in bed and you're one of those people that brings the phone into bed with you. I don't recommend it, but you can talk to the AI in bed if you really want to. And then foundation number five is to automatically record your online and ideally as well your in-person meetings. So, you probably do Zoom calls or Google Meets or that kind of thing. And there are loads of free AI tools that let you automatically record and transcribe those Zoom calls. The one we've been using for the last 5 years is called Grain. And if you want something free, then Fathom is usually one that I recommend. Okay, so that is it for week number one of learning how to use AI, getting these five foundations in place. If you're watching this and you haven't done any of these five things, 100% I'd recommend pausing the video, doing those things, and then coming back because all of the rest is going to build on this foundation. Phase two is using AI as

your coach. And this is going to be week number two. Now, this is where we're actually going to start using AI in a meaningful way beyond just being a Google replacement. But the key thing to know is we are not yet asking AI to do our work for us. We're asking it to help us think better about the work that we are already doing ourselves. For example, on my team, we've got Nicole, who's our social media manager responsible for growing my Instagram. Now, Nicole could go to Chad or Claude or whatever, and she could say, "I am a social media manager tasked with growing an Instagram profile from 1 million followers to 1. 2 million followers in the next 90 days. The account belongs to productivity YouTuber Aliabadal. What are the highest leverage things I should focus on? What mistakes do you see people in my role commonly make? What questions should I be asking my manager to make sure I'm set up for success? " Or for example, we have Gio, who's one of my team members who leads our student success for our lifestyle business academy, which is our like online business mentorship program. And she could say, "I run student success for a high ticket business mentorship program. " Currently, the biggest thing our students are struggling with is defining their niche and coming up with a reasonable offer within a twoe period. We find that a lot of them tend to overthink and overanalyze before taking action. How could I be thinking about how to solve this particular problem? And then to use an example from my own life as the business owner, I might go to Claude and say something like, "My goal for 2026 is to grow our business's revenue from $5 million to $10 million, and I think the biggest lever we have for that is our new lifestyle business academy product. Can you interview me, ask me a bunch of questions, and help me figure out what are the key levers I should do as it relates to annual planning and quarterly planning for 2026? " And so if you don't have your own like coach that's helping you with your job or a business or even if you do like I've got a couple of coaches that I work with, it's still very helpful to use the AI as a kind of thought buddy/coach to be able to ask you questions that can then help you come up with insights that can help improve your performance at your job or your business. Now this is where the fact that we are recording all of our calls then also really helps. For example, you could have a team meeting, meeting with your manager, direct report and you could take the transcript of that call and you could ask the AI to give you insights based on that call. So, for example, Nicole, who's in charge of my Instagram, could say, "This is a recording of a conversation I had with my manager, Angus, where he was coaching me on how I can be thinking about our Instagram strategy better. Based on this conversation, can you suggest a curriculum for me to follow over the next 2 weeks to improve my skills? " That would be a totally reasonable thing to do. If I've done a coaching session with our students in the Lifestyle Business Academy, the whole thing is recorded and transcribed so I can then chuck it into AI and I can say, "This was a coaching session that I ran for my students in my lifestyle business academy. Based on the transcript, I want you to tease out the key themes that came up, the key struggles that the students were struggling with, so that I can use it to help improve our core curriculum, and while you're there, please do give me feedback on my own teaching style and any blind spots that you notice. Another really useful prompt is you can literally ask the AI to interview you about your job. You could say something like, "I want you to interview me about what I actually do in my role and help me identify what's high leverage and what's probably a waste of time. " And I guarantee that if you just use that super simple prompt with literally any job that you have, you could probably find ways to actually just do a better job and waste less time doing things that really don't move the needle. Oh, by the way, if any of this stuff is confusing, don't worry. We've got a link down below to a totally free Google Doc, which is like an AI getting started curriculum that you can just download, copy into your own Google Drive, and you can just like follow it along if you like. Now, obviously, we want to give the caveat around AI limitations. Yes, the AI is cool. It's really helpful to have as a thought partner. But the way that I think about the AI tools is that it's sort of like having a very smart colleague who reads a lot of books, but who doesn't have much context on anything other than the knowledge that they've gotten from books. And so it's very useful to be able to talk to that person to kind of mirror your own thoughts or ask you questions or help you think deeper about something, but I would be very careful about taking its advice and treating that advice as gospel. Like you really want to make sure that the advice that you actually agree with the advice rather than just blindly following what it says. So, by the end of week two, if you're following along with this method, you should hopefully have by this point the habit of turning to AI whenever you're stuck with anything in your personal or professional life. And also, even if you're not stuck, just as a way of optimizing your performance for whatever goals you want to work towards even more. Oh, by the way, if you have made it this far into the video and you really want to nail your foundational understanding of AI rather than just knowing how to use the tools, then one way I found super helpful for doing this is Brilliant, who are very kindly sponsoring this video. I have been using and loving Brilliant since like 2019. And what I love about the product is that it helps you get way better at maths and coding and