Latest / The Koerner Office - Business Ideas and Deep Dives with Chris Koerner / The Easiest Way to Profit From AI Right Now⏐Ep. #278
Transcript
- 0:00I'm here to report back. The state of corporate America
- 0:02is worse than you're assuming. It's worse than the echo chamber
- 0:05that Twitter is paints it to be. They know nothing.
- 0:08You get in there really quickly, you're going to see, holy crap,
- 0:11they don't even know how to Google something, let alone use
- 0:14AI. To most people who are sitting
- 0:16there that want to be entrepreneurs, that want to
- 0:18figure out how do I do something with AI, I'm going to say
- 0:20something that I think it's going to shock you.
- 0:21I think it's going to shock your audience.
- 0:22How do we make money here, Nick? Chris, I promise you we're going
- 0:25to get there. So you may have been wondering
- 0:30what happened to Nick, what happened to Hold Cobros?
- 0:32What happened to your best friend Chris?
- 0:34I'll tell you what happened. He took the last four months off
- 0:37podcasting of making content. He went underground and he has
- 0:40been deep in the AI world making agents, playing with Open Claw
- 0:44and Clawed. And we just reconnected for the
- 0:46first time in months so he could tell me everything he's learned
- 0:48and how we can make money off of everything he's learned.
- 0:51So if you're interested in AI agents, Open Claw Clawed all the
- 0:54latest and greatest and AI explained in a simple user
- 0:57friendly way and how you can monetize it.
- 0:59This is the episode for you. Welcome back Nick, I want you to
- 1:02prove this tweet wrong. That's the purpose of this
- 1:04episode, OK? Really, I'm in.
- 1:06Yeah, All right, I'm going to read this out loud to you.
- 1:08You ready? 100% I have 10 agents running
- 1:11while I sleep. No one is prepared for AGI in
- 1:13two years now. So what are you building, bro?
- 1:16All my smartest friends are vibe coding until 3:00 AM every
- 1:18night. It's all about agency.
- 1:19Intelligence is a commodity, man.
- 1:21So what are you building? Do you even study exponentials?
- 1:24Have you seen the latest METR chart?
- 1:26You're going to be stuck in permanent underclass bro.
- 1:29So what are you building? Did you even set up open claw?
- 1:32I'm maxing out my token budget every day man.
- 1:34So what are you building? I promise you I'm 10X more
- 1:37productive bro. You just don't understand.
- 1:39Please bro just I know you use this stuff every day too but you
- 1:41must not be prompting a right? Please bro 1.3 million views
- 1:46dude prove me wrong. Go to my profile, swipe right,
- 1:48swipe right. That's my right.
- 1:50That's that's my profile, that's mine.
- 1:54All right, Scroll down there, there.
- 1:56Click on that one, dude. So I can keep the lights on.
- 1:59And I just, like, read all night.
- 2:01It's insane, bro. So what are you even building,
- 2:04though, bro? You don't understand.
- 2:06My smartest friends are running copper wire through the walls
- 2:08and can ring a bell on the other side.
- 2:10Yeah, but what are you even building?
- 2:12Didn't you see they're starting with lights?
- 2:13Like, actual lights on a Christmas tree at the Edison
- 2:16office. They just freaking decorations,
- 2:17bro. But what are you building, dude?
- 2:20They're projecting pictures onto a wall and people are paying
- 2:22money to watch that. Like this Technology runs
- 2:25everything. Everything.
- 2:26That was my response to this freaking dude's.
- 2:29Tweet. I didn't know you responded to
- 2:30this. I'm completely surprised I never
- 2:32saw this tweet. This is amazing.
- 2:34OK, Yeah. Because it's like, it's like,
- 2:37OK, yeah, 99% of the things that people are messing around with
- 2:40don't have a use case and they won't have a use case, but how
- 2:42are people supposed to build the skill?
- 2:44Mohammed, edit out Nick reading that tweet because it makes me
- 2:47look bad. Sorry.
- 2:48Keep going. No, dude, but you, you know what
- 2:51I mean? Like the original tweet has a
- 2:53point. What are you building?
- 2:54What have you built, right, Because who's that guy?
- 2:56Alex Finn, I think, who's like every single day coming on
- 2:59Twitter, he's like, bro, this freaking changed my life, my
- 3:02life. Like he what is he really built?
- 3:04I don't even know he's actually built.
- 3:05But the flip side is this technology is changing so
- 3:08quickly. People are learning it so
- 3:10quickly. By the time you actually start
- 3:12seeing results, it's came over, bro.
- 3:14It's so soon in the cycle to be pulling out the what are you
- 3:18even building? What are you even done with
- 3:19that? We don't know.
- 3:20We don't know what it's going to look like yet.
- 3:21And the reason I pulled up the analogy of electricity and
- 3:24Thomas Edison is because there was a big lack, like it was a
- 3:27novelty for a long period of time.
- 3:28Like, what are you going to do with electricity?
- 3:30Oh, you can lock your house. You you're going to read books
- 3:34after hours. You, you know what I mean?
- 3:35Like people just didn't get it until all of a sudden people got
- 3:38used to the new technology and creative people were like, I
- 3:42think I could use this for, you know, whatever it is.
- 3:44And pretty soon we've got our modern technology.
- 3:46So I just think it's a, a pretty dumb way to approach it.
- 3:50What do you think? I'm convinced you sold me.
- 3:52I'm not easily influenced. Yes, yes, I'm an influencer.
- 3:57You're an influencer of influencers.
- 3:59Can I show you something? Real quick, before we pivot, I
- 4:02just want to add an analogy to your analogy of electricity.
- 4:05So last night I went to Chipotle with some kids at church that
- 4:08are starting a business. They bought this printer.
- 4:11It's not a 3D printer, it's not a normal printer.
- 4:13It's it's a printer that can print anything on anything.
- 4:16OK, so like if you want to like engrave your Tumblr and you
- 4:20know, and it has a rounded surface or make AT shirt or
- 4:22whatever or make like a 30 foot long banner, you can print it.
- 4:26And so they wanted business ideas from me and where we
- 4:29finished the conversation was like, dude, you need to put this
- 4:32in your backpack, take it to school, plug it in and print
- 4:34stuff on demand. Like you need to use this as
- 4:36your vehicle for rapid testing and prototyping because there's
- 4:41an endless amount of things you can print.
