Rich Schefren Sharing AI Breakthroughs 5/30/25

Welcome. Hello, hello, hello, everyone. Just getting Instagram to go live as well. And then we are good to get going today. So welcome. Cool. All right. I'm on Instagram as well. Welcome. And if you're just joining me, which I just started, so obviously you are, my name is Rich Sheffrin and been marketing online for about 25 years and just have started this live stream where I'm really talking about using AI to improve human performance across the board. And I'm motivated to do this primarily because I launched a program at the end of February, beginning of March, called Zenith Mind os that helps people get their AI to know them better than they know themselves or better than anyone knows them. And because of that, then be able to work with them to create improvement and to move closer to your ideal life every step of the way. That's what my program does. And that's why I want to talk about AI to improve human performance. The first thing today, that in the news, that was kind of a big deal and not sure how many people saw it or not, but Perplexity. Perplexity was like a great search engine, AI search engine. But just yesterday they rolled out a, a tool creator, right? Like, so you can create a dashboard, you can create anything that you want. And so yesterday I played around with two different, two different types of agents really, or actually I started a third too. So I'm going to tell you about that. But some really, really interesting things. So first and foremost was Perplexity. And you know, I don't know about you, but for me, right, like, you know, AI has been around for a while and I finally have like developed some pretty impressive prompting skills, self taught. And my prompts are pretty intense, but I haven't really gone past that. You know, just these big mega prompts. They could be like seven pages long and they can do amazing things, but haven't really taken it further really with that much other than doing deep research for these types of live streams. And I'm going to share a bunch of that research with you today, but I haven't really taken the next leap. And the next leap, I think, is creating your own tools that run on your own system. You know, there's been a big explosion of vibe coding, but now you can do it in almost anything, right? Like maybe not full stack kind of development like you might with some of the tools that are made for coding, but you can still do quite a bit. So yesterday what I did was I started thinking about where do I sometimes waste time, right? But waste time in A way that actually leads to something beneficial. But if I could figure out a way to do it without having to spend the time, I'd probably be better off. And so one of those things is looking for mega prompts, right? I will oftentimes, like when I'm on substack, I'll look for prompts. When I'm on Reddit, I will look for prompts. And I decided that that's really not a great use of my time. So yesterday I went to manuscript. No, not Manus. I'm sorry, I went to genspark. Let me go there for a second. That's not how it's supposed to look. Let's try that again. Hopefully I can show that was the plan today. Hopefully this works. Why would that. Oh, okay, wait, I'll go here. Go back to this for a second and then I'll go here. Genspark. Where's genspark? I don't see it. Google Chrome. Genspark. All right, there we go.

All right.

Hopefully you see that? And by the way, if you're watching, please say hi. Just curious to see where people are watching from. I see Lorna's watching on YouTube and I see that I'm live on Instagram, but don't see much more than that. So would love to know if you're watching me, where you're watching me. Anyway, you can see here I wrote, hi there. If I want to scrape the community on Reddit prompt engineering and find every post labeled Prompt text, showcase and capture the prompts given in these posts, what would be the best way? So it starts really simply. What I keep teaching people with AI is that you don't need to know how to do anything. You just need to know what it is you want. That's what I wanted. I wanted a way of scraping this on a frequent basis. Then it starts giving me different approaches I could do. I said, I'm running a Mac studio and I'm somewhat technically challenged. Can you walk me through step by step using the Python approach? If you think that's the best way to go, Then I start creating all of these. Every time you see me, these are screenshots telling genspark, right, what it was that I was seeing on screen. So, like, I would. Genspark would say, okay, like, do this, right? Then do this, then do this, then that, then that, right? And then, you know, do all these things. And then I would, right? And then this is like, I'm on Reddit and I'm showing what I see on Reddit, right? And so it knows my credentials, and so it keeps the programming going until we get to the end. And in the end, now, every day in my. Well, this is twice a week in my documents, there is a text file of all the different prompts that are being shared on Reddit, right? And so, like, that's one of the first things that I've ever created. And I just wanted to show you again how it all started. The only thing that I did was, hi there. If I want to scrape the community on Reddit or prompt engineering and find every post labeled right, what would be the best way? That's all you need to do, right? And so I think that, that, like, I'm sharing that because some of you are like, yeah, no shit, Sherlock. And I'm sure there are some of you like that. But for me, this was really like the first time that I actually created something that now runs in perpetuity outside of my system, finding stuff for me, bringing it back. And I went to Perplexity and did the same thing. This was for something else that I'll show you a little later. I'll show you this. I would like to create a tool that surfs the web daily to find the best mega prompt shared on Reddit, substack, Twitter and everywhere else. Great prompts get posted and collect them for me each day. Now it actually creates a diagram of how it's going to work, right? That's the diagram. And then it's explaining how to do it.