computer science through step-by-step interactive lessons and personalized practice where you genuinely learn by doing. Brilliant has helped me get a foundational understanding of crypto and all the cryptocurrencies and cryptographic stuff that goes on behind the scenes there. It's helped me get a foundational understanding of how algorithms work and how programming and Python works. And recently their how AI works course has been absolutely brilliant. And they basically break down how large language models like Chad GBT actually function behind the scenes which is a incredibly fascinating and also it helps you actually use AI better when you understand how it works. The other cool thing about Brilliant is that they really focus on problem solving rather than just getting you to watch videos. So yes, they give you content to help you understand the concept, but then they give you a problem with like an interactive interface that involves using that concept to solve the problem and that just makes learning way more fun and way more interactive. The courses are crafted by a world-class team of researchers from MIT and Harvard and Stanford and they're designed for ages 10 through 110. So whether you're a beginner or you're looking to level up your skills, there's going to be something for you. So if you would like to learn for free on Brilliant for a full 30 days, go to brilliant. org/aliabadal or scan the QR code on screen or click the link down in the video description and that will all get you 20% off the annual premium subscription. So thank you so much Brilliant for sponsoring this video and let's get back to it. All

right, so now we're moving on to phase three where things get a little bit more interesting because so far we have only asked AI to help us think. We haven't actually asked it to do anything other than that just yet. And phase three is where that's going to change. So phase three is using AI as your worker. And this is going to be weeks three and four. So a twoe phase. Now here is where we're going to get the AI to actually start doing stuff for us. The mistake I've seen most people make is that they will go straight to this phase and they'll say something like, "Write me this Instagram post. " And if you do that, the output is probably going to be very generic and it's not good. Instead, I'd recommend thinking about it using the 10 AT10 rule. I think I got this from Dan Martell's book, Buy Back Your Time. This is like a system you can use to delegate stuff to real humans. Given that we're basically treating our AI as a very smart intern that we can talk to in a chat window at all times, we're going to be using the 108010 rule. Now, the 108010 rule is that you are going to do the first 10% of the job. You're going to ask the AI to do the middle 80% and then you're going to do the final 10% for a taste check, for a vibe check, for like quality assurance on the thing. You absolutely do not want to be jumping to trying to get the AI to do 100% of the work for you because that's when you end up with absolute garbage. So, let's say Nicole, who's in charge of my Instagram, needs to come up with content ideas for a filming day. That's a core part of her job. The noob version of this would be to go to Chad BT or whatever and say, "Hey, come up with 50 content ideas for my Instagram. " But if instead Nicole were to say, "This is a transcript from Aliabdal's latest YouTube video. Here are three Instagram reels from competitors that performed really well this month. Here is our current content strategy doc that explains our target audience and brand voice. Based on all this, give me 20 hook ideas that would work as Instagram reels. Focus on counterintuitive takes and pattern interrupts. So Nicole's given the AI a lot of context. She's copied and pasted a bunch of stuff and therefore the AI is generally able to give a much better response. And then Nicole is obviously going to go through the list of content ideas that the AI has given. And she's not just going to take all of them wholesale. She's going to use her own taste and her own discernment to decide which of the ones she actually agrees with. And then what you can always do is you can, you know, let's say you've gotten 20 ideas from the AI and you like five of them. You can just copy and paste those five back into the AI and say, "Hey, these were the five I've resonated the most with. Give me 50 more ideas along this vein. " And then of the 50 more ideas it generates, maybe you'll like another 10 of them. So now you've just generated 15 content ideas without having to do a lot of the heavy lifting yourself. But crucially, all 15 of those have been human vetted so that they actually have taste and discernment applied to them rather than just being AI slop. Now, I want to do a little tangent at this point about the idea of taste. This is the key thing that separates the people who use AI well from the people who don't. For whatever you're doing, for your work or your business or your personal life, you should have an intuitive feel for what is good and what is bad. If you're a total beginner to the thing, you might not have developed that taste yet, but over time, as you become a better professional, you will develop a feel for like, okay, this is actually a good piece of content. This is a good sales page. You have this feeling of taste around what is good and what is not. And the biggest issue with AI is that if you are asking it to generate stuff for you, it will produce stuff that hopefully you'll think, h I don't really like it. You'll get that internal feeling of like cringe at the output that the AI has produced. That's a very good sign because that means your bar for taste like what you think is actually good is here and the AI has not quite met that bar and your job is to give it feedback like you would a junior team member or an intern. So by the end of week four, the whole idea is that if you start doing this with anything that a chat GBT window can respond to, like anything to do with writing or strategy, and you're developing the habit of any time you're doing anything at all, you're asking yourself, could I ask the AI to do this in addition to me so that I can test the response of the AI? And it's really a process of experimentation and testing. Okay, so at this point in phase three, you're already ahead of most people because you're using the 108010 rule to communicate with your AI inter. But the issue is every time you use AI right now, you're starting from scratch. You're opening up a new chbt clawed window or whatever. And so instead of starting from scratch every time, what if the AI could actually get better and better over time so that the hundth time you do something is actually way better than the first time you do the thing?