- 4:43You can print anything on anything.
- 4:44Like you could print the the wrestling team's weight classes
- 4:47on their shirts. Cool.
- 4:49You could print like they're on the track and field team.
- 4:51A lot of the track and field runners, they customize their
- 4:53spikes right on their shoes. You could print something on
- 4:56that, like their PR on the spike.
- 4:58Like we could sit here for 10 hours and just scratch the
- 5:00surface of all the ways you could make money with this
- 5:03thing. But don't try to go down those
- 5:04rabbit holes. Just use this as your rapid
- 5:07testing machine and just learn. Where's the demand?
- 5:10How much will people pay? What colors do they like?
- 5:12Do they want tumblers? Do they want their phone cases,
- 5:15engraved stickers, whatever? That's what you should do with
- 5:18this and leave your options open and then just chase the energy.
- 5:21And what we're talking about here is like, use Claude, use
- 5:24AI, use LLMS, use agents as your rapid testing vehicles.
- 5:28Like use it to have fun and to learn, but to don't, don't get
- 5:30married to anything just yet. Just be ready to pounce.
- 5:34And when something comes up, you're going to have all the
- 5:36skills necessary to execute on that thing.
- 5:38I'm gonna say something right now that I think is gonna shock
- 5:40you. I think is gonna shock your
- 5:41audience. To most people who are sitting
- 5:44there that want to be entrepreneurs, that want to
- 5:46figure out how do I do something with AI, I'm gonna say
- 5:47something. Go get a job. 100% go get a job.
- 5:53And I'm going to prove to you right now why I think it's, I
- 5:55think it's the best thing that you can do if you're earlier in
- 5:57your career, especially to work for a company.
- 5:59So first thing I'm going to show you, this is nuts.
- 6:02Chris is like, I'm going to hold it in.
- 6:04I'm I'm going to give him some rope.
- 6:05I'm going to give him some latitude here.
- 6:07I just love that you're speaking in like Stephen Bartlett clips.
- 6:10You're like, I'm going to say something that's shocking to
- 6:13everyone in the sound of my voice.
- 6:16Clip it. Clip it.
- 6:18Send it. If you're on Twitter, you've
- 6:19seen this. This is a graph or explain to
- 6:23our audio listeners. Yeah.
- 6:24So this is a chart, or a graph with 2500 dots.
- 6:27Each dot represents 3.2 million people.
- 6:29So effectively what we're doing is we're showing the entire
- 6:31population of the world. 8.1 is in there.
- 6:33Yeah, 8.1 billion people. Most of the dots are Gray.
- 6:36And then you've got kind of like maybe a tenth of the dots that
- 6:38are green, and then you've got some yellow dots and some red
- 6:40dots. So the Gray dots are people in
- 6:42the world who have never interacted with AI.
- 6:45They've never heard of AI, and that is like, let's just save
- 6:4880% of the world. Then you've got this green band,
- 6:50which is 10% of the world who have used a free chat bot.
- 6:55So they've gone and sat down with ChatGPT or they've used
- 6:57Gemini or whatever. And they made me ask it a
- 6:59question like, how do I cook a pot roast?
- 7:01If I ain't got no pot? I ain't got no roast, no.
- 7:04And then there's like the yellow part, which are now 123-4567.
- 7:10So 7 * 3.2. We're talking about 24 million
- 7:12people who pay $20.00 a month for AI.
- 7:16And then there's one little tiny dot dot here of people who
- 7:20actually use coding scaffolding. So what we're talking about here
- 7:23is Codex from Open AI. We're talking about code.
- 7:26We're talking about replit lovable cursor.
- 7:29Anyways, So when you look at this, you're like, holy crap,
- 7:32there's a lot of people who have no idea what's going on.
- 7:36And if you just lived in the Twitter bubble, which is
- 7:38basically the yellow people, you would think it's ubiquitous at
- 7:40this point, but it's not. We're very early in the use
- 7:43cases, and as we're early in the use cases, we're also early in
- 7:47the way that the technology is advancing.
- 7:49So I'm going to show this chart real quick.
- 7:51This is Gemini. They just released a model.
- 7:53And if you look at this, nobody's going to know what arc
- 7:55AGI is, but effectively what it is.
- 7:56It's measuring reasoning and knowledge.
- 7:58And you're like, holy crap, 84.6%.
- 8:01Oh, and there's Claude Opus 4.6, which just released.
- 8:04That's at 68.8%. Whoa.
- 8:06It's way higher than that. But there's not really any
- 8:08context for what this means, right?
- 8:11You don't really know. But when I look here, this was
- 8:14released in December. This was measuring the models
- 8:17that were out at that point, how many tasks they could do in
- 8:19human time and basically 1 output.
- 8:22And so in December we had. This is like an AI version of
- 8:26horsepower. It's like a way to measure it.
- 8:27Sure. Yeah.
- 8:29So GPT 5 in December in like 1 iteration.
- 8:32I asked it a question to to code something and then a code
- 8:34something. It could do about two hours of
- 8:37human work instead of 1 output of a prompt with 50% accuracy.
- 8:42Now forget that for a second, but that's just the measurement
- 8:44that we're using. Claude Opus 4.5 in December,
- 8:48this is December, could do about 5 hours.
- 8:50So like, dude, that's pretty good.
- 8:52And at that point it was doubling every four months.
- 8:55So you would expect December, April, OK, By April, we should
- 8:59be able to have these models doing 10 human hours in one
- 9:03output. Well, this is what actually
- 9:04happened because Cloud Opus 4.6 is released.
- 9:07This is where that 4.5 was on that graph.
- 9:10And now if you look way up here, 15 hours in one output.
- 9:15So if I ask it something now, I'm like, hey, code this, blah,
- 9:17blah, blah. It's output would do the same
- 9:20amount of work of 15 human hours in one output.
- 9:23So it didn't double so from 1:00. 5 to 15 in a month or two.
- 9:27A month and a half, yeah. So it tripled in a month and 1/2
- 9:30when it had been doubling every four months.
- 9:32What's the accuracy difference? It's the same.