Right.

And it's giving me the different elements that I'm going to need. And here it is, right? So this is a sample daily report. So collection by source, right? Chat GPT prompt. Genius. There were 15 prompts. Three were high quality prompt engineering. There were eight prompts, four were high quality. Do you see this?

Right?

And so that is like, now I can do it where I can have this running in the background and I can actually really just start to take activities that have been like, wasting my time and actually spend that time on more productive things, eliminate the likelihood that I would get distracted on those platforms and then start surfing other stuff and waste time and. And I don't have to do it anymore because I know, I know that I'm capturing everything. So I wanted to show you that. Now, the next thing is that I haven't really been using meta much. I don't know if you have. So every day I'll show you this real quick here. Every day, these reports are all generated for me at night, and all of these reports are deep Research about what has transpired over the last 24 hours as it relates to AI improving human performance. And so the. The first thing was that I read in some of the research. Well, let me just kind of like now take you through some of this. All right, so I'd probably like to go in order, back to front. But you can see, like, all this research that would have required a ridiculous number of people and time to create is now done for me completely on autopilot. So let's like, kind of like I'll show you what man has found for me yesterday, and hopefully I'm here, right? Yeah. Okay, so you can see here, right, that yesterday I was sharing some research that showed that, you know, we have to be careful that we don't have AI do everything for us because our brain could atrophy because we. We stop spending time critical thinking. Here's though, a different research study.

Right.

And here, after just 25 hours of training. Let me see if I can make it bigger for you guys. Hold on. Yep.

All right.

After just 25 hours of training with AI systems, participants showed measurable improvements in cognitive performance daily, directly contradicting the popular narrative that AI makes humans dumber.

Right.

This research suggests that the future of cognitive enhancement lies not in AI replacing human thinking, but in creating a symbiotic relationship where each augments the other's capabilities. And that is what I try and teach in Zenith Mind os. It's about developing a copilot, a copilot that knows you so damn well that all the ways that you weasel out of work or all the ways that you don't get done, what needs to get done are kind of blocked in advance and at least for me, and keeping me on track. So that was one from Manus that I wanted to share. Another one that I wanted to talk to you about was this, which I found really fascinating. And this is really interesting. So where is it? Right here it is. Okay, so another. This research came out March 23rd. So it's two months ago, but still kind of interesting. And it is that by using natural language processing, Right. With wearable technology can be extremely useful for mental health support. So what does that mean? Well, like, I have here the limitless pendant, right? This records me 24 7. I also have here. Oh, I guess I'm charging it right now. Yikes. So I'm not wearing it, right? Oh, no, I am. I'm wearing right here a B computer. And this records also everything. It also tracks my emails, everywhere I go, et cetera, and the Battery here lasts a week, so it's really useful. But what this research is about is really fascinating. That by listening to what, how like, the subject, like me in this particular case, speaks, right. My language patterns, these systems can identify early warning signs of mental health deterioration with unprecedented accuracy. There's no doubt that right around the corner from this will be a wearable that does natural language processing, that will pick up on Alzheimer's, right. When you start forgetting more than you used to, when you seem confused. And it will know that based on your language, which is pretty insane, I think. I hope you guys. Oh, YouTube. Where can we learn about Zenith OS? You know, I have a link for that. Let me find that real quick while I'm talking to you guys. I know that it's somewhere in here, I think. Let's see. Let me just pull this up and then I will. Let's see. Let's see. All right, I'm looking for. It's never easy to find these things. You know what? I'll do it later or I'll do it in the next week, but you can go to. Let me just see here. I should be able to do it right here. Hold on. This is where, like, having a producer would be nice. Or anything other than just me. All right, let's see. I'm looking for the link. Damn it. Can't find it. All right, so I will get back to that. Next is though, I thought was kind of cool, too. I'm just scrolling here to show it to you. Hold on.

Was.

Neuralink. Neuralink had their third patient actually get the brain chip. And the latest patient reportedly achieved control of digital interfaces within minutes of activation, demonstrating significant improvements in the technology's integration and speed. Which is insane, right? I mean, it's totally insane. The whole idea of a brain chip actually being real and actually being done and actually working. So that's what Manus had to share with us today, if we go to Claude.

Right?

