AI Business Manifesto Live - August 11 2026

Am I live? There we go. Hello everyone, welcome. It is Tuesday, August 11th, noon Eastern time, and, uh, we're talking about AI. We're talking about judgment in AI. We're talking about how to get your AI to do a better job of really understanding your intent and being able to execute on your behalf. And, uh, one thing that I—

what's going on? Oh, the Twitter account. Okay, so we're just gonna go live. I'll do that.

And, uh, I, I definitely want to talk to you about judgment. I want you to— a lot of people have been telling me recently that they've been having more challenges getting what they want most out of AI, like as far as, um, the as far as getting the outputs that they want. Good afternoon, Mahatma, and good afternoon, One, or Onik, David Onik.

All right.

Good morning, David. Yeah, so a lot of people are having challenges these days with— hey, Biggie, and hey, Explore— with getting the outputs that they want, but it's so much bigger than that. And that's really what I am excited to share with you. So I have this challenge that I'm leading later on today. And one thing that I want to share, I was going to share it on there first, but we're not, but I thought I'd share it now. And so this is the very first time that I'm sharing this idea. And I think that it is a pretty profound concept. All right. So I created a screenshot of just a note that I want to share with you guys. And the— oh, that's great. So for those of you that signed up for the Imprint Challenge, that's what I was working on this morning. And this is something that I shared with my team earlier, and it's really the core idea. And so I hope that— let me just see if I can grab it here.

So hold on, I'm just going to slides and PDFs. PDFs, yes. No. AI Performance Lab, your computer. I don't want to upload the file. Can I just share it on my desk? Let's see. Share screen. Crap, Mike, you drive me crazy.

All right, there we go. And it's called the Judgment Age Thesis.

And there we go. Share. All right.

Did that work?

Why didn't that work?

Bear with me, guys. I promise you it's worth it. And I know that Let me do, okay, image file. Maybe I can just do an image file. All right, desktop.

Where is, no. This is not gonna work.

One more time.

Just share screen, sorry guys. Ah! All right, there we go.

All right. I will start talking in a second if I can't, uh, get this going.

Why won't this share?

I have another way of sharing it. All right, uh, one of these days I'm gonna get all the technical things out of the way and it's all gonna be smooth and wonderful, but until then I'll have to do it this way. And, uh, let's see.

All right, let's just put this here.

All right. And I know you might not be able to read it. So I'm actually going to move it and I'm going to put it to the side here, and then I'll try and make it a little bit bigger too.

Let's see if I can, and then I can scroll maybe. Can I? Yeah. All right.

All right. And if you're watching on, if you're watching on Instagram, at least you can hear me explain it. And I apologize for that. I can move it over a little bit. Yeah, you're not gonna be able to see it anyway. Okay, the Judgment Age thesis. The causal mechanism. The Judgment Age is not merely a period in which intelligence becomes abundant while judgment remains scarce, although that's part of it, right? That describes the outcome but misses the mechanism. The deeper change is a bottleneck inversion. AI increases the amount of work that can be attempted. The work creates more decisions, exceptions, trade-offs, branching paths, and consequences. Human attention cannot expand at the same rate. As modeling agents work longer without supervision, they must carry judgment, not merely instructions. When models' competence and harness converge, competitive advantage moves to the judgment directing them. The Judgment Age begins when machine intelligence becomes abundant, autonomous, and persistent enough that the volume and speed of possible action exceed human attention. The bottleneck is no longer whether AI can perform the work. It is whether AI can recognize what matters, choose among plausible paths, apply the right trade-offs, and know when to stop or escalate. As access to models, competence, and agentic infrastructure converge, judgment becomes the last non-commoditized layer of competitive advantage. And here's the causal chain. So if you got— if you get this, if you get these 2 last parts here, right? If you get these 2 last parts, you get it all. You understand it, right? And it's simply this: more intelligence creates more possible action. More action creates more judgment. We're talking about your decision-making. We're talking about how you perceive things and what you perceive, what you don't perceive. And so what I'm going to do on this Imprint Challenge Is one, help them set up their imprint to capture their feedback because their feedback is their tacit knowledge, their unconscious competence, their, the things that they know that they can't articulate spoken. It's in reaction. Does that make sense to you guys? That just even that concept that judgment Your perspective, your feedback, right, is the only way that you get access to your tastes, preferences, etc. Okay, so this is a good question. Let's see what Marcus has to ask. Okay, how do judgment and automa— automations— I'm not really Very articulate today.