That is where we get to phase four, which is using AI as a system. And for most people, this is probably going to take around 1 to two months to get really comfortable with this approach to AI. Now, at this point, I'd like to offer you an analogy, which is imagine that you are baking a cake. You got the ingredients of the cake and you bake the cake and like you know, you follow a basic recipe and it comes out okay cuz it's the first time you bake the cake. And now you find yourself having to bake this cake every day because it's part of your job. And over time you start finding yourself wanting to add sugar at the end of the process because the cake is just a little too not sweet. And so at some point you might ask yourself, wait a minute, what if I were to just modify the recipe so that I just added more sugar up front? And you look at the recipe and you see the recipe involves one cup of sugar or whatever. you're like, "All right, let me try 1. 5 cups of sugar to see if the cake has my desired sweetness. " So, you give it a go. You add an extra 1/2 cup of sugar, and then you see how the cake turns out, and you're like, "Oh, that's actually pretty solid. " So, then you update the recipe. So, from now on, you're using 1. 5 cups of sugar rather than one cup of sugar. Then, to continue this analogy, let's say you realize after a while that, huh, it always seems a little bit dry, and I find myself wanting to add chocolate sauce at the end of it just to add that like moisture. You might then think to yourself, wait a minute, why don't I experiment with the recipe? What if I were to add the chocolate sauce earlier on in the process? What might that look like? So, you add the chocolate sauce earlier. You experiment with it and lo and behold, it turns out amazing. And you're like, great, I'm going to update my recipe so that the chocolate sauce is in there every single time. And this is how your grandma's cake recipe ended up being passed down the generations because she probably worked on it hundreds of times and developed that recipe over time. What the hell does any of this have to do with AI? Well, I'm glad you asked because it's the same kind of thing when we are working with AI. The idea is that we are trying to build our own prompt library using prompt engineering. So the very first time, let's say Nicole, our social media manager, the very first time she uses Chad GP to generate content ideas, she might say, "Here is a transcript of some content that my boss Ali Bal has produced. Give me 50 Instagram real hook ideas from it. " That would be version one. That's version one of the recipe. Then she sees how it goes. She realizes the hooks are a little bit generic. And so she adds to the prompt, "Make sure each hook uses a pattern interrupt or a controversial take. Avoid anything that sounds like generic advice. " She's like, "All right, maybe if I just ask the AI to not be generic, maybe it won't be generic. " And then she looks at the results of that same prompt and she's like, "Oh, this was actually a little bit better. " So then she updates the recipe and calls it version 2 v2. Then she notices that like, you know, the output is giving is that the hooks are a bit too long to actually be a sentence that I could say out on an Instagram reel. So then she updates the prompt to say, "Make sure each hook is under 20 words. " Again, she tests the output and she realizes, "Oh, wait a minute. That was actually better. So let me update my recipe. prompt. " So, it's now a V4 prompt or V3 or whatever. Then she finds I give a feedback saying, "I really don't like rhetorical questions cuz rhetorical questions are just never something that I use in real life. " And so, she then updates the recipe to say, "Make sure you never use rhetorical questions. " And that is V5 of the prompt. At this point, Nicole could use an app like Text Expander. She could create a keyboard shortcut for the phrase, I don't know, IG, Instagram hooks, and then by typing IG into any kind of text bar, it would automatically expand out and give her the whole prompt. And this would be an example of prompt engineering to generate a prompt library. So, in Nicole's case, she would have a prompt for hook generation. turning a transcript into a LinkedIn post, for example, she might have a prompt for um analyzing a competitor's Instagram account and figuring out like what they're doing that we could basically steal ideas from. And so, if you were to apply this to your own work, then by month three, you'll have a prompt library. And now, over time, all of these prompts are just getting better because you're able to add more context to them as the workflow changes or evolves. The other thing you can then do when you have this sort of systemized list of prompts in your prompt library is you can start experimenting with different AI models. So maybe you're just using the basic free version of Chad GPT for your hook generation or whatever, but then you think, you know what, let me try the free trial of Chad GPT Pro and see if Chad GBT 5. 1 is any better. And over time you realize that like actually certain models work best with certain prompts. And at that point you might be like, man, I'm getting so much value from this. I might as well just get the pro subscription to Chadyp and Claude and Gemini or ask your workplace to pay for it or whatever the situation is. In my case, I have a paid subscription to all of these things because I find them incredibly useful. At this point, or maybe even before this point, you might have realized that actually there's a bunch of tasks you might need to do in your work or in your business that actually cannot just be done through a text interface. Like maybe you need to create slide decks. And so you can find AI tools for that. There's GMA, there's beautiful. ai, Figma, Slides now has some AI generation tools applied to it. The mistake people make here is getting overwhelmed with all the choices. Oh my god, there's a hundred new AI tools coming out every week. How do I know which one to use? It's like, don't worry about it. find the AI tools that are helping you with your specific use cases with your work or with your business. Now, everything at this point has required you to be in the loop. But wouldn't it be absolutely sick if you didn't even need to talk to the AI? What if you could just set up a kind of system once and then have the AI automatically running in the background doing the work for you? And this is where we get to phase five, which is AI

