In this August 20, 2026 livestream, Rich Schefren opens by apologizing for missing a prior session and greets a small live audience. He shares observations from recent work with marketing legend Jay Abraham and his own team. A central technical insight is that the newest, most powerful LLMs (e.g., Codex 5.6 Sol Ultra, Fable 5 Ultra, Opus 5 Ultra) are less reliable for long‑run, instruction‑following tasks, though they excel at high‑level planning and problem‑solving. Schefren advises users to match model tier to the task rather than default to the top tier, warning that over‑paying for the most powerful models can reduce consistency. He then outlines his strategic AI positioning: focusing on AI that enhances human judgment, tacit knowledge, and self‑knowledge, and aims to be known for that niche. Schefren discusses business goals for 2027, emphasizing the need for at least one evergreen funnel feeding the Zenith Mind product, launching 1‑2 front‑end products at breakeven, and leveraging an 80% cross‑sell rate to drive growth. He mentions the upcoming "Connect the Dots" event with Jay Abraham, the expanding public‑facing role of Ashley on the team, and how the founder’s day‑to‑day involvement is decreasing, giving him more freedom. Personal updates include improved vision in his right eye and a scheduled cataract surgery for his left eye, as well as increased rope‑jumping activity. He closes by reiterating the model‑selection insight, the December‑31 2026 planning question for 2027, and wishes the audience a productive weekend.
Chapters
Hey, hey. See a few people. Hey, Mahatma. Hey, Sean. Hey, Big. And I am— I'm looking at yesterday's. That's really weird. How bizarre. Yeah. And it's saying, oh, this is yesterday's livestream. I wonder if anyone will be on today. That's so bizarre because it says live August 20th. Huh? Anyone there? Let's see. Anyone? Because I have stuff I want to share. Anyone there? Yesterday, there was. Yesterday I got so busy with Jay, I completely spaced about this livestream. So I apologize if you guys were on and were waiting and I never showed. The— hey, Mahatma. Oh, cool. All right. So we got a few people here. And I'm outside. I'm sweaty, as you can tell. I have spent the last 2 days with Jay and worked with his team as well. His team is pretty far along. There's a bunch of work to do with Jay. And I think at first, it was more showing him what's possible, how helpful it can be for someone like him and I. And so that was kind of interesting. And but I had a really, really productive beneficial and useful, um, session the other day with one of my top groups, uh, Maximum Profits. And, uh, in Maximum Profits, uh, a couple people were having this challenge. I was having this challenge too. And, uh, I'm curious how many of you have this challenge. Uh, hey, John. Um, so tell me if you've noticed that the new models Uh, you know, Codex 5.6, Fable 5, Opus 5, if you've noticed that they're a little less consistent. If you can let me know yes or no. Yes, you've noticed that they're a little less consistent. No, it seems, seems better than, you know, 4.8, 5.5, stuff like that. So John, Kevin, Mahatma, like, tell me what your thoughts are. Okay, so Mahatma says yes, even Opus 4.8. Codex is just killing it right now for me. All right, John, are you using 5.6 SOL or are you using something else? Got it. All right, so now let me see if I can— I want to see if I can upload a PDF. Let's see, your computer. All right, currently PowerPoint PDF. Okay, upload file. Let's see. All right, so I downloads, um, let's see, the decision cheat sheet. Uh, can I do more than one? No, I can only do one. All right, so decision cheat sheet. Do that one. I'm gonna do one more here. My computer upload. The big idea in plain English. Muhammad said I just had to retain on basic judgment. It doesn't seem to have sound logic. That's Definitely the case. I'd love to have a look, Rich. I'm not certain. Got it. Okay, so a bunch of us have experienced a lot less reliability, but what we found was, one, it was really reflective of how someone was using it. I was guilty of it myself too. Those of us that were really using the top models for too much were the ones that were experiencing a lot of these difficulties in consistency and reliability. So I did an extensive amount of deep research on it and found this. I want to see if I can make this bigger. Hold on. Can I? No. Let's go back. Let's see if we can make this bigger. Can I? How do I make it bigger? Can you read that? I don't know if you can read that. I'm going to remove myself. Maybe if I remove myself from this page, let's see, maybe I can make it bigger if I remove myself from this. That, um, I haven't heard that advice shared anywhere. And so— oh, if you were talking— oh damn. Well, did you read the page? Did you read the page? Can you hear me now? Can you hear me now? Hello, hello, hello, hello, hello. Let's see. Can you hear me now? Can you hear me now? Hello, hello, hello. It's good. All right. Okay. So what you're not hearing anywhere that I just wanted to convey today, and that's really the Sound is going out. Okay. What I wanted to convey to you is that for long runs, the top models are actually less reliable. And when I say top model, I'm saying like using ChatGPT 5.6 Sol Ultra. Right, or Fable Ultra. And those are great for high-level thinking, high-level problem solving. They are not good for long runs. They will do everything wrong. They will not follow directions. And it's the same for— so it's what I'm talking about is Codex 5.6 Soul Ultra, Fable 5 Ultra. Opus 5 Ultra or Max, but whatever the top settings are, don't think because it's important and you really want it to be awesome that that means that you need the top model. That's what I thought, that's what some of my clients thought, and that was a mistake. So we were spending extra money for less reliability. For whatever reason, the top models are better at planning and thinking and worse at at following instructions. So that's that. Um, the time with Jay was really, really awesome. I think one of the things that we worked on together was really figuring out what Jay is all about. Um, where should he come up if someone's searching in AI for him? And that's a big discussion that hopefully each and every one of you have thought about. I know that I've spent a lot of time thinking about it and didn't do any kind of large language model optimization, AEO, because I didn't know what I wanted to come up for. And now I know what I want to come up for. Now I want to come up for AI as it relates to judgment, as it relates to tacit knowledge, as it relates to self-knowledge, and those 3 coming together to drive AI. And