How to Use AI to Grow Revenue in 2026

Everyone's talking about AI in terms of using it for themselves personally, but not enough people are talking about how to use AI to drive more revenues or more specifically, how to grow a business. So in this video, I'm going to cover how to think about it from a high level. I'm also going to show you some workflows that we use and the different tools that you should be thinking about as of this recording. So let's get into it. So the challenge with AI right now is that a lot of companies, maybe 9% of companies are actually shipping AI at scale. The other 91%, they're experimenting or they just haven't started at all. And this is the gap that most businesses have right now. And the way we think about it right now with our business is that we have people at four different levels over here. You have unacceptable, capable, adoptive, transformative. So you have people that are unacceptable, which means they don't use AI at all. You have people who are capable, maybe they're using Claude, they're using ChatGPT, more like a search engine. Then you have people who are adoptive. So adoptive over here, that means people that know how to build end-to-end workflows. And I'm going to show you a chart in a second to explain what I mean by that. And then I'm going to show you more end-to-end workflows, but also closed loops as well, which we'll get into in a second. Transformative, these are people that are building these closed loops. Most people that are hovering around unacceptable, they work in open loops and you cannot work that way. So this is where the unacceptable live right now. They're working the same way that they've worked in like 2023, for example. Okay, even 2024, even 2025. Okay, open loops where it's like, hey, I'm going to ping you over here on Slack. Can you check this over here? Can you give me an update on this over here? What are the notes? What's the handoff over here? Hey, please don't forget this. Hey, just following up over here. That way it doesn't work anymore because you have a human in the loop. Then you have a lot of manual follow-up and then status unknown. And if the humans forget as well, then the work leaks out. And so you have a lot of waste that ends up happening. So output exists. Ownership is fuzzy. What ends up happening here is you have this ad hoc work and nobody enjoys that. That's the old way of working. It doesn't work anymore. That's an unacceptable way of working. Then you have people who are using ChatGPT. Maybe they're dabbling a little bit. Maybe they're using Claude, okay? And obviously capable is the bar is going to continue to shift over time. But right now, at least, people who are adoptive from an AI standpoint, these are people who are building end-to-end workflows. That means is you have a human that will input and define what they're looking for. So maybe they want to create a landing page. Maybe they want to create ad creatives. Maybe they want to create cold email infrastructure for sales. Maybe they want to do something else for business. Maybe they want a CFO. They want a CFO type analysis, okay? For the numbers that they have. So you have these known steps over here. And what happens is, by the way, within these steps, the AI is thinking. That's what an end-to-end workflow looks like from an AI standpoint. So human inputs, there's AI thinking in the middle. And then you get a deliverable for the human to review. And that's what happens. Okay, that's a true end-to-end workflow. You have a map path and limited learning. And then this is a repeatable process that can happen. But just keep in mind, end-to-end workflow when it comes to AI, AI has to be thinking over here, all right? So that's more on the adoptive side of things, okay? Now, transformative would be someone who knows how to build closed loops. So a closed loop is where you think about input, human inputs. You have execution, output. You get customer signal as well. AI agent will come in, measure, will think for you. You'll learn. And actually, probably there's a robot at all these steps over here. And then what will happen is you will build a skill.md file for this loop, and it can continue to repeat. Now, the key thing for this is that you have the agent that's learning over time. It's recursively self-improving. And so you can think about Andre Karpathy. He has this skill known as the auto-research skill in his GitHub, where it's constantly looking at things. It's constantly figuring out how to optimize things and make them better over time. And you're going to have to define what the definition of done or success looks like. And then this closed loop can continue to get better over time. And obviously, the human's going to continue to tweak this closed loop and try to make it better. But the whole idea here is that you have a measurable outcome. There's auto-follow-up. You codify a skill.md file where basically it's the process documentation. And then the system continually improves over time. Maybe you can pull from customer signals as well and look at the data that you have. And so what happens is the work compounds instead of having to restart. Because when you run an end-to-end workflow, you're going to have to start it up again next time. You're going to have to start it up again, right? And so there's a handful of things that you can do here, but we for sure need to get away from open loops. The world that we're moving into has end-to-end workflows. And these end-to-end workflows eventually will fit into these closed loops. And it goes over and over and over again. And what ends up happening is you have a self-improving operating system. So humans can focus more on doing human work only. Okay, so let me tell you where this is all going now. Once you understand kind of the philosophy in terms of how you should be evaluating your team and how you should be thinking about building when it comes to AI for business. And let's jump over to my other screen. So by the way, when it comes to business, one of the biggest challenges right now is that 72% of executives can feel the benefit of AI. They can see the benefit of it. But 10% of the workforce, only 10% of the workforce, they don't feel it, right? They don't see it. They don't feel it. They see people talking about it. They look at the threat of maybe job loss and it doesn't feel good to them, which is why, at least for in America, a lot of people are actually against AI, which is weird because when you look at people in China, they're actually very for AI. Anyway, the gap isn't a technology. It's more so the implementation piece. Oftentimes people will talk about change management. What I will say with change management is this. Change