as infrastructure. This is probably going to be month number four onwards for most people. And you could spend literally years going deeper and deeper into this rabbit hole. So again, to use Nicole as an example, part of her job is to take transcripts from the YouTube videos that I've created and to figure out what are some of the interesting points I made in these YouTube videos, which we could then film separately as like an Instagram reel or write up as an Instagram carousel. But even if you're doing prompt engineering, prompt library, 108010 rule, you are still finding yourself manually doing stuff. And at a certain point, you might think, man, we're releasing so much content. I'm having to do this manual repetitive task like for 3 hours every week. I wonder, is there a way I could automate this? And this is where we get into the rabbit hole of AI automation. And there's like a few different like levels of depth here. So level one would be using AI automation that's built into tools that you are already using. So Nicole might discover, for example, that hey, our editors edit videos in Premiere Pro. We use a plug-in called Fire Cut, which is an AI plugin to speed up editing. And Fire Cut automatically generates a transcript and can automatically just chuck it into a Google Drive. Fantastic. That's one step of the process that's already been automated that we don't really need to do anything for. Then you've got level two, which would be using simple automation tools like Zapier or like make. com. Basically, these are connector tools that let you connect different apps together. So you can say every time there's a new Zoom call recording, I want you to automatically get a transcription from the Zoom recording and then I want you to run that transcription through chat GPT using this particular prompt which we get from our prompt library and then I want you to give me the output as a Slack message. You can do that kind of stuff with Zapier or Make. com. Level three is where you graduate to a more powerful automation tool, something like N8N, which I've been playing a lot with recently. Now, these give you more granular control over these automations that you're building, but they do require a little bit of more technical knowledge, but you can build much more fancy, sophisticated workflows. And when you're at this phase and you start watching video tutorials on YouTube about how to use NA10, all of that stuff starts to make sense and you start getting way more ideas for what you could automate if you wanted to in your professional or your personal life. The other example could be Gio, who's our head of student success for a lifestyle business academy. One use of automations for her that we are currently trying to build is that every time we have a coaching call with one of our students, all of that is automatically recorded by a grain and it's automatically transcribed. So every week, can we run a weekly automation that automatically takes all of the transcripts for all of the students coaching calls combines that with the conversation we've had our students based on the Slack support channels cuz we give all of our students one-on-one Slack support. And then based on its knowledge of where the students are in the lifestyle business building roadmap based on our own like CRM which we have in notion, can the automation take all of this data and automatically give us a weekly report for every student that summarizes what are their wins and what are they struggling with and what are the areas where they might need support next week. That would be sick because currently that's a manual process that takes hours and hours every Friday. And if we could automate that work, it would save our coaches a lot of admin time and it would mean that they can actually spend more time talking to the students rather than doing admin and trying to sort of put data together all in one place. And then level four of automation is where you're not just using built-in sort of connector tools that drag and drop stuff. This is where you're actually building your own AI apps. You might not be building them to try and sell them to the market because that's actually kind of hard. But you can totally build your own internal tools that you are using within your workplace or within your business. But to be honest, a lot of this is overkill mostly for most things. You don't really need to at least right now go much beyond like learning how to use Zapio to connect things together. So, as you get into this world of automations, you start realizing that like, whoa, there's this infinite rabbit hole of things I could automate. And then the discipline becomes basically deciding what is actually worth automating versus what's worth continuing to do manually versus even better, what's a process that you actually don't need that you could just delete from your work or from your business. Now, in this video, I've given you a lot of examples of team members of mine. But if you're interested in my own personal workflow for how I actually use AI as an entrepreneur and business owner, then I've got this video over here that walks through exactly how I use AI to create new features for the software that we're building. how I use it to generate content ideas for building my personal brand and that shares a bunch more data around like the input process output method of using AI to improve your productivity. So that'll be right over here. Thank you so much for watching and I will see you