If You’re Not Using AI Like This, You’re Behind

I think a week or two ago, we talked about the K-shaped economy, right? And so, I want to review what that is and talk about where that's going. And then, I found another article yesterday talking about— the headline is, you've been kicked out of the arena, you just don't know it yet. Because everyone's talking about you got to be the man in the arena and all that, right? But people are just like, no, you've been kicked out of the arena, you just don't know yet. And I want to get your reaction to that.

Okay.

So, let me share my first screen here. And I apologize, I'm on a laptop. I hope this doesn't break. Neil, can you see this okay?

Yeah.

Okay, cool. So I'm going to show the graph first. So it says, look, this is the K-shaped economy. Okay. So basically what you're seeing on the screen over here is you're seeing a line, you're seeing a graph. Okay. And then chronologically, it's just like one line. But then that line diverges into one that, that scales up. Okay. The graph goes up and then the line graph goes down. Okay. That's where it splits off. That's why it looks like a K. So you have high agency, which is the— this is the bifurcation point where it's moving up and these are high agency plus AI mastery people. Okay, then you have low agency people who are managed by AI. So, and I literally think this is—

Before you go into it, do you want to explain high agency, low agency? Because Eric here is not talking about ad agencies or marketing agencies. If anyone's listening and can't see the graph.

Oh, totally, dude. I got to give you some insight. I spoke to one of these, these ad agency people too. He works with a holding company, so we should talk about that. But on the agency side, okay, high agency means that if you're a high agency person, that means that you're someone that's very proactive, you get things done, and you just figure things out, right? Like, you know, everyone knows those people. You're very resilient. You're very— you have a very strong bias to action. Am I making sense there?

Yeah, you're— you execute, you're a go-getter, you don't procrastinate, you don't wait for someone to tell you what to do. You just go and do stuff on your own.

And the world revolves around these people. People are attracted to these people, right? Low agency is like, okay, maybe you just want to— you're the exact opposite. Maybe you don't want to have a bias to action. Maybe you just want to chill, right? Maybe you just want a 9 to 5, which is okay, right? Maybe, maybe you're in a different phase of your life, but that's lower agency. You don't want to get things done that quickly. So his point here is that in 10 years there's going to be 2 classes of people. So economists call it the K-shaped economy, and the next 2 to 3 years will decide which line you're on. So I don't think there's a lot of time here, right? So You're going to have an overclass that uses AI as a lever to build wealth, automate income, and make decisions at a speed that no human can compete with alone. And then you're going to have an underclass that gets managed by it. So imagine this, like, you need— not you need— Neil, when— because you've worked as a janitor before, right? And I've worked at like— I think I've worked at a fast food restaurant. No, no, no. I worked at Circuit City before. You worked at it.

You clean toilets, right? Yeah, I clean toilets.

Yeah, you clean toilets, right? So, so my point is, you know, if you're a toilet cleaner in the future, you're probably going to get managed by an AI, right? Like early in your career, like, you know, the data entry job that I was doing, my first job out of college, I'd be managed by an AI.

So what this graph is just saying is there's a K-shaped economy, or what's the overall point?

The overall point is you're going to see with high agency, people are going to like be the ones managing the AIs, and then low agency people will get managed by AIs. That's the entire point because people are just growing so fast right now. Yeah, yeah, yeah.

When I look at the K-shaped economy and like the graph, I look at even pre-AI, we were already heading to a K-shaped economy. I think AI just accelerates it.