Hold on.

How did judgment and automations go together?

Hold on. All right, there we go.

Example: In your copywriting arena, different sales letters compete. You allow AI to find the winner and compete again. How do you execute judgment on of which is the best results in an AI automation? without interfering? Well, you would have to, right? Like, so the— it's a great question, Marcus. And so for those that don't know, um, he's talking about an arena. So I have created lots of different arenas, and the purpose of an arena is for AI agents to compete with one another. I found, I found a long time ago that when agents were competing with one another, they would do better than if they weren't. And so I took the complete frameworks of different well-known copywriters from their courses, from their bodies of work, et cetera, and created these different agents that write in their voice, plan in like approach copy in their voice, et cetera. And then critics behind it to make sure that it follows all of the methodology. And then I created these arenas so that you could take like a creative brief and you could say, I want a sales letter for this. And then all of them would write sales letters, like 6 of them. would compete with one another. There would be a judge. The judge would award a winner, and then they would go back and make changes and compete as many times as you wanted for the sales letter to get better. So that's how it works today. And so Marcus is asking, how do you put your own judgment in there? And the way that you'd put your own judgment in there is that you really, in that particular case, Marcus, you wouldn't necessarily want your own judgment to be there. Just like when I'm hiring a copywriter, I want their judgment writing the copy, not my judgment. But my judgment to think about whether it will work or not, whether there's anything really bad in that. So there'll still be some elements of my imprint behind it, but it would only be as a Like, as it's moving towards production, I would still let the arena award the winner. But I also want to be clear that, like, I don't want anyone to think that I've got it all figured out because I don't have it all figured out. Like, I think that, like, one of the things that's really cool is, um, I was talking to ChatGPT about this yesterday. In fact, and I was one of the original buyers of the Oura Ring, like when it, when it first came out. I don't remember if it was like a Kickstarter or an Indiegogo or whatever it was. I bought it like at its infancy. I bought it so early that I never have to pay a monthly fee, right? And I get like offered upgrades like every time they have one. And I was talking to ChatGPT because like in the last year or so, I feel like I've aged more. And we were talking about the Oura Ring and how I've had it for close to 10 years now. I've never taken it off. So the amount of biometric data I have over the last 10 years is insane. Every night I went to sleep, how many hours I slept, you know, how much I exercised, et cetera. And having all that information is so valuable now as I wanna look at like what changed in my life. Having something know how you decide. And be able to sit in your shoes is no small task. It's a huge task, and we are at the infancy of it. In fact, humanity has never been able to do it except through apprenticeship. So this is an entirely new idea. And when you think about it that way, and I hope you do, humanity has, has advanced so much in the last several thousands of years because of our ability to, one, have language, to capture that language, to be able to record it and print it and share it. And so knowledge has compounded over the several thousands of years, but judgment hasn't. And maybe now it can. And so the first step is the imprint, but understand that you don't like set up the imprint today and now you have this computer that knows and can sit in for you tomorrow. That like, That doesn't happen. It can happen a lot faster than an apprenticeship can happen, but it still takes time. And so it is so valuable. Now, to just that, but then there's this added component. And if you've read the manifesto, and I sure hope you did, it is, it's eye-opening. And I promise you, it will shift your perspective as it relates to AI. And you can get that at strategicprofits.com/manifesto. No opt-in, no nothing. Just share it. And if you get value out of it, I'd love you to post it somewhere and share it. Um, let's say that your judgment is flawed. I would say that my judgment is flawed. And why would I say that my judgment is flawed? Because I have baggage. I'm a procrastinator. I, you know, I'm a little flaky at times, ADD, et cetera. So what I've done is not only capture my judgment and my decision-making, but also be very clear with AI what is ideal and what isn't so that it can help me actually go in the right direction. This is the beginning of this whole line of thinking. And the whole— so there are short-term benefits of having this, like you don't have to babysit your AI as much, right? Your AI can fill in and therefore Like require less of your attention, time, energy, and effort. You will have more confidence in its ability to execute for you, but that will be incremental. It will not be like revolutionary. It won't be like automatically overnight, like holy shit. But it's so worth it anyway. And if you're going to be working with AI. Then it certainly is something that you're going to want to do. That's a great question, John. So John asked, how do we sharpen our judgment? And I think that that is a great question. And I want to think about that because I've never I've thought about how to learn better, how to like, you know, develop better distinctions, but not exactly how to improve your judgment. I'm gonna revisit that, John. I'm actually gonna look into that like when we're done today. But good question. How do you keep the judgments realistic and not pretend in the copy arena? I'm not sure where you're going with that. I'm not sure where you're going with that. There are claims that are made in copy that are backed up by proof. Um, so I'm not sure where you think that, um, judgment's being realistic. All right, let's see. Okay, Rich, how do you capture judgment if it's fleeting and can only be experienced in the moment? Great question, Sean. So, okay, the same way an apprentice would. So what made apprenticeship work, or what makes apprenticeship work, is proximity, being around the expert, witnessing enough instances of how they make decisions and what are all the exceptions and what do they notice for, what do they not notice for, and all the decision-making criteria. So for all of time, that's been the only way to capture that. And what did I— that's the only— that's been the only way to capture that. But understand that one, right? Like, so, okay, first, Let's say that you're prompting and you're prompting correctly. You're, you're giving information to your AI correctly, which is very outcome-oriented. So you're less concerned about what the AI does. You're much more concerned about the output it gives you and will it do what it is that you ultimately want. Now you're giving feedback. Okay, so there is a hook that fires every time you press enter. That is looking at what is being sent back to the AI, and it is determining whether that's feedback or not. If it's feedback, right, it captures it. It captures that, and it captures the context window that it was referring to. So now you have your— like, that would never work to my audience. I would never say that because of XYZ.