- 9:35It's 50%. So 50% is the bar is the
- 9:37threshold, basically. OK, so it's not necessarily
- 9:39saying like, oh, great, it's going to replace everything, but
- 9:41that's just the best metric that we have to measure how well it's
- 9:44actually accomplishing human tasks in a specific period of
- 9:48time. This, when I saw this for the
- 9:49first time, was mind blowing because I'm like, OK, this isn't
- 9:52just a gradual, we're getting into the exponential territory.
- 9:55Who knows what the, what the coming models are looking at.
- 9:57So in December, you know, this, I had a house fire.
- 10:00That's why I'm sitting in this house right now.
- 10:01I haven't podcasted for a little while.
- 10:03And I did a lot of thinking like, all right, if I believe
- 10:07that AI is the future, there's no better use of my time than to
- 10:10invest in and leverage something in AI.
- 10:13But what is that? And I, I looked at starting
- 10:15other businesses, investing in businesses, and I wanted to have
- 10:18the biggest impact possible. And a friend of mine who works
- 10:20for a large public trading company reached out to me and he
- 10:23was like, hey, would you ever think about coming back?
- 10:25And for whatever reason, it was just like, yeah, I want to see
- 10:29if all this stuff that I've been working on actually works at a
- 10:32company. It's like you need.
- 10:34It's like you became a Formula One racer and you have no car to
- 10:37drive. That's exactly right.
- 10:38It's like I live in a retirement community and I need a golf
- 10:41cart, but someone gave me a Ferrari and I'm like, I'm sorry,
- 10:43guys. I, I got to go.
- 10:44I got to, I got to go drive this thing.
- 10:46I can't have this thing cooped up in my garage all day long.
- 10:49And so I was like, like, I didn't want to, you know, this,
- 10:52we talked about it. It's like, I don't want to get a
- 10:54job, but there was no other opportunity where I felt like I
- 10:57could have a bigger impact and see if these things actually
- 10:59work right. So I'm here to report back.
- 11:02The state of corporate America is worse than you're assuming.
- 11:05It's worse than the echo chamber that Twitter is paints it to be.
- 11:08They know nothing. They know nothing.
- 11:10I went to an off site with the executive leadership team.
- 11:14I had a conversation with them. All of a sudden a like 15 minute
- 11:17conversation ballooned into a 5 hour conversation because they
- 11:20were just asking me questions. How do you do that?
- 11:23You can do that right now. And by the end of that
- 11:25conversation that the CEO said to me, this is the first time I
- 11:29actually realized it's happening.
- 11:30Like I have an insane amount of urgency because I finally get
- 11:34it. I thought it was talk.
- 11:36I finally get it. And the stuff that I've been
- 11:38showing them is not anything mind blowing to people who are
- 11:41listening to this right now, right?
- 11:42But to people who like have real jobs, they actually have to like
- 11:45do something on a daily basis, like they're doing accounting or
- 11:47they're doing caregiving. They don't have time to go play
- 11:50around with this stuff. They don't know what it actually
- 11:51looks like. And so when you come and show
- 11:53them like, oh, that's interesting.
- 11:55We just had a conversation. I recorded it.
- 11:57Let me upload the transcript. It's going to spit out APDF of
- 11:59what we talked about and there'll be a presentation form.
- 12:01It's like I just came from the future and showed them The
- 12:04Jetsons. And like what I've done now is
- 12:07I've manipulated time. You're still living in a cave
- 12:10300 years ago. Nobody was a toy.
- 12:15But seriously, because for so long that it's been viewed as a
- 12:17toy that now they're like, oh crap, it's kind of here.
- 12:21Like the promise of it is kind of here.
- 12:23And so if you can go and work at a company, do for a year, who
- 12:27cares? And just get an idea of what are
- 12:30their pain points? What are they struggling with?
- 12:32These are billion, multibillion dollar corporations that have
- 12:35real problems. You get in there really quickly,
- 12:37you're going to see, holy crap, they don't.
- 12:39They don't even know how to like Google something, let alone use
- 12:42AI. In my opinion, there's this huge
- 12:44gap to bring the current state to the future by individuals who
- 12:49are humble enough to say, yeah, I'd rather start something.
- 12:52But I think if I went and actually worked at a company,
- 12:54I'll learn their pain points and then I can do whatever the freak
- 12:56I want whenever I want because people will think that I'm a
- 12:59magician. Cool, but how do we make money
- 13:01on this, Nick? Dude, we're going to get there.
- 13:04We're going to get there. July all.
- 13:05Right. OK, OK, OK.
- 13:06You are a magician in the context of what they know to be
- 13:09true, right? 100.
- 13:10Percent, I could go learn an easy magic trick on YouTube and
- 13:13show it to a three-year old and they'd think I'm amazing.
- 13:16A 30 year old would not. So in that kids eyes I'm a
- 13:19magician. You are a magician in their
- 13:21eyes, therefore you are a magician.
- 13:23One of the members of the team was like, when I started showing
- 13:26them all the stuff that I was using, they were like, I'm glad
- 13:30you should. I'm just going to say something.
- 13:31I think I speak for everybody. I was intimidated by you.
- 13:34Like I thought that you were a genius as genius.
- 13:36She's like, but I'm not saying that you're not.
- 13:38But now I get it. Like I get how you were able to
- 13:40do so much in such a short period of time, like capture
- 13:43information, synthesize it, create things.
- 13:46And I'm not working, you know, 80 hours a week.
- 13:49This is pretty simple stuff like I'm synthesizing data.
- 13:51So I have a couple of use cases and I like legitimately think
- 13:55this is the template that could make somebody.
- 13:57You're not going to make $1,000,000 tomorrow, but this
- 13:59will set you up, I think, for the rest of your life just to be
- 14:01a professional bringing people into the future.
- 14:04Why are you smirking? I'm drooling.
- 14:06I'm on the edge of my seat right now.
- 14:07You're still talking in Steven Bartlett clips and I'm just
- 14:10loving all of this. Chris, can I show you to answer
- 14:14that question? What have you built?
- 14:15Can I show you some of the stuff I've built?
- 14:17All right, here we go. I guess it's go time and I'll
- 14:21show you this. This was I just wanted to see
- 14:23competitors in this space and so I literally went, it's all
- 14:27publicly available. What space are we talking?