Yeah, completely. And, you know, we kind of talked about this before, but, you know, I said that money amplifies your nature, AI amplifies your intelligence. And what's interesting is that, you know, a lot of people here see the same thing, right? And we're kind of talking about that right now. So to close out this point, He just says, this guy Miles says, if you're reading this now and you haven't built systems with AI, haven't automated a single workflow, haven't used it to create anything that makes you money or makes you irreplaceable, you are currently on the wrong line. That's not an insult. You have the agency, right? You have the will, your own willpower to change your trajectory right now. But 6 months from now, the gap will be twice as wide. And a year from now, it might not be crossable. And I actually see that. And what I'll say to everyone listening to this is if you're looking for a job right now, and you're into this stuff, both Neil and I are hiring and we're looking for people with amazing agency because to Neil's point earlier, everything is a talent game more than ever now. Everything is a talent game that I think the AI stuff actually reinforces the need for strong talent because last week we did one, an episode talking about slot cannons. We need to hire people with good judgment. Everything is downstream of hiring. Quick break. I want to tell you about something that we've been building for years at Single Grain. We've been doing SEO and content for clients. And the biggest problem is always execution. Everyone knows what to do, but doing it at scale is brutal. So we built ClickFlow. It's an AI SEO product that creates content that actually sounds like your brand, not just generic AI garbage. It also helps with other areas such as internal linking, creating FAQs, and more. We use it ourselves. Our clients use it. One user reported saving over 90 hours a month alone. If AI SEO matters to your business, you can try it for free for 14 days at clickflow.com. All right, back to it.

One of my favorite AI quotes is actually from Eric, and he said it either a week ago or 2 weeks ago when we were recording the podcast. AI brings out the best in you or the worst in you. And I 100% agree with what Eric said, and I love that quote, and I've been using it in which if you are an A player and you're just a go-getter, you execute on your own. AI just makes you even better, more dangerous, more valuable. If you're lazy and you don't like working, AI brings out the worst in you in which you try to figure out how to get AI to do as much of your job as possible. And I see marketers being like, oh, look at this, it created a website for me, it's creating content for me, it's doing outreach to try to build links for me. And they're like, this is great. Look how much of my job I automated. And then they're telling me like, yeah, this technology is amazing. Everyone needs to use it. But then I see what they're doing with their time. They set up a lot of these systems and tools to help them out so that way they can go run around and take photos or go travel or hike or whatever they want. And I'm like, so you're telling me I don't need to pay you? Now, the people who are showing me this are friends and they don't really work for me. But it's just like, if they show that to me and they were working for me, hypothetically, I'd be like, so you're telling me you want to get paid to not work? And yes, it's great how you can use AI to do a lot of stuff, but you should be using your time for other things that AI is not as great at. And when people have the attitude of, oh, I just automated everything so I don't need to work, it's just like, those are people who are just going to be replaced. And they're not going to be backfilled. And AI really does bring out the best in people from what I've seen and the worst. And people need to have the right mentality in which it's a great tool. It can do a lot of what you're doing and make you more efficient and do even some things better. But that opens up a lot of time for you to do other things. If you decide to use that time to do other things like personal stuff and expect to get paid for it, You're in for a reckoning.

Dude, you know what's crazy? I was getting a haircut before I flew out to Florida on Sunday, right? And this is the first time I got a haircut before I was using Open Claw. And so with my barber, typically we just kind of, you know, I'd be doom scrolling on my phone and we'd be kind of, you know, talking about whatever, like small talk, right? Hey, what video games are you playing right now? Blah, blah, blah. How's your family? Right? That type of stuff. This time, as he's talking to me, I'm not doom scrolling. I actually ended up shipping, like when I say shipping, like my open call was doing a bunch of tasks. So he cuts my hair, takes 30 minutes or so. In that 30 minutes, I pushed it to do 7 tasks, right? Whereas before, I was like, these are like tasks like a typical human would do, right? Whether we're like strategizing on things or like doing a PRD or something like that, right? I'm like, holy crap. And the thing is like, before I'm going— like Neil talked about AI being an amplifier. Before I'm going to bed, I'm like on my phone, like it's like 7 PM because we're done with dinner and all that, right? And then literally until 10:30 PM, I'm on open call. I'm just pushing stuff, right? And so, I'm saying that I've never had more fun at work, but also like I've never worked harder either because the entire thing feels like a video game now. And I saw David Hensel. So, David Hensel, you know, we used to play a game together. So, I was 12 years old or so and he was like, I think he's like 21 years old or something like that. We played a game called EverQuest. And I'm like, dude, this is just like EverQuest. He's like, it's exactly like EverQuest. It's like World of Warcraft, right? I'm like, holy crap. You can literally build whatever you want, whenever you want, and it's only going to get crazier from here. And that's why it's so exciting. So I literally show them all OpenClaw, and a lot of them are using Claude code. Some of them are like maybe 1 or 2 people are just on ChatGPT, but we all feel behind and that's every single day. But we all know that there's no time to slack off. Like now's the time to press like even harder.