The first story is about Perplexity, and it's called Labs. I tried it today, right? And I showed it to you here, right? And that's the tool, and that's the design for it. And. And all you do is you click that little button here. I don't know if you see it here. Like, that's. It's called Labs, and allows you to create programs, stuff like that, from scratch. So this represents a shift from AI assistants to AI workers that complete complex tasks independently. So that's the first thing. If you've used Perplexity in the past, what you can now do with Perplexity is game changer. You can actually start creating yourself tools. And those tools I would recommend think about, like, the things that you do that are somewhat productive but are kind of a waste of time too. And those would be the things to automate first, to kind of get them off your plate.

Right.

The next thing, the reason I was showing you Meta earlier is because Meta hits 1 billion users and it's kind of crazy, right, that you can't use meta yet, or at least I don't know how to. To actually extract everything from like my Facebook and my Instagram to tell me more about myself and the people I talk to and stuff like that. But right now, I don't see that happening. Hey, Sam. Oh, I owe you a call. I never launched any of that stuff. That's why I never called you because I really wanted to focus on Zenith. But I'll. I'll follow up with you next week or over the weekend. Sam, if you're around the. So that's Perplexity. Okay. That is Meta. Was there anything else I wanted to share? Okay, there's this big push now with wearables, right? Oh, okay. Let me go here. This was interesting too. Sorry about that, guys. Got to jump back for a second. Okay. Yeah. This next piece came from Claude, right? And it confirms that AI saves about 5.4% of work hours for the average person. But most people are not average, right? And if you are a highly skilled worker, right. A 40% gain. And GitHub data shows an 88% gain for developers using AI. So the amount of productivity gains are starting to compound. And they're compounding for people with deep expertise, because with that deep expertise, they are able to be able to command the AI to do things that you and I wouldn't ever think of.

Right?

Like, I was talking to a gentleman that owns a production studio and he was talking about, you know, what's going on with AI, and he's got these editors. And I'm like, well, you have to understand that these editors know a lot more than the average person. So the way they can edit with AI is going to be insanely better than the way an average person does. And they'll be able to do amazing things. And I hope that that's been your experience too. It's like you can generate. Anyone can generate copy, right, with AI. But if you happen to know copy really well, you'll be able to generate much better copy because of the distinctions you have and the way that you're able to instruct the AI. So not surprising, but nonetheless confirmation. Next story is all about this AI wellness revolution. And yesterday I showed you how I loaded in my blood tests, my supplements, and my genetics into ChatGPT. And it gave me a very comprehensive overview of my health and my biological markers, et cetera. I am very much into wearables. I'm very much into wearables that, you know, record, but then also wearables that track my biometrics. Like my Apple watch, right? Like my whoop strap, like my aura ring and other devices, too. And the reason why is it stuck with me since the very first presentation that I ever did on AI, the very first presentation I ever did on AI I did at Marketing Funnel Automation, Todd Brown's event, and it was in 2018. And I tried to implore people that, you know, the AI revolution is coming, but you won't be able to participate in it if you don't have your own data. And so everything in my business, as well as my personal life, I try to amass data, right, like, of what's going on. And so my body is no different than my business as far as tracking metrics and trying to improve. And over the last, I don't know, five years, my health has gotten better and better each year. My HRV is going up, my resting heart rate is going down, my body fat percentages are going down, the amount of lean body mass I have is going up. And so all these things, right, are because I'm able to monitor everything, and I know how well things are working for me. It's the same reason why I go for a blood test every three months and then kind of use the blood test with the supplements I'm taking with my doctor to adjust, right? So that's whoop is combined with OpenAI. And we were talking about Aura the other day because they have instituted glucose tracking and also meal tracking by taking pictures. So Aura now has this comprehensive view, right? The next thing, which really is just for the people that are much more into AI than probably you or I, is this. And it's all about that. Models are getting smaller and smaller, right? But more powerful and more powerful. So this signals the end of, well, this is what I'm looking at right now, Signals. I guess I can make this bigger, too. Sorry about that, Signals. The end of bigger is better era in AI development makes advanced AI more accessible without massive requirements, which means we're closer and closer to you running your own AI at home without any very complicated machines, which is kind of cool. I think. I hope you guys think it's cool too. And once again, if you're watching this, wherever you're watching this, let me know and just say hi. It's nice to say hi to people. You know, it seems like. Seems like Instagram's the least talkative, but I see people here now on Facebook, on YouTube. Oh, great. And on Twitter.

All right.

Trying to see if there's anything that I wanted to share with you from, like. Yeah, let's. Let's go to. I want to go to the chat. GPT research and it. And also you could look at how all these different ones are doing the research. It's kind of interesting, but Grammarly got a billion dollars in funding. You know, I've kind of forgotten about Grammarly. I don't know if you have too, with all the, like, advent of all the tools that have been out there. But. But that's a big number, a billion dollars. And I guess they're trying to stay relevant and so kind of interesting, but not a big story for me there.