- 14:28About thank you the home health and Hospice senior living and I
- 14:32get curious. So I'm like, I wonder what
- 14:33everybody else is doing this publicly traded companies.
- 14:36They publicly report things. I want to go and see what
- 14:38they're doing. And so I just went and pulled
- 14:39their ten KS and 10 QS, which are quarterly or yearly earnings
- 14:42calls. And I scraped all their
- 14:44transcripts because I wanted to see like, so honestly, I went to
- 14:49Gary and I said, Gary, tell me how to do this.
- 14:52And Gary was like, oh, hello, this is my little British
- 14:54gentleman. So this is Gary.
- 14:56Gary's my claw. Bot to us what we're looking at.
- 14:58So Gary's my claw bot that I set up.
- 15:00I set it up a couple of weeks ago in the beginning of
- 15:02February. And I'll, I'll explain in more
- 15:04detail what the claw bot is because I have a whole thing on
- 15:06it, but effectively what it is. It's like the first window into
- 15:10mass adopted usage of agents. So you bring you bring Gary on
- 15:15and get well, not it's not Gary, it's open claw, but I call him
- 15:18Gary. I brought Gary on and I gave him
- 15:19access to my emails and my texts and my Google accounts and
- 15:24online and everything that I have.
- 15:27And now I just go through him and I ask him questions.
- 15:29And so the first thing we just started iterating was like, how
- 15:31could I get these? I was like, Oh let me go look.
- 15:34It goes and searches online. This is locally hosted, right?
- 15:36Yeah, I don't host it on the web at this point.
- 15:38That scares me. So this is locally hosted on my
- 15:40Mac. I will probably turn another one
- 15:44of my Macs just into this so that it's running all of the
- 15:46time. But right now it's locally
- 15:48hosted on the Mac that I use. There's a couple of different
- 15:50ways that you can do, Gary or not Gary, But Clobop, you can
- 15:53run a model locally on your computer and there are models
- 15:56that have been created that are open source for free.
- 15:59OK. But if you want to use a model
- 16:01like Gemini or Opus, OK, OK, GPT, I've got to pay for that
- 16:05inference, of course. Yeah.
- 16:07So every time I use it, I'm paying for that token usage.
- 16:09Sorry, could you build like a clog bot with DeepSeek and have
- 16:13it essentially be free because it's open source and then you're
- 16:16essentially you essentially have all the security features you
- 16:18need because it's locally hosted on your computer?
- 16:22Like how complicated is it for the average listener or watcher
- 16:25right now, someone ages 20 to 50, that's Internet, native
- 16:29Internet literate. How complicated is something
- 16:32like this for them to set up just as you've done it?
- 16:34It's complicated, but it's not impossible.
- 16:36It, like I would say the hard part is it does take a lot of
- 16:39time because it's not just downloading the software and
- 16:42putting it on your computer and running it.
- 16:43And I'll, I'll show this later. Like you, you've got to have a
- 16:46plan for how you're going to use it.
- 16:48That's what people miss with AI. Dude, I remember us talking
- 16:51about this a year and a half ago where we were like, if you can
- 16:54prepare and just think, what would I use an employee for?
- 16:57Now you're going to use AI for that.
- 16:58That's how you should use AI and, and people still understand
- 17:02that concept. So if you understand like, oh, I
- 17:04know exactly what I would use them for, here's the training
- 17:06material. Here's how I'm going to oversee
- 17:07them, then it becomes a lot easier.
- 17:10But if you don't, then this is like this messy iterative
- 17:12process. So, so downloading it and
- 17:14putting it on your computer, fairly simple, but like getting
- 17:17the mileage out of it, that's where I feel like people are
- 17:21falling short because they're buying a Ferrari when all they
- 17:23needed was a golf cart. And so like, of course, they're
- 17:26like, So what do I do now? Well, like, did you, you live in
- 17:29like South southern Florida's retirement communities?
- 17:31You can't go faster than 15 miles an hour.
- 17:32So buying the Ferrari was kind of a waste.
- 17:34The cool thing though, is like, so this is a, a model that came
- 17:37out, I don't know, a couple of weeks ago.
- 17:39I mean, the model itself hasn't come out, but it's the update
- 17:42came out a couple weeks ago and you can see how it's benchmarked
- 17:44against open AI Gemini and plot. All right, it's about as good.
- 17:50And the thing about this model is it's an open model.
- 17:53So clod Gemini, all those other those are closed models, the
- 17:57proprietary models, they sell you the ability to use those
- 17:59models. This model, this minimax is
- 18:02open. Anybody could go and download it
- 18:03and they could adjust the weights on it and they could
- 18:05kind of make it their own little LLM.
- 18:07So this you could download and run locally on your computer.
- 18:11And if that was the case, then you're not using any inference
- 18:13from Anthropic or Open AI or. Gemini, which means.
- 18:17So, so if I've got something that I want to do on my
- 18:20computer, if I want AI to go and do something on my behalf that
- 18:24cost money, and it cost money in the form of compute from one of
- 18:28these companies, because I'm using them, I'm effectively
- 18:30renting their model to do something, search the web,
- 18:34scrape databases, send an e-mail, campaign, whatever it
- 18:37is. But I'm I have to pay something
- 18:39in order to do that. For models like Gemini, you're
- 18:44paying like $5 per million tokens.
- 18:47For a model like Open AI's newest model, you're probably
- 18:50paying $10 per million tokens. For Quads Opus 4.6, you're
- 18:53paying $25 per million tokens. But it's so good.
- 18:56But it's so good. It's so good.
- 18:58I mean, that's why, I mean, that's why it's you get it
- 19:03anyways. So yes, theoretically you could
- 19:06download this, run this locally, and then you're not paying for
- 19:08any of that imprints and you're running Clodbot on a on a local
- 19:12server. But it's not going to be as
- 19:13good. It depends on your use case.
- 19:15You might not. 100 percent, 100% it may, it may not be as good
- 19:19depending on what you need it for and back to the analogy like
- 19:22you think you need a Ferrari, you probably just need a golf
- 19:24cart. Like you just just be real with
- 19:25yourself if. We're syncing calendar events or
- 19:28yeah. Dude.