Right?

That plus what you were talking about gets put together plus a bunch of other elements. And so now the whole decision has been captured. Now that goes to another AI, right? That then determines whether that really was feedback or not, and then looks and builds out the profile, then comes back to you to like, is this right or not? So that You express your judgment, right? Like whenever you're reacting to anything, that's what I mean by it's fleeting and it can only be experienced in the moment, but you can capture it because AI is the first tool that's always there and knows exactly what it is you're referring to. It can capture that. It can capture the whole process. And for those of you who are not going to be on the challenge that starts today, You can go to GitHub and get the imprint. It's there for free. This is— I know this might not be normal in our market, right? To have written a report requiring no opt-in, to create the software and give that away for free on GitHub. And there's no launch behind this. This isn't all like the free content leading somewhere. This is that profound. This is why, like, and this was, if you were around 20 years ago, this is how it was when I was talking about attention back 20 years ago. And the people that got it did extremely well. And that's what brought me close with Gary Vaynerchuk and everyone else, because that was what the people who got it could see very clearly what most people couldn't see about attention. And I, it's now The same about judgment. So let's see, what's this? There, there are skills, so there's a framework. You can have three rounds of judgments and appear valuable, but are not deep enough. I've tested and ran into this. I.e., I know what a C.J. or a B.J. real copy would be. Yeah. Okay. So you're, what you're asking me, Mahatma, is, are you asking me about the imprint and how it relates to this? Or are you asking me about how to improve the sales letter arena? If you're asking about how to improve the sales letter arena, then I would have a deeper conversation with either Codex 5.6 or Opus 5 or Fable 5. Um, and have a conversation about that individual skill copywriter and how to take it deeper. And you might wanna ensure that when you're running the arena, that it's capturing the thought pads, because those thought pads will actually give you insight into what the agent was thinking. I think you are confusing judgment with a skill, and they're very different, right? And because a skill is a method that is a sequence of steps, a judgment is, can be captured. You can't capture dead copywriters. Preferences, you can model their writing, and it can be modeled better is what you're saying. And I would agree with you, but it's modeling it better. It's not creating judgment. Judgment right now does not exist in AI. We are, we are on the edge, the frontier of this. So hopefully that makes sense. I guess there's like a lag. In the response, like, you know, and thank you guys for being responsive today, a lot more responses. So I really appreciate that. Um, any other questions? Cool. So Mahatma's got it. Any other questions? I gotta get this setup working better because there's stuff— there is stuff I really want to show you.

All right.

So for those that got on a little bit later, we were talking about the Judgment Age and why it's so important and how the Imprint and Atlas work together. Um, so Biggie asks, is— are Imprint and Atlas a complete memory system, or would I still need or want additional layers? No, you'd still want additional layers, Biggie.

This was the question.