Right.

Perplexity unveils labs. We're going to see that in almost every single report since that is what was announced yesterday. And notice how like in the research, right. Like it's, it's summarizing the story.

Right.

Executive summary, why it matters. Live stream angles.

Right.

Like if I wanted to share it, I could share. You know, start with this. Right. And slide content.

Right.

Perplexity, we talked about Maxium. I don't really understand this tool and yet it's come up in the research about five times already. And it's a. It's a site to attach your wearables to and your blood test. So I just did that this morning. Here it is. And you know, my max score right now is 55. I'm a horrible sleeper. My sleep duration improved by 1 hour and 9 minutes. My efficiency decreased. And this is like, I don't know why I would pay for this, to be honest with you. It seems like it's just pulling from. I can get the same information from my wearables. I don't know why I'm pulling in my wearables to this. Although it's nice to know that my resting heart rate is 56. So I don't see this as really. The whole promise of this tool is that you have like Olympic level coaches that have programmed this. So if you're trying to be in the best physical shape possible, this would be potentially what you would use as it integrates with all your wearables and it is kind of giving you a consolidated view. I wasn't impressed, to be honest. So maybe something will change, but right now, not really. That's maxium right here, right? I mean, it. It analyzes all this, but so does a lot of other tools. So I don't know why people would want that per se, but I wanted to test it out for you guys. And then the whole thing about Vibe coding, you know, I have a lot of friends that have already made this move and are starting to do more than what I just did with Perplexity or. Or with gemsmark. This is more like really coding and creating tools. And that's damn damn cool. So I'm trying to think here. Okay, let's see what else I wanted to share. I tried to go to context AI, but I couldn't make it work. So I'm not going to talk about that. All right, so I circled something here. Okay, so let's read it. Other tech content creators have been buzzing about AI's impact on personal productivity. Too many daily AI newsletters and blogs highlight new AI tools listed and recap big launches. Mainstream outlets ran features on how AI is transforming work and creativity from coding to art. Productivity gurus on YouTube, think Ali Abdaal, Thomas Frank have been sharing their favorite AI hacks for focus efficiency, while AI researchers on Twitter dissect the latest papers on AI driven learning and memory. Overall, the chatter converges on one theme, AI as the enabler for doing more with less effort. And I hope that that's been your experience. Too deep work with AI, but not distracted work. So can you work with it just intently? Right from Cal Newport, this is what we're building, basically taking a second brain and perpetual memory so that everything is remembered and everything you read is brought into it and highlight and study and stuff like that. And also all your personal information. That's what we're working on with Zenith Mind os and I'm excited when we'll release that. Then the other thing is, let's see, we have AIs like bringing in all kinds of data. We know this already. So let's go down to noteworthy mentions. Trusted therapists, right? They're beating real therapists, even though they don't want to say that. When you read the research, it shows. Now here we go again. So I was talking about this yesterday. I don't know if it's cross memory, cross conversation memory, but here we go. That Claude now has, with 3.5, has launched a 200,000 token context window, roughly 150,000 words of memory. This means clauds can retain lengthy conversations or entire books worth of information in one go for users. It's a step towards AI that can remember context indefinitely, making it much more useful for long term projects. Right, so that's kind of interesting. And then this is really fascinating. I haven't played with it yet, but you know, when I created. I've said this before, but I'll say it again. When I, when I created Zenith mind, I used ChatGPT because ChatGPT was the only one that had the ability to remember conversations outside of that conversation. But it looks like here that if I move this over a little. There we go. Following up on an early coverage of AI Chatbot struggling with context Google Gemini's AI is now able to remember past conversations to tailor its answers. Announced in a quiet update, it will bring chat history context into new replies. A quality of life boost for users tired of re explaining themselves. If this is go, if they're going to move more in the direction of chat GPT, then this is an insanely, insanely big update because what it means is, is that it can help you so much more. It's the reason why ChatGPT is so helpful. And so that is ChatGPT the next thing. And this will probably be one of the last then want to show you a few things. Let's go. The most comprehensive when I do the deep research, if you're interested in what deep research tool right now seems to generate the, the deepest work, it is Gemini. So Gemini, if I go here, Gemini, you know, gives me like something serious.

Right.

So the first thing that is really interesting here and I, I know this is going to be a much bigger deal over time for sure.

Right.

But I'm already seeing it. Is this.

Right?