- 19:28Responding to emails. You're you're not lighting the
- 19:32world on fire. You're, you're sending a cold
- 19:34e-mail campaign. OK, like let's let's be real
- 19:36with it. So I went and I said I want to
- 19:39know what the other publicly traded companies are doing.
- 19:41And so I asked Gary. I said Gary, how would I find
- 19:43this information? It was like this information is
- 19:45available on the Internet and it went and it found an API.
- 19:50So I went to Ninja API or API Ninjas, which I didn't know was
- 19:53a website. And I was like, oh, buy this
- 19:55$20.00 a month. You get access to all these
- 19:56other AP is I'm like cool. So I go there, I get all of the
- 20:0010 KS and then I'm like, well, I want to display it.
- 20:03And so it starts telling me how to display it and I end up
- 20:05building the site where. I can see when the next public
- 20:09reporting from these companies are, how they've done over the
- 20:11last quarter, where the revenue is, how they look.
- 20:14I can, I can go all the way and see just the themes, you know,
- 20:16how are things looking this quarter?
- 20:18I can see what questions analysts have been asking.
- 20:21And so from this, like I, I showed some of the people at my
- 20:23company like, hey, that's pretty cool.
- 20:24And they're like, actually, did you know how long it takes to
- 20:27prepare for quarterly earnings calls?
- 20:28I was like, no. And they start telling me about
- 20:31their quarterly earnings call. So it gave me this idea.
- 20:32I'm like, I wonder if I could help people prepare for 1/4
- 20:34learning calls because what you have to do is you have to take
- 20:36all. The data.
- 20:37How much could you charge for something like that?
- 20:38My broski, you're about to see. So these people, they have to
- 20:42report. It's an SEC filing requirement,
- 20:44right? Because they're publicly traded.
- 20:46They've got to do on a quarterly basis and they got all the stuff
- 20:47that they got to report, the financials, whatever.
- 20:49Their time is worth. Their time is worth how?
- 20:51Much an hour, so they get together and they spend like a
- 20:53couple days 1/4 every quarter. The whole executive team like
- 20:56talking about should we use great or should we use amazing?
- 21:00Should we use the word wonderful or should we use the word
- 21:03exceptional? You know, like they're just,
- 21:05they're debating these words back and forth and they're like,
- 21:08so my mind just starts going, I'm like, dude, all I have to do
- 21:11is look at your past transcripts.
- 21:12Easy peasy. I create a voice for you.
- 21:14I know exactly what the template's going to be.
- 21:16You dump the data in, I'll give you a first draft.
- 21:17I'll save you a day. And so I went and sure enough, I
- 21:21created this earnings pipeline. Stop scrambling before earnings
- 21:25start running a pipeline. And it's literally intake as
- 21:27your strategy. It's a workshop script
- 21:29refinement and export. And this walks you through
- 21:32exactly what you need to do if you're a publicly traded company
- 21:35to get the output that you want. Sorry, why do you not own
- 21:37earningspipeline.com? It's available.
- 21:39You need to take that. You do it, you're better at just
- 21:41don't pay that. I'll pay you.
- 21:44I honestly didn't know. I don't do anything unless Gary
- 21:47tells me. Now my wife's like, did your
- 21:49best friend Gary tell you that? I'm like, yeah, he's a really
- 21:53good guy. So like, this is the landing
- 21:55page for it. But you could go scrape all of
- 21:58the publicly reported 10 KS, 10 QS build profiles for each one
- 22:02of the executives, build a template for what exactly they
- 22:05talk about and when they talk about it.
- 22:06And then an ingestion pipeline with like I've done to be like,
- 22:09just dump all of your data in here and I'll populate this for
- 22:12you. And then you have a first draft
- 22:13because that's all this is. You populate it.
- 22:14Then you go through like the strategy session of AI making
- 22:17some recommendations, then you workshop it.
- 22:20Then you actually generate the first script, then you refine it
- 22:22and then you export it. And like, again, the thing that
- 22:25people miss here is they think AI just does everything.
- 22:28It doesn't. It's a good help.
- 22:30But like you still have to workshop things.
- 22:32You still have to massage the messaging.
- 22:33And so this is an actual representation of me working an
- 22:38earning call. Yeah.
- 22:39I spent 12 hours like this is to the point I could sell it,
- 22:42Chris. It's kind of like for me, you
- 22:44know, me, I don't finish stuff like this is finished.
- 22:47This is this. I like, I could take this to
- 22:48market. It's pretty, it's pretty
- 22:50bonkers. And so when I show this to
- 22:51people and they're like, holy crap, as an executive, like you
- 22:55just got the four highest paid people on the entire company 10
- 22:58hours a year back in their time. That's that's high leverage,
- 23:01that's high value, right. So anyways, this this was a fun
- 23:06one. Took me like 12 hours, but these
- 23:08are toys. I what I really want to show is
- 23:12what I think is the most amazing thing and that's scary.
- 23:14So open claw is really cool. It's got some protocols built
- 23:18into it. It's pretty plug and play in the
- 23:20way that it remembers things. People talk about let's go
- 23:23infinite the memory. It doesn't have infinite memory.
- 23:25It like takes time to learn things, but as it learns you, it
- 23:28adds things to different files so that every time it loads it
- 23:31remembers, oh, Chris doesn't like it when I spit out this
- 23:35output or Chris doesn't like it when I do XYZ.
- 23:37I was already in the process of like creating something that was
- 23:40my second brain and then Open Claw came out and I just adapted
- 23:42it to it. But we get bombarded with stuff,
- 23:45especially if you're an executive and this is just like
- 23:48whatever tons of messages per week.
- 23:50There's literally no way for you to process all of this
- 23:52information. The only way for you though to
- 23:55get out of AI what the promises is if you have clean data that
- 24:00it can extract, analyze and then spit out actionable information
- 24:04to you if the data is not. In, garbage out.
- 24:06Exactly. If the data is not clean, you
- 24:08don't get anything good. So a lot of these companies have
- 24:10data, but they're not, they're not cleaning yet.
- 24:12So I'm like, I'm looking at this.
- 24:13Give me an example of clean data in, clean data out or vice
- 24:16versa. OK, so here's dirty data.
- 24:20Dirty data, Boom, boom, you're a company.