Yeah. And when you look at OpenClaw, a lot of people are telling me like, oh, I can't use it because I don't have a Mac Mini. Well, you can just spin up an AWS instance and there's Pop it on there. You don't need a Mac Mini.

I would be scared of— that's the only one word of caution there. I'd be, I'd be concerned with the security items there.

Because you set up a whole new instance and a whole new account on AWS.

Yeah, but that's— so the way I have it set up, right, it's when you put it on AWS or like a virtual private server or something like that, there's a lot of security issues around it. And even the way I have it set up right now on a Mac Mini, it's not bulletproof. So I would just say with OpenClaw, Like, it's nice and I use it. I'll use it for my own stuff, but we definitely wouldn't use it for clients because there's too many security issues.

Right.

And so we would— we want like a SOC 2 compliant version, which is something that we're actually actively working on building right now. And we just want to manage it ultimately because people like your clients, ultimately like our clients, they don't care if it's OpenClaw, ClosedClaw, or like whatever it is exactly, right? They just want outcomes. They just want outcomes.

Yeah, and if you're going to use it on an Amazon instance, you shouldn't be using it for anything sensitive, anything customer-facing. You should just be using it for testing and experimental stuff, in my opinion, and wait for OpenAI, who has a deal. And Eric called it on our last podcast. Do you remember you were saying like, yeah, the OpenCLoud founder is in San Francisco, probably doing a deal with, you know, like one of these big companies like OpenAI or Anthropic or one of them. I forgot which ones you were naming. Or Facebook, and he ended up doing a deal with OpenAI. And I'm waiting for OpenAI to release a version that's safe for us to all use, and then it's on.

Yeah, we'll see. I mean, it's, you know, we at Single Grain right now, we're already working on a SOC 2 compliant version. And there's a handful like Manus, for example, which is owned by Meta. They kind of have their own version of like an agent as well. So I think a lot of these are moving into this space. So we'll see what happens. The one thing I wanted to call out too, actually, this will be interesting, Neil, and then I want to come over to this article. So I met with someone at one of these— Google. Okay, so I met with someone at Google, and he actually manages some of these holdco relationships, these traditional agency holdco relationships. He works directly with some of these CEOs. So I think you would find this interesting because this is your space. So I want to get your reactions to some of these and Kind of go from there. Are you ready?

So to be clear, this is a person you met that is a Google employee. And then he deals with all the agencies and he specifically deals with the agency CEOs and stuff like that.

Yeah, some of these agencies hold co-CEOs and I think there's a handful of these types of people, but, you know, very smart guy and I've known him for a while. So I was like, okay, so what's the deal right now? And I was just, I'll kind of fill you in on the conversation. I was just like, Okay, he's like, dude, Eric, the churn right now, these holdcos say their churn is 20% annually. It's more like 30% and the churn is going up. And so he's like, when I talk to these holdco CEOs, I'm like, what are you guys going to do about this churn? Because the churn is really high, right? And none of them have a plan. I'm like, none of them have a plan? He's like, yeah, none of them. All their plan is to grow top line right now. It's just to continue to grow top line. And he's like, the problem that I see with a lot of these agencies is that they're not AI native, right? And so he's like, you guys are like, you know, AI native, or at least he perceives me to be, right? And so he's like, they're going to have— because the way they're structured, they have— they used to throw people. The problem was a labor arbitrage. That arbitrage is not going to work as well in the next, you know, 5, 10 years or so. And so he's like— and also I'm like, I was like, hey, entertain me here because he used to work in tech, he used to work in SaaS, he used to work in e-commerce. I was like, Hey, you've worked in all these industries. What percent of these people do you find impressive on the agency side? He's like, honestly, I don't look forward to having most of these conversations. Maybe only 5% of people are impressive on the agency side, which we've kind of joked about, right? I was like, okay, well, entertain me here. On the e-commerce side, on the startup side, on software side, what percent of people are generally impressive, right? He's like 15 to 20%. So that's 3 to 4 times more. So I'll pause for a second. What are your reactions to this so far?