so that's where I'm spending my time, and that's where I want to be known for, and that's where I want to surface. And so we were working on that for Jay. And, you know, if you're not someone that AI recommends in your area, that's problematic long-term, and you might want to try and fix it. Sooner rather than later. And, uh, anyway, so that's that. If you didn't know, this is my backyard. I don't know if you get— oh, well, no, the camera's behind. I thought the camera's in the computer. My bad. Um, and my whole team is upstairs right now because we have, uh, we have the, uh, Jay Abraham Connect the Dots starting, uh, tomorrow. Did Jay— let's see here. Did Jay add any new perspective to how you're approaching your own vision, or was it just focused on helping Jay? It was primarily, uh, focused on helping Jay, but, um, I know what Jay wants for me. It's the same thing that— it's annoying, hold on. Uh, it's the same thing that, uh, that Mark Ford also wants for me, that all my mentors, they all think I should be spending a lot more time with bigger and, uh, smarter companies, and, uh, that I could have a bigger impact with bigger companies. And, um, that's something that I've been told my whole life and, uh, didn't really have anything I wanted to share with big companies. And the big companies I've worked with, I haven't necessarily enjoyed Um, but I, but I'm in a better place than ever before for that to actually happen. And so actually, I'm really excited about my team talking to Jay because Jay and I share a lot of qualities and characteristics. And one of the things that I'm really excited about and why I say now is like a good time for it is You know, we have like a core— we have a core positioning of what we're focused on as far as understanding AI and AI knowing you better than you know yourself, understanding your judgment and your decision-making and your tacit knowledge, and then applying it to business, right? And so that's what we're all about. And what's great is, is that ZenithMind, Connect the Dots, uh, Force Multiplier— none of these things require me. And, uh, my team is Does most, if not all, of the heavy lifting at this point. And that's more so today than ever before in my business life, which means that I'm less necessary for the day-to-day. And because I'm less necessary for the day-to-day, that gives me more freedom. So one of the things that— and I guess I'll end with this today, and it's something that I think you might want to think about for yourself. That's so annoying, this flashing light. Hold on a second, guys. What the hell? I'll just turn it off so I'll be more lit on one side. Um, Nicole, who runs the company, um, put out a really good question. It's a question we use for all our clients, but something that we haven't used very much internally. With our clients in our different programs, one of the questions that we like to start the programs with is what needs to be true when it's over, whether that's in 2 days from now, 2 weeks from now, or 2 months from now. What needs to be true that isn't true right now to make this the best investment of your life? And we work very hard then to deliver that to people. Um, the— but Nicole sent out this memo about planning for 2027, and it's like, what has to be true By December 31st, 2026, in order for 2027 to be this number, the best year yet, so on and so forth? And I think that's a really great question to ask yourself. Like, for me, when that question was asked, it was like, okay, well, we need at least one evergreen funnel that feeds Zenith Mind. And that could be an evergreen imprint, it could be an evergreen webinar, but something there with 30 days of, uh, sales data behind it so that we know like what we can afford to spend on advertising and acquisition there. And so either a Zenith Mind webinar or an imprint challenge, either one loading into Zenith Mind. And then we have a bunch of front-end products that the team has been working on that I would like to see at least 2, 1 or 2 of those, uh, out deployed and, um, acquiring at least at breakeven. And so that's what the team and I are working on to make true, um, before December 31st. And, um, everything else in our business is like working really well right now. We have over 80% of the people that buy Zenith Mind end up buying other programs from us, which is an insane number. And so knowing that if we had more acquisition, we'd be even bigger and more successful, that's kind of where we're moving towards. So I have the team here today, uh, as we prep for Connect the Dots tomorrow and Sunday. Also so Ashley and I can record some videos because Ashley will be taking a bigger role and being part of the face of the company so that I don't have to be on every webinar, and, uh, or at least not lead it. And, um, just a super exciting time. And anytime I spend with Jay always like gets my mind opened more. Uh, one of the most creative, smartest men I've ever met in my life. And, um, yeah, so that's that. And, uh, I will, uh, I hope everyone has a wonderful weekend. Um, oh wait, one more question. Let's see. How would you compare yourself today? Let's see. Um, I came into January of this year with lots of questions about like what we would be doing, how we would be doing. I think we started Connect the Dots this year, and we did Force Multiplier this year. So it's been a great year. I also have my vision back in my right eye. So, and September 1st, I have the cataract operation for my left eye. So then I'll never, you know, so I had a short stint wearing glasses, but never need to wear glasses again. Um, back to jumping rope more so than I was in January, so that's good. Yeah, so I'd say there's been, uh, a steady improvement, uh, in my life. I'm older though, so that's the only part that's worse. Uh, anyway, so I hope everyone has a wonderful weekend. Thanks for tuning in. Sorry about yesterday and the day before, I think, um, because of Jay being here, I just totally lost track of time. We'll have a great event here for the next 2 days at Connect the Dots. And, and then I'll be back here on Monday. I'm still like playing with this idea of whether I should be doing these or not. My AI and I are going back and forth of whether this is a useful amount of time for me or not. And, but thanks for playing along with me. And thanks for being here. And thanks for listening. And I hope you got value today. If you've been using the top models for everything, you did, because that was a big aha for all of us. And, uh, also think about what needs to be true by December 31st in order for next year to be the best year ever. And with that, great weekend, higher profits. Uh, talk to everyone on Monday. Bye.