Yep. Stark warnings from anthropop anthropic CEO about significant job cuts stand in contrast to a more optimistic narrative of AI. Now 25% of white collar jobs, like new hires for white collar jobs are down by 25% and that's not an aberration. You're going to see that grow. It's really scary, right, for recent college graduates because they don't have yet the experience necessary to get the best out of AI. And they don't, you know, they don't have the deep knowledge of the industry either yet. So this is problematic and I think that we're going to see a lot more of that and I don't think anyone is surprised by that. The next is what's called ambient AI and ambient AI is Kind of like this wrist thing I'm wearing that's recording all the time, or my necklace that's recording all the time. It's like not having to input anything, but yet AI is being constantly updated. So that's ambient AI. It's like you could have ambient AI in the office. It would know when I'm here, it would know when I'm working, it would know when I'm not working, when I'm not standing at my desk, so on and so forth. And that would make it that much better because it's expanding the context, right? And because of that it is. It makes AI seamlessly integrate into your life, right? So it's less about engaging with a tool and just having that. Now, what's kind of cool about the way that the Gemini does it is that, you know, okay, so the first big story was Perplexity Labs. We went over that, right. The next is autonomous agents kind of spend a little bit of time on that. Okay. AI powered smart glasses and rings, right? This is along the lines of getting that information. And no matter what, AI therapists are beating real therapists. AI redefines creative workflows, right? And then we have the actual story. So the first story is the Perplexity one. And I just want to share what it says on the bottom here because I thought that you guys might like this. See, like a small business owner could use Perplexity Labs to create a real time sales dashboard by describing the metrics they want to track the tool, then pulls data from their E commerce platform and visualize it all without requiring a developer or any coding experience.

Right?

So that's what. Like that. Like think about, like how you manage your business and what metrics you might like to see. And now you can use a tool like Perplexity or Claude to actually create that for you. So when I went to Claude, do I have cloud here? Let me go here for a second. And I asked it to create the same thing that I asked that I asked Jen Spark yesterday, but this time not just using Reddit, but using also substack and stuff like that. All I said was I'd like to create a tool that serves the web daily to find the best mega prompt shared on Reddit, substack, Twitter and everywhere else that great prompts get posted and collected. You can see now that it started to create all of the software that I will need. Then if I tell it that I'm very technically like an idiot, it will then walk me through exactly how to set it up. So let's go back to Gemini for a second and see if there was anything else I wanted to really share with you that was really like stood out. I told you about being able to just monitor language and be able to pick up on mental health issues. This is really. The wearables are going to get bigger and bigger that market for sure. And so I would say that while I could go further.

Right.

I think we've covered enough. We've done a half hour. I want to thank you for joining me. I gotta figure out a way to engage people on Instagram a little bit better because I see like I've got 30 people over here, but only like one or two on Instagram. So somehow, maybe it's because of the formatting of this. I got to figure out. Maybe I'll do it off a cell phone or something. But I want to thank you guys for being here. And if you're really interested in GEM in Zenith mind, let me share with you. I'm gonna find the. Let me see if I can find the link. So let's see here. I think it would be a little bit. Oh, there it is. Hold on. Okay, opening that up. Okay, everybody got big plans for the weekend. I hope you do. I'm pulling this up right now and I'll actually just show you the page if I can. There we go. Okay. I run two different computers, so that's what I'm doing right now. I'm getting it from one computer over to the other. Texted it to myself. Very not. There we go. And then here. And where is Chrome? There go there. So it's steal our winners IO webinar registration. I don't know if you can see now. You probably can't see that. I will put that here. That's where you can actually register for the next webinar that I do. And this is the. Let's see. What am I looking at here? I'm not looking at the right thing because I haven't hit enter. All right, so I do it on Thursday. So I think I'm doing one next Thursday. I'm not sure, but. And this can be found at. Well, I just put it in the chat. So there you go. And thanks for being here. Thanks for listening. Thanks for being interested in AI. I'm telling you that the future is going to belong to those of us who augment ourselves with AI. So don't try and replace, not try and have it do everything for us, but really be able to do everything that we do much, much better. And that's what it's doing for me, and I hope that's what it's doing for you. And I want everyone to have a wonderful weekend. I have clients coming in this weekend, so I'll be busy then heading to Orlando on Monday afternoon, I think, or Sunday, I'm not sure. And then I go to Mexico and hang out with Russell Brunson and a few other people at the end of the week, so I'm busy. Hope you're busy, too. And business has been booming, and I think it's mostly because of the AI thing. And I will see everyone on Monday, so have a wonderful weekend. To higher profits and beyond. This is Rich Efrin, over and out.