Yeah, I could see how startups and smaller businesses are going to be more innovative because they don't have all the legacy, archaic bullcrap to deal with or politics, so they can do more innovative stuff. But the first reaction I get when you're telling me all of this is, this person doesn't seem to enjoy their job. They should go find a new job.

Could be, could be.

But that's my first reaction. On the churn metrics, it depends which companies you're talking about, but you have to keep in mind some of them are publicly traded, so I doubt that they're fudging their numbers on churn. He may not be talking about all publicly traded companies, and I'm assuming that is the case, but the publicly traded ones, I don't think they're churning at high numbers like 30% and stuff like that. People tend to forget, and this is the big thing about agencies. And I saw Eric on an X thread about this. You were debating with someone, okay?

Oh yeah, it just came from this conversation. Yeah.

So the thing with agencies is agencies are hired for people more than anything else. So if you're a company, a big organization like Procter Gamble, And you're looking to figure out things, you're hiring Accenture or McKinsey to help you with AI. In many cases, you agree with that statement, right? If you look at Accenture and McKinsey, their businesses in some areas are struggling, but the AI parts are booming. And we see the same thing in marketing people. And it goes back to the topic Eric and I talked about earlier, where it's all about hiring and the quality of people that you're recruiting, right? You just want amazing talent. So a lot of companies, what they do is they hire agencies because they have amazing talent. Not all agencies do, but the ones that do have amazing talent tend to do well because people hire you for the people and they want you to just figure out their problems for them. I'm not saying that'll always be the case, but that has been the case historically, and I believe that's going to be the case over the next few years at least.

I think that's always going to be the case. People want something to just do the work for them, something or someone. But he made some other interesting points here. So he's like, one, you know, I'm like, so what do you see agency people as? And so they are hiring from the people, right? But he's like, now there are no longer any differentiators. He's like, these agencies are just literally like budget managers. And I was like, okay, so explain, like, why do you think that? He's like, dude, there— and I shared this in the thread, but he's like, there are software companies in India that will go to like an agency, like maybe a Procter Gamble, and like, oh, you're Procter Gamble, you're working with this agency over here. What's your, you know, what are they charging you? Right? It's like, Could be, you know, 10%, 15%, 20%, sometimes 5% in terms of ad budget. You agree with that? Yes.

Uh-huh. Okay.

So he's like, oh, let's say hypothetically some, some agency is charging like, you know, 10% or 15%. He's like, okay, well, if you're paying, you know, 10% or something like that, how about we just charge you 4%? Right. And he's seeing these, these upstarts in India with whatever software that they have, they're taking hundreds of millions of dollars of budget away. And he's like, dude, he's like, this is just going to continue to happen on the AI-native side. I'm like, so what are people going to do? He's like, unfortunately, um, and he does— I think he does like his job. He works with really smart people. But he's like, dude, even people at Google, very smart, they don't want to rock the boat and they don't want to change anything, right? And so they'd rather not rock the boat and say, hey, these things are changing. They're just like, let's just keep it how it is. And that's why some people are like, yeah, maybe it's going to take a little longer for UBI to come, or in your case, maybe not at all.

So So I'm in a little bit of a different camp than this Google person, and this is just my own take. I've spoken at Google's offices a lot, at many different offices all throughout the world. Google and Facebook, I'll throw them also in the same bucket. They're really big on, they will build a technology that'll do all the bidding and everything, and they can leverage AI. And Mark Zuckerberg's even gone out being like, why do you need agencies? I don't know if you've heard him say that, but—

Yeah.