- 24:23You have 10s of thousands of contracts, OK, but they're all
- 24:26saved in different folders and all of your DME contracts are
- 24:30saved in like what's this is the durable medical equipment
- 24:33contracts are saved in Some of them are saved in a DME contract
- 24:37folder. Some of them are saved in a
- 24:39durable medical equipment folder.
- 24:40Some of them are saved in a hospital equipment folder.
- 24:42Some of them are saved in a skilled nursing facility
- 24:44equipment folder. The data is just in a bunch of
- 24:46different places. There's no rhyme or reason to
- 24:48it. Second layer that could be the
- 24:49problem is you've got data laying around that is in PDF,
- 24:53it's in Word, it's in HTML, it's in a JPG.
- 24:56And so AI looks at that and they're like, I don't know how
- 24:59to extract all of those at the same time, or the naming
- 25:02conventions are off. Maybe you're like, I'm going to
- 25:05save all of the ABC code information in this folder.
- 25:09I'm going to call it ABC code, but you accidentally name it Abd
- 25:13one time or or you accidentally name it BCD one time, right?
- 25:16Like just common mistakes. So how does AI find all that
- 25:20stuff? It can't if the data is not
- 25:22clean. Now clean data would be clear
- 25:24taxonomy, which a taxonomy would just be a hierarchical way for
- 25:28you to to identify information and everybody stays with kind of
- 25:32the same saving file naming structure, right?
- 25:37So for me, what I did is I was like, I want to create Gary.
- 25:39I want to create my own AI bot that like knows me.
- 25:42So I went and I exported all of my ChatGPT data, all of my claw
- 25:46data, all of my texts because I can connect to my iMessage
- 25:51through an MC. Never.
- 25:52Deleted any text. Type never never through an MCP
- 25:55server. This is going to haunt me, isn't
- 25:57it? You're about.
- 25:57To come. Dude, dude, my boy, I love it
- 26:01right now. Right now I just pull up the
- 26:03most disgusting creep. No, there's nothing.
- 26:07Anyway, there's nothing in there.
- 26:08So it took all of that data and I was just like, I want you to
- 26:12take that and standardize it for me.
- 26:14And so it did, I mean, it took some time and some prompting,
- 26:16but it didn't standardize it for me.
- 26:18And then I was like, all right, we got to structure it.
- 26:20And so we started structuring it and putting it in a way where
- 26:23it's easily retrievable so that I can go and search and then I
- 26:26can create automations on top of it.
- 26:27And then eventually you can create agents.
- 26:29But if the data is dirty, you can't get good outputs.
- 26:32So that would be like the first thing I would say is just
- 26:35helping companies get clean data is a multi $1,000,000 business
- 26:38just going in there and being like I'll help you organize this
- 26:41and helping them clean. How can that be automated?
- 26:44Or is that just like the handheld messy process that it
- 26:47is? The cleanup, yeah, it can't be
- 26:50automated until you've built the system for it.
- 26:53You can get 80% of it done, but that last 20% takes a long time.
- 26:58So for example, if you've got a company that's got 10s of
- 27:00thousands of files, I could probably run AI on it and
- 27:04categorize everything to 80% confidence.
- 27:07But there's still let's say 20,000 files in there that you
- 27:10don't know where they go. How do you then figure out how
- 27:13to clean that stuff up? And that that's where somebody
- 27:16who has a lot of experience can come and be like, oh, that's
- 27:18simple, just do XY and Z. But even beyond that, putting in
- 27:23a format where it can query and use the data is something in and
- 27:27of itself. So you know this Gemini, Well,
- 27:30let me back up these models. These LLMS have what's called a
- 27:33context window. So think of it like this.
- 27:36If I wanted to buy a business, I'd go talk to Chris, right?
- 27:39Chris knows everything about businesses.
- 27:41He's like done every business, seen every business under the
- 27:43sun. But in order to get him up to
- 27:44speed, I've got to spend a couple of hours with him to tell
- 27:46him like, hey, look at this is how much the business costs.
- 27:48This is the market, this is the industry, this is the demand.
- 27:50This is how much money I have. And then by the end of that hour
- 27:52or so, Chris can give me an actual response.
- 27:55He can say like, oh, you should do XY and Z.
- 27:57So Chris, he's the model. He's clawed, he's Gemini, he's
- 28:01been trained on all this data, and he's just knowledgeable.
- 28:04That hour of me spending time to get him up to speed, that's the
- 28:07context window. That context window is pretty
- 28:09finite. And for a long time it only went
- 28:11up to like 200,000 tokens. It finally just hit a million
- 28:14tokens. But even with a million tokens,
- 28:16if we're talking about 10s of thousands of documents, we're
- 28:17talking about billions of tokens.
- 28:19It can, it literally cannot. It's not physically possible for
- 28:22it to query all of that data and return things, and so you've got
- 28:25to organize this data in a way where you parse it more.
- 28:28Efficient. You chunk it and you clean it
- 28:30and you tag it and you embed it, whatever.
- 28:31And then you set up these systems.
- 28:33So I've been doing some of this stuff and I was like, I want to
- 28:35build this for myself. So I exported all my ChatGPT
- 28:37conversations, all my clawed, all my emails, everything.
- 28:40And I implemented open claw. So open claw has access to
- 28:44everything. This is just like a, these are
- 28:46the seven files in open claw. So open Claw has a soul because
- 28:49you want, if you want to give it a personality, the user MD is
- 28:51just like what you want open clock to know about you.
- 28:54If you want any agents, run. This is where you would put
- 28:56instructions for agents. The memory, this is long term
- 28:59memory. These are things that you want
- 29:00open clock to sort of always remember.
- 29:02I work at XYZ company. I have XYZ skill set tools.
- 29:07You can go and research what agents and tools are the
- 29:10heartbeat. This is just like jobs that come
- 29:12every hour, 2 hours that continue to run.
- 29:15And then if you want to give it an identity, so those are like
- 29:17the seven basic files that it builds over time.
- 29:19And like the beauty of open claws, the more you use it, the
- 29:21more open claw builds these automatically.
- 29:24So what I built was like, I had already had this, I added this,
- 29:28I had a people framework, so I asked it and I like, I had a
- 29:31whole set of prompts to do this where it's like, how do I manage
- 29:33relationships? How do I manage failure?