Um, you would, uh, Now, if you are part of any of our products, like if you're Zenith Mind Elite or anything like that, then you have everything that you would need and you'd put Imprint and Atlas into it as well. But if you don't, um, Imprint and Atlas can work with any other system out there. In fact, Atlas has already got connectors for, uh, Hermes, for Open Brain, for G-Brain. And I would imagine that you could do the same with Imprint. So these are layers of memory. This is not the main memory layer, which you can get pretty much anywhere, and that's really where people are fighting. As far as the judgment layer and Atlas, there's nothing— there's not right now. There is no— there is no other Imprint. There's no other version of it. There's nobody that offers anything like it. And same with Atlas. And I don't say that as like a proud thing. I say that as I hope that changes. I really hope that other people pick up the ball and run with it. Just like when I invented the automated webinar or the free lead magnet or the drip content system. I mean, we were the first to do all these things. So this is the same, right? Like I'm the guy that creates and sees the vision and then I move on. And so my hope is, is that There are much better tools that come out over time, but yeah, you'll need a base layer of memory. But the good news is, is that every desktop app offers that now, and every system out there is plug and play with that. So that's relatively easy, and the Imprint and Atlas kind of come on top of that. Uh, let's see. Yes, I'm not a PPC manager and do not run or understand PPC management for our business.

All right.

So, uh, I'm not sure where you're going with that, Kevin John, but, um, the— so are you asking that in relation to the Imprint and Atlas? I hope not. Um, the, uh, because you'd need someone's judgment, right? And so I talk about this in the manifesto though. If you could identify who is a great PPC manager or someone who's famous for being a PPC manager and did deep research on their methodology and their thinking process, et cetera, You could create a very good skill. It's not their judgment, but you're trying to clone best practice as an agent. And that you can certainly do. Now, the imprint is what captures— is the imprint and Atlas the same? Now, imprint captures your feedback, right? So that it can build out your judgment. Atlas keeps a record of all your beliefs. And anytime anything changes, it looks for what else might change and surfaces them. so that you can decide whether these beliefs that are kind of layered, of whether they're impacted. And no other memory system looks at the, like the blast radius of a, of a belief change or a preference change. I wanna see the keywords in PPC campaigns visually in a map. All right. So then you would just ask your AI to do that. I don't know what, where you're going with this, Kevin. If I wanted to see the keywords in PPC campaign visually in a map, I would just ask my AI for exactly that. Um, hey Lisa, can Claude help us build and learn to interpret the point of campaigns, the results of performance, and make inference and insights for management actions? Yes, it can. It can do it better than a human, and it can do it, and you can set it up in about 20 minutes, uh, Kevin. So it's relatively easy. And instead of asking me these questions, you know who you should ask? Take a guess, Kevin, who you should ask. It's not a trick question. Since I don't know how long the delay is here, we won't necessarily wait for Kevin's answer. Kevin, tell Claude that— yes, exactly, Mahatma. Tell Claude, ask Claude the exact questions you asked me, and it will take it from there.

Yep.