management is difficult because humans are very hard to change. We just don't want to change. We're stuck in our ways. And inertia is a very real thing because I remember when I asked my dad, when the internet first came out and workforce is starting to adopt it, I asked him because he worked for a few companies as a computer programmer. I was like, hey, how long did it take for your companies to adopt email? Five years, okay? So five years to adopt email. And maybe that was acceptable back then. But when you have compounding intelligence at the tip of your fingertips, you can't actually afford to wait five years. It's just way too long. So the way you want to think about this from a change management standpoint in your business is that you actually want to think about all the tools that you have, all the connectors, like your meta ads, your Google ads, your analytics, your CRMs, your sales intelligence tools, whatever that you have exactly. You need to make sure that these things are talking to each other. So the big challenge right now, at least in the old days, is that a lot of these tools didn't talk to each other. That has changed now. So you see all these characters, they're not really talking to each other. So when they don't talk to each other, you don't get the benefit of compounding. So tools don't talk. Knowledge stays siloed. Change management stalls. And pilots never ship. So your whole idea here in terms of using AI the right way for business to grow your revenues is to think about, okay, what are the different levels of AI that are acceptable? And what do we need to think about building? But also, if we're going to think about change management and get all these tools talking to each other, we need a clear solution for that to make it very easy, which I'm going to show you in a moment. So you want to make sure that you have one brain. Okay, we call it single brain. There's people that call it world brain or unified intelligence. Whatever it is exactly, just look at your GitHub, your HubSpot, your Salesforce, Search Console, Google Analytics, Slack. Whatever business tools that you have that are critical, you want to connect into one brain. And that's why we call ours single brain. And we've been implementing for ourselves. We've been implementing it for clients. And look, I'm not saying you need to buy single brain. I'm just saying that whatever it is exactly, you need to think about some form of unified intelligence that is working within the tools that you have already, which might be Microsoft Teams or it might be Slack. Because if your team is already talking in there, they're already used to it. There is no inertia. And not only that, when you think about your team talking to these agents, it's no different than talking to human beings. In fact, they're collaborating with these agents and they feel like these agents are part of their team. And so the resistance, their defenses drop. That's how you want to think about this, right? So you want to think about the org chart changing quite a bit. You have this unified intelligence over here. You have all these people, like sales, SEO, paid media, ops. They're talking to this brain over here and then they're asking for what they need. And then there's a fleet commander over here that tells these agents what to do. And they have sub-agents below. So the way we think about it is we have these specialized agents that you're going to need for your company for business to grow faster. And maybe you start with one first. And then you think about adding more agents, more sub-agents, things like that. And then you want to think about giving your team these agents as well. So what I will call out here in terms of using AI the right way, you can't just be over here. Oh yeah, we bought our chat GPT Teams subscription and that's it. Oh yeah, we did cloud teams and that's it. Gemini, that's it. That cannot be enough because when you have people using these, so do-it-yourself is over here. But when you have done with you where the specialized agents are collaborating with you, you're going to be much stronger than the people over here. But the brain ultimately, the single brain that lives over here, that this is where the leverage is. Because this brain has context of your entire company, your processes, your documentation, all the stuff that you have, all the Loom videos that you made before, the CRM context, sales intelligence context. It's all over here, which can guide over here, which then eliminates the need for you to type in all these things manually. You would just work over here mostly and then work over here. Really these two parts over here. And ultimately what that means is you, your team, is now focused on doing human strategic work only. Maybe not just strategic work, but human work only. While the robots focus on doing robot work. And that's how you use AI the right way, right? You have it live in Slack. And I'm going to show you in a moment how I go about using it inside all the things that I ask it to do. It is my strategic thought partner. And it's weird to me because I was looking at YouTube and nobody's talking about this right now. I'm like, why is nobody talking about this right now? This is crazy. Everybody talks about AI, but nobody talks about the implementation piece. So I'm going to give you an example right here and then I'm going to pop over to my Slack screen. By the way, if you want to grow faster, you need to have a single brain unified intelligence sitting inside of your chat. So it could be inside of Slack or Teams. But you can see right here, this bot's working and it's creating ad creatives. It is doing data pulls from meta, from Google. You can pull your SEO data. So imagine having all these data connectors. You can ask whatever question you want. Get the data pulls a lot faster. Your team can see it as well. They can choose to execute. And then you can run your other specialist agents that you have inside. So we have ad creative agents. We have email infrastructure agents. There's all these agents that can do a bunch of things. And the whole idea is they're all playing together. They're playing with your team as well. That way you're going to be able to just move a lot faster and then grow a lot faster. So check it out. Just go to singlebrain.com. Singlebrain.com. We'll see you inside. So here, this is an example where I'm talking to an agent. Give me a 7-8 post on the ads. Okay, here's what the ads look like over here. Oh, we're burning too much money over here. And oh, you know, we want to refresh this creative over here. So you can see it's multiplayer. First, it was me asking and Christy's asking. Here's our other agent, Picasso, which is our