He has. He's big on that. And from their perspective, just think of it this way. Let's say you're paying someone 10% to manage your ad spend. In many cases, it's more. It's very rarely less unless you're spending an arm and a leg because it's hard to do it at 5% and actually make it profitable. Facebook and Google rather have that money go to them. Imagine if everyone was spending an average of 10% more. Okay, if Google does over $200 and something billion a year just in ad spend revenue, let's just use the $200 billion. I know they do more than that in ad revenue. If they did 10% more, that's extra $20 billion, that straight drops to the bottom line and that's pure profit. They would go from $116 billion a year to $136 billion a year in pure profit. That's a lot of extra cash. They would love that money. It's in their benefit to just go and get it. Now, keep in mind, I'm biased, and I would say Eric's also biased because we also run agencies, so we're on the other side of it. And we've seen companies just leverage the technology, and we have a lot of clients who use the AI Max, PMax, and the list goes on and on. And when we put a human in the loop, including our fees, In all cases where we have customers paying us, because if we can't do this, the customer doesn't keep paying us, we produce a better ROI than them only using the AI tech that Google and Facebook are giving. So when we charge fees, it is black and white because you have to remember we're dealing with business people, CFOs, finance people, and I'm going to use basic simple example here. If a company is spending $10 million on ads and that $10 million brings in $40 million in revenue, not profit, just $40 million in revenue, and let's say it's profitable for them, and now they're spending $10 million and you, Neil, are taking $1 million in fees, all right? So let's say they're spending— let's just keep it simple. Let's say instead of spending $10 million on Google, you're spending $9 million on Google, and they're spending $1 million to you, if you generate them $40 million in revenue, they're not happy. They better see $45, $50, $60 million in revenue, so then that way it's a financial win for them, and we're able to do that for customers, and I'm assuming Eric is also able to do that for customers because if he wasn't, he's shaking his head that he is. If he wasn't, he would be terminated. I am. Right? But we're seeing We're seeing companies, and we've seen this already, them going to not just places in India, we see this also in the US. Companies make the pitch how they can use AI and do this for them in an automated way and charge 1%, 2%. We're not even seeing 5%, we're seeing 1%, 2%. And we've actually seen some of our customers switch. We've seen 3 switch so far in the last 6 months from this. And of the 3, what do you think's happened if you had to take a guess, Eric?

A percentage are coming back.

You're right. 2 have come back. The 3rd one has not hit us up back yet. The 3rd one ended up switching. Dude, we just had a conversation on this on Tuesday. This is Thursday as we're recording this, but we go over lost accounts. So internally every week we go over lost accounts that we performed exceptionally well, And it is in their best benefit to come back. And the reason we cover this is, you know, we have game plans on how do we get them back over time. Because if something is better for the client and it's a better financial pitch for the client, you should be able to convince them logically to come back over time. The one that hasn't come back yet, it's been roughly 2 and a half months. And, you know, with campaigns, you can analyze and see some of the data publicly, not all of the data. But I believe we'll be able to— maybe they don't come back to us because they didn't like the fees, but it was more profitable than what we believe they're doing right now. But I believe they'll make a change within another month or 2. And the reason being is even if someone's cheaper, but in the grand scheme of things, again, if you were spending— instead of spending $10 million on Google and $1 million And then you switch to us and then you're saying, oh, we're spending $9 million on Google and $1 million with NP Digital, but we're generating more revenue. Oh, we found another company that's just using AI. We can spend $9.5 million with Google and they're only going to take $500 grand in fees. Oh, this gets us more ad inventory. But then your revenue starts tanking. You're going to start questioning some things and you're going to end up starting to make changes. We're finding Not technology being bad. We're finding the winning combination being technology with A players. In other words, rock stars combined with AI. And I think that's the caveat. When Eric talks about agencies or this Google person talks about agencies, and we're not talking about our agencies, we're talking about other people's agencies. Some of these agencies have amazing talent. Some of them, especially as they scale and they get really big to like $10, $15 billion in revenue, You do have some amazing talent, but the majority of their talent is mediocre and average. And what we're seeing is the agencies who are focused on hiring amazing people and not bloating their headcount, they're doing well when they combine AI. The ones that are just hiring for the sake of hiring, those are the ones that are struggling right now.