- 29:36Like it knows if I showed you some of the stuff right now,
- 29:38you'd be like, that's pretty spot on.
- 29:40The response was just two words. You don't.
- 29:42You don't. Yeah, you don't.
- 29:45So when I'm talking to it now, like literally Gary will be like
- 29:49Nick, it kind of feels like you're spiraling here or like
- 29:51Nick kind of feels like you need to get back to this person.
- 29:53Oh, Nick, you probably see, I know you said you should take
- 29:55that on, but that's not a good idea because you take on too
- 29:57much stuff. It's incredibly helpful to have
- 30:00from from the get go. It also builds like a taxonomy
- 30:03for me. It also created a bunch of
- 30:05projects. So I like, I was going into this
- 30:08job and I was like, I just want to be organized and make sure
- 30:10that I'm not letting stuff fall through the cracks.
- 30:12And so it just created all the projects, all the people, it
- 30:15took all the conversations that I had and it synthesized them
- 30:18into these documents. And so when I started, I had
- 30:21open claw that was already working.
- 30:22And then I layered Gary on top of it.
- 30:24And now literally I can say, Hey, what do I have outstanding
- 30:27to so? And So what did I say I was
- 30:28going to do? It will remind me because I've
- 30:30got jobs set up for it to come and remind me and say like, Hey,
- 30:33remember, at the end of the day, you're supposed to get XYZ thing
- 30:36to so. And so it will preemptively give
- 30:39me a spreadsheet based on what I said I was going to do is like,
- 30:42Hey, does this look good? Like just give me a first draft
- 30:45of this stuff. But it's because I went through
- 30:47and spent the time so that I had the context to understand my
- 30:50people framework commitments, the decision framework, my
- 30:54personal context, the taxonomy, the extraction methodology, all
- 30:58that stuff. And so you can see like this is
- 31:00all the crap that obviously it's spit out anyways.
- 31:04And the way that it's built is like literally queryable.
- 31:07I can query just about anything that I want to know.
- 31:10So if I'm like, hey, what did Chris and I talk about the last
- 31:12time? If I'm getting ready for a
- 31:13meeting, if I'm getting ready to give a presentation, Like I
- 31:16can't tell you how many times in the last few weeks people have
- 31:19asked me to do something and I'll like come prepare to a
- 31:21meeting with a presentation. And people are like, what?
- 31:24You did what now? Yeah, right.
- 31:26Because like the old paradigm is this took you 5 hours.
- 31:29The new paradigm is it took me like 5 minutes.
- 31:32All the setup took me a long time.
- 31:33But now I'm just able, I'm able to access it.
- 31:36So anyways, I'm gonna I'm gonna stop talking because I feel like
- 31:39I'm just on a heater. OK, so with everything that you
- 31:41built for yourself, how much of your contacts window did you use
- 31:45up it? Depends on how it's being used
- 31:48and when I'm utilizing it. So if I'm asking it specific
- 31:52questions about people, it will go then and look at the people
- 31:55file and pull it. So it's not loading in the
- 31:58context every single time, but it is loading into the context
- 32:01when I'm asking about. Specific.
- 32:03Is that what the rag is? So no.
- 32:07Explain what RAG is. All right, so remember how we
- 32:10were talking about how you've got all this data?
- 32:13So if I've got 10s of millions of tokens of data, but I can
- 32:17only ever ingest 100,000 tokens, how do you make things
- 32:20queryable? Well, there's this RAG approach,
- 32:22which is retrieval augmented generation.
- 32:25And So what you do is you tag all of that data with metadata.
- 32:30So for example, think of it like a library if there's a book
- 32:33that's written on ancient Rome. Like a Dewey Decimal system.
- 32:36A Dewey Decimal system. Yeah, it's going to be like it's
- 32:38in row A, column B, categorized with the rest of these things,
- 32:42right? So if you search a word, it's
- 32:45going to pull up where that might be located and then allow
- 32:48you to access that stuff. The Gary and open clause a
- 32:52little bit different. It's it's on this thing called
- 32:54QMD, which is quick markdown. It's not a vector data.
- 32:58This is like way too technical. The gist is it allows for
- 33:02semantic searches and the results are much more accurate.
- 33:05So everything that that I would have are semantic searches
- 33:09because they're meetings, they're being transcribed,
- 33:12right? If I had a big database with
- 33:14numbers like the, you know, maybe I'm.
- 33:16Maybe I'm using more of a computer?
- 33:17You're using it like ChatGPT, not like a coder would use it to
- 33:20search. It exactly exactly.
- 33:22A code database. The reason though, that it's
- 33:25important is because now it unlocks all of that data that
- 33:28I've had sitting there, all of that context.
- 33:31I don't get to the middle of a conversation with Gary and all
- 33:35of a sudden he's like, you can't use me anymore.
- 33:37I've run out of memory because it's constantly updating itself.
- 33:41I don't get stuck in the middle of a conversation with them.
- 33:43Like if the memory is persistent, it's very helpful.
- 33:47It's it's fantastic. So if I were to say the lowest
- 33:50hanging fruit though that I've seen, it's so dumb because I can
- 33:53show all of this stuff and it's like agents and skills and OMD
- 33:57files. Whatever.
- 33:58I'll tell you right now. Here's the 8020.
- 33:59If you want to unlock the most value, record your meetings,
- 34:02period. Record your meetings, transcribe
- 34:04them. Have a vehicle or a way for you
- 34:07to actually get a summary and a synopsis of that and then build
- 34:10in yourself some type of an accountability mechanism for you
- 34:12to then say, hey, this was a do out that you committed to.
- 34:15I will make you millions of dollars.
- 34:17I've seen it now because right now the traditional way within
- 34:20companies is like, what are they doing?
- 34:23They're writing something down or they're like they're trying
- 34:25to remember it, or maybe they use Copilot, which sucks.
- 34:28There is no way for them to capture what was done in that
- 34:31meeting. Save it to some type of a
- 34:34archival system that you can then access and query later and
- 34:37then follow up with individuals like that's, that's always been
- 34:40the hardest part, right? Follow up, follow up and follow
- 34:42through. I said I was going to do one
- 34:43thing and I didn't do it. Why didn't I do it?
- 34:45Well, maybe you forgot, Maybe something slipped through the
- 34:47cracks. But if you just record meetings,
- 34:49document what was said, and then put it in a place where you can
- 34:52go and get back to you later or build something that reminds
- 34:54you, you're ahead of 95% of people because they're not using
- 34:57it for that right now. People get so tripped up on
- 35:00like, I'm going to build this agent or I'm going to build this
- 35:02skill. No.
- 35:03Literally record a meeting, summarize it, put in a
- 35:07transcript, put it somewhere that you're going to check in,
- 35:09and then all of a sudden you've got this superpower because
- 35:12you've got this massive database that you can go back to.
- 35:15How can people make money learning how to do this and then
- 35:18doing it for individuals or for companies?
- 35:20Is that a viable opportunity right now?
- 35:22Like a picture If if I'm an executive watching this video
- 35:25right now, I'm like trying to find your contact info, right?
- 35:29Because yeah, because I'm like seeing this and I'm overwhelmed.
- 35:33And it's like, and this isn't a sales pitch like Nick has
- 35:35nothing to sell us, but it's like, I feel like people.
- 35:37Did. People could learn how to do
- 35:40what you've done and and charge for it.
- 35:43So the first thing I would say is you and I are so freaking
- 35:46lucky. Like we're so lucky that over
- 35:50the last two years we've just been able to like dabble, you
- 35:54know, like how does that work? Oh, that's interesting.
- 35:57And we just Start learning about it.
- 35:58So just devoting the time to this, you're ahead of 95% of
- 36:02people because they don't, they don't have the time and they
- 36:04don't want to make the time. And by the time they get home
- 36:05from work, from doing all the things that they're supposed to
- 36:07be doing, they don't have the time to like ingest this
- 36:09information and then figure out a way that makes it applicable.
- 36:12So the first thing I would say is just learn.
- 36:13Just learn, bro. The second thing that I would
- 36:16say is anytime that you've been within an organization where
- 36:19they're like, I wish there was a better way to do this, that's an
- 36:21opportunity. If someone's using a
- 36:23spreadsheet, that's an opportunity.
- 36:24If you're on a meeting that could have been an e-mail,
- 36:26that's an opportunity. But like.
- 36:28How do we make money here, Nick? Chris, I promise you we're going
- 36:31to get there. OK?
- 36:32Give me a minute to finish this thought.
- 36:33Yes, Sir. If you're on a meeting that
- 36:36could have been an e-mail, that's an opportunity.
- 36:38If somebody let something slip through the cracks, that's an
- 36:40opportunity. I think that now the cost of
- 36:42building custom code, I didn't even show you all the other
- 36:45stuff I have like little survey software or tracking things.
- 36:51The cost of custom code is so low that you can build
- 36:53customized tools that save people 80% of their time.
- 36:57And it doesn't have to be like it on the mass corporate scale.
- 36:59It can be on the small scale. So anyways, first one would be
- 37:01learn. The second thing is I think
- 37:04there's huge demand for a corporations just to be in the
- 37:08know. If you get educated and you just
- 37:10cold call, literally, you could set up a claw bot to be like
- 37:13here, all the public security companies, because I can go and
- 37:15scrape all of that data, hold e-mail, go and find the
- 37:18executive information because all of that information is also
- 37:20public hold e-mail every single one of those executives and say,
- 37:23hi, I'm Nick. I've been deep in the AI in the
- 37:27last year and a half. I know what's coming around the
- 37:29corner and 95% of your competitors don't.
- 37:31I'd love to have 5 minutes with you so that I can update you on
- 37:34what's coming down the Pike. You're probably going to get
- 37:35rejected by most people and you can probably perfect that sales
- 37:37pitch much. But like you, you would do
- 37:40amazing what this called outreach.
- 37:41But if you can get on their calendar and just have like what
- 37:44I had with that executive team, like within a couple of minutes,
- 37:47they're going to be like, oh, I get it, I get it.
- 37:50I want this guy every single week just giving me an update.
- 37:53They've asked me to like, hey, would you just, would you do a
- 37:56course for the next 12 weeks, one hour a week for the
- 37:59executive team. They want to know.
- 38:00They just, they don't know how to use it.
- 38:03They're kept from the truth because they know not where to
- 38:04find it. Chris.
- 38:07So, oh, I can't wait for like the few people to be like, oh,
- 38:11brother elder, but just putting yourself in a position where you
- 38:15can relay yourself as a subject matter expert.
- 38:18And again, this is like 2010 on social media where it's like,
- 38:22oh, you have a Facebook account. Will you run social for us?
- 38:24That's, that's what it's like in AI right now.
- 38:26Oh, you kind of use Claude. Can you run AI for us Once a
- 38:29week? You could come in and and pay
- 38:31consulting services. Like do you remember when you
- 38:34went and met with that unnamed billionaire and he was just
- 38:38asking you questions, like just extracting information from you?
- 38:42I think just doing that session alone, you could charge a couple
- 38:45$1000 just to give these executives a taste of kind of
- 38:49what's coming around the corner because they, they don't know,
- 38:51they don't have the time to do it.
- 38:52And that that's kind of where I was like, of course you don't
- 38:55know. Of course all of your day is
- 38:57spent managing people. The second you meet somebody
- 39:00like me, you're like, OK, I get it.
- 39:02Holy crap, train is coming. I'm about to get hit.
- 39:04So I think an executive boot camp, I think weekly
- 39:09roundtables. I think a fractional AI officer
- 39:13is 100% in the offering. It is more of a newsletter, but
- 39:17some type of a briefing service. You don't even have to be an
- 39:20expert in in vibe coding, just like, hey, I'm going to keep you
- 39:23up to date. I do think custom vibe coded
- 39:25tools are massive. You and I have talked about that
- 39:27for an AI agency for an, you know, a very long period of
- 39:31time. Probably the biggest unlock
- 39:32though, is if you can figure out how to get proprietary data sets
- 39:36within an organization accessible to that organization.
- 39:40Massive unlock because right now they have no way to do it.
- 39:43They're like, what can you get? Power BI and I can do a sequel
- 39:45database like that. That's hard.
- 39:47Somebody has to have a skill set to do that.
- 39:50If you can get an AI UI on top of that data so that people can
- 39:54just search, hey, show me where we have the largest deficiency
- 39:