creative agent. We have Oracle, which is our analytics agent. Here's our spend performance on data. So imagine this. You're getting data pulls quickly, which will lead to more questions. And you go back and forth with it. And you can strategize. And then what happens is you see Christy here asking for three variations of two different ad formats. Boom, it kicks it out. Picasso is the creative specialist when it comes to ad creatives. And that's how you think about this, right? So the question really is, why do you want to build for 22 weeks when you can build in 22 minutes? So I'll say it again. Why build for 22 weeks when you can build in 22 minutes? And that's the whole idea with all this. You want to have this living inside of your Slack or your Microsoft Teams so you can compress time. So you no longer need to think about what do we need to do about change management. It doesn't matter when you have this inside and everyone sees it, it's in public and everyone's using it. I'm telling you, it changes the game completely. We have people that say I'm at least 40% faster or I feel like when I don't have this, I'm drinking soup with a fork. This will unlock the way that you work. And that's why it's super important for everyone to learn how to work like this. I believe that most companies are going to be working like this in the next 12 months or so. So you might as well get on this now. You might as well start compounding on this right now. Because if you don't, well, your competitors will be doing that. So this is how I work with AI for my business. So you can see I'm inside of Slack right now. And this is my Hermes agent. We actually call it Hermes or Hermes because of the logo over here. So it's Hermes. And you can see literally I'm strategizing with it. Like I might see something interesting on X, for example, and I'll share it over here. I'll say, hey, what would this do for us? Now it's my thought part because people are posting interesting stuff. There's new launches on X all the time. So two of these can run DeepSeq v4 Flash. And so I actually have two DJX Sparks. And so this is what it's talking about. And so threat level. Okay, here's what you should be thinking over here. Should I be doing this? And then basically I can go back and forth with it. I can ask a follow-up question or I can say, hey, you know, audit our BeatClaw challenge, which is our AI challenge over here. So this is a GitHub repo that we have that has our BeatClaw challenge where people need to beat AI in order to join our team. And so it's doing all this work over here. And it's like, hey, did I harden it? Did I make the challenge harder? And it says mostly hardened, right? And it verified the repo over here. I'm just going back and forth with it. And so the more I connect with this thing, the stronger it gets. So it has the X API in there. It has access to my GitHub repo. And so I'm just like, okay, well, like what should I do? Like can people still easily game the test? Yes, they can still game it. Here's what you should think about. Here's what I would change. Remove public cloud baselines, collapse public rubrics, add private follow-up variants, require some source artifacts, use hidden scoring dimensions, add or defense. And then I can say this, watch. Hey, so how much harder would this make it if we add all the things that you want to add over here on a scale of one to 10? Then I'll just let this work. And then by the way, earlier, I was just making a YouTube video on the different slash goal commands that we've added. So slash goal allows us to basically do work more overnight, for example. So agents can do work overnight. It can queue up more work to do. It can basically auto-optimize work to do as well. And so I wanted to make a video talking about the work that we did. And so this can be used for content as well. So when I'm looking at this, we have slash goal overnight. We have night queue. We have AI optimization lab, all these like AI optimization lab metrics become hypotheses. So pull from the data that we have, like it'll pull from Metricool, our social media data, it'll pull from Stripe, it'll pull from Mixpanel. And then all of a sudden you see here, hypotheses become scored experiments, experiments become approved tests, results feed a learning store. And this turns into a compounding growth loop. Who doesn't want that? So if you are a strategic thinker and I'm like, oh, all the work that we did, let's make this into a diagram. By the way, I have a design.mb file where it doesn't just look generic. Sure, this might look AI generated, but I was then using this to go through my YouTube videos and I was explaining as I was going. And so it's easier for people to follow along. And I'm like, oh, by the way, what other goals do we add this week? Okay, so I wanted to better explain the video. You can do that now. By the way, Peter Steinberger tweeted this little chart. I'm like, what is this thing over here? What does this do for us? So this is useful, not what we need for Codex Bar. AI cost observability is becoming table stakes for agent fleets. Here's what it shows. Here's what it does for us. And then do we have a version of this right now? Is there something I could take a look at? I want a dashboard that I can quickly click on to see how we're using stuff just in general. So let me know. Feel free to ask me any clarifying questions. By the way, I'm using Whisperflow to dictate that out. And then come back over here. And I went over to the BeatClaw challenge. So I was like, hey, how much harder would it make? By the way, we're going back and forth now. We're going back and forth. So 8 out of 10 harder. Curve version is maybe 4 out of 10 hard to game. It feels there's a lazy AI slot, but it's more applicant. So you know what? Okay, yeah, let's do that. Let's do 9 out of 10. Do we need to goal this? So you see what happens. And then all of a sudden, I have all these multiple threads that we're pulling on over here. And then you can use slash goal to finish these things. And so this is actually the way to work with AI. And I have no idea why there's not that much content, whether it's on X or YouTube on this stuff right now. It's just crazy to me. Why is it all the people that are just talking about how they're using their AI assistants and they're not talking about using it for business? That blows my mind. And that's why I made this video for you. So if you enjoyed this video, you want to check out this next one over here in terms of how I'm using Hermes as a revenue agent and OpenClaw as a revenue agent that you can use. And so I'm going to let this work over here slash goal while you check out that next video over there. And I'll see you inside.