Yep. Yeah, you guys get it. So there you go. I really would love to tackle something so we could talk tomorrow about how to just improve your judgment. I've certainly read a lot, taken a lot of notes on that. Anyone else have anything? That they would like me to focus on tomorrow. Maybe we come up with the idea today. We can rock on a little bit better tomorrow if we know what the topic is already. Okay, let's see. How do I— how do I do the imprint if I want to draw on all my past and current strengths on my zone of genius if I'm not currently in a role or business to capture input from? There's quite a few ways. If you've been using AI for a while, your AI keeps a transcript of everything that you've ever said to it. You could have it go through and look for times where you articulated your preferences, where you articulated certain feedback, And then have it find all of those and then sit down with you and discuss them. And because you don't want anything thrown in, you need to vet every single thing, because if anything's wrong, it throws the whole thing off. But you can do it that way. And in fact, I've done certain like import jobs like that because, you know, the imprint is relatively new, but I've been doing a lot with AI for years. And so I want to go back and capture that. The problem is, is that Any output could— like, lots of different judgments could lead to the same output. So the only way that you can, like, capture anything from the past is by going through all the transcripts. So hopefully that helps, uh, W. Scott. And first you'd have to download the imprint, which you can get at GitHub. Um, and now you can also— it's also a pip command where you can just pip Yeah, no, that's I think for Atlas. I'm sorry, Pip is for Atlas. But you can just go to GitHub and grab it, W. Scott. All right, I have one vote for how to improve judgment, like your own judgment. Anyone else have anything specific they'd love to work on? All right, well, okay, let's see. What about the success principles behind good judgment? They seem to be gold to me. Have AI learned the principles behind it? How can you tell when your judgment is superior to AI's as AI keeps improving the graphs change, or it's having the system to build better judgment in a current field with research training videos? AI will understand and be able to act. Paul, great question, right? But I think it's a little bit different than that. I think I'm just going to get rid of the question so I don't have to like Leno. I think the— instead of judgment for a second, let's just use point of view, right? Like there really is no centered point of view of the AI. And so that is what we're talking about. And AI has every judgment, every point of view, right? It doesn't. And we're looking for points of difference that allow us to have something that no one else has. And that's where the imprint and capturing our judgment is really valuable and capturing our point of view and our taste. And there are many ways to go after a market. There are different types of customers. Which customers do you want? What do you sell them? What type of product do you start with? These are not black and white answers. There is no one right answer. And so these are the kinds of things that really are determined by your perspective, your point of view, your judgment. Hopefully that helps, Paul, and I'm sure we'll have more discussions about it. All right, so Ruth, forgive my ignorance, but I am new to this environment. No reason to apologize, Ruth. Are there resources available on how to work with the tools in GitHub, like a step-by-step guide for a newbie? Thanks in advance for the guidance. Ruth, great question. You don't need a step-by-step guide because you have your AI. So you would have your AI, like if you're using a code, like either Codex or Claude Code, you'd say, I heard about this memory layer for AI that Rich Schefren has. Can you search it on GitHub? It's called Imprint and it's by Rich Schefren, right? It'll find it. And if you don't find it, then just do a search and grab the link. And then ask your AI to check it out. What does it think of it? Um, have— ask it to install it, teach you about it, tell you what you need to know to use it. You know, the AI that you have in front of you, Ruth, is the best teacher, especially when it's something that you want it to do for you. It will be able to teach you and do it for you at the very same time. So that's what I would suggest, Ruth. What is same and different in your eyes about human point of view tendencies versus AI tendencies now and next? Well, AI is always changing, right? Like, that's why sometimes it feels like you have the smartest idiot assistant in the world. The perspective is always changing and it really doesn't have a perspective. It doesn't hold firm on any point of view. It will bend very quickly if you push it. So I would say that is definitely a part of it. But this is a bigger question, Miguel. Probably better, Miguel, to ask support on that. That would take too long for me to answer. How do you confirm that your prompt is well detailed that AI will give the appropriate answer so you don't get a faint or surface answer? It depends on a bunch of different factors. It depends on a bunch of different factors. One, are you giving it the right context? Does it have your preferences for deep answers? Do you have any kind of critic or checks that come back to make sure that the work was done? You can do contracts to make sure, like between agents. So there's a lot that you can do and it really depends where the challenge is showing up for you. But that's a very easy challenge to deal with. And one, like, I don't know which AI you're using, so I can't speak to it specifically. If you're finding that your answers are too shallow, I would just press it to say, where did you get that answer? I need them. I need you to do a lot more work. I need you to take it a lot deeper. I need you to impress me with what you find because this is very not impressive. That's exactly right, Lisa. And then have it start building out your advantage and capturing your perspective. All right, I was wondering if it'd be a good demonstration project to serve as an example of something to build and use judgment. Oh, that's what you were talking about. Okay, so like build a skill that would manage a pay-per-click campaign, is that what you're talking about? All right, that's at least an idea. It's better than nothing. I will write that down. So we have one idea is judgment, like just how to improve human judgment. The other is how to build like Either a skill or an agent, judgment, and then the other is skill or agent to manage pay-per-click. Got it. Cool, cool, cool.

All right.

All right, guys. It's 12:36. I'm gonna jet because I got an interview to do. I have the imprint to do. I have a Q&A call to do. I'm a busy boy today. So tomorrow we're either going to talk about how to improve your own judgment, a better explanation of the difference between human judgment and AI judgment, or we'll build a pay-per-click— Kevin, you're going to have to remind me. Oh no. Okay. Keywords and targeted I'll take a picture of it.

Hold on. Picture.

I didn't take a picture.

Ah, there we go. All right, there we go.

Um, so, uh, yep, cool. I will take a look at that. And thanks for being here, guys. I will see you tomorrow at noon on Facebook, on LinkedIn, on YouTube, and Instagram, and we'll see if we can get X back up. And until tomorrow, to higher profits beyond, Rich Sheffrin over and out.