Yeah. So the thing I want to bring up is there's that 2x2, you know, the matrix again. I'm going to pull this up one more time just so everyone can see it.

So let me—

okay, here it is.

This is a slot cannon metric.

Yeah, sorry. So I think it's important to call out because it's— it is because this is exactly what we're talking about, right? So this is, this is what I have on the screen over here. So this is a 2 by 2, those of you that are just listening. So on the very bottom left, you have people who do not use AI and people who do not have good judgment. These, unfortunately, this is deadweight on your company. Okay. Now, Let's say if you have someone who has poor judgment and they use AI, that means they become slop cannons, which means they just add a bunch of gunk and garbage to the internet because they're just lazy in the way that they use it. Okay. Now, if you move to the bottom right here, you have people who have good judgment, but they don't use AI right now. So these are people, at least for right now in today's day and age, these are steady hands.

Okay.

So these are people that are, you know, they're going to get the job done overall, but they're not enhanced by AI. But if you have someone that has good judgment and they use AI, these are turbo brains. They're supercharged and they get things, they get way more done. They're already smart people anyway. These are the people that you ultimately want. And so you can't just say like, I might come in, it's like, don't get me wrong, I love AI, but I'm like, guys, guys, guys, we need to focus on talent right now. And people aren't quite— there's some people on my team that don't quite get what I mean, but I'm like, no, no, no, everything comes downstream of talent. We can talk about AI all we want, but if we hire poorly, we're just going to have a bunch of slop cannons on the team. Yeah.

And I want to give an example of a slop cannon. I'm not going to reveal the names because I know them, but I've been explicit to them. So let's call one of them John and the other one Michelle. Okay.

Michelle's your wife's name.

Yeah, Michelle's my wife's name, but my wife's not a marketer. She has a harder job than both Eric and I. She's a stay-at-home mom and she has to deal with 2 kids who are not rational. So, uh, which they'll be up in an hour and a half. Okay, so, uh, John says lovable.dev and build apps and websites with AI fast no-code builder. So then Michelle responds, but did you build anything? Was looking into it. John says, I built 10 websites, sends screenshots, sends a whole image of a full-blown landing page and a website. And John's like, takes a day, LOL. Michelle says, wow, familiar logo. Michelle's saying that because John's website is pretty much a ripoff of my website. I don't think you need AI to do that. All right. And the logo is straight up a ripoff of my ad agency website. And Michelle says, wow, familiar logo. And John says, imagine a Claude plugin, but in a builder. Neil even did a testimonial for my website. Laughing my— I'm not going to say the last 2, it's A-O. L-M-A-O. And John's like, can't believe you call yourself a marketer, Michelle. And you know, they're going into this. John's not a bad person. Great person. Like him as a friend. I've told John this to their face maybe 15, 20 times. What is John doing in your matrix here?

He's creating slop.

He's creating slop. This is no value. It's not going to convert well. You shouldn't use AI to just make up random testimonials. You could use it to copy other people's websites if you want to do that, but this is someone that doesn't need to be paid. If you're being paid to just copy someone else's website, why do they even need to pay you to do that? They can just do that themselves. There's no innovation here. Yes, you could use AI to help you build a website, but shouldn't you be hiring someone who has a solid sense of direction and a lot of experience so that way they don't need a designer, but they can use AI to build something that actually converts better? They understand your ideal customer profile, they're doing customer research and talking to customers to figuring out what angle you need to differentiate yourself versus another— all the other competitors out there. That is using AI to fine-tune their copy and going back and forth with AI to fine-tune it and making your headlines better, your CTAs better, running tests, etc. But like people like this who are using AI in the example I gave you for John, they are going to be replaced. And this is not how you should use AI as a marketer.

And we will see you.

We'll see you next week.

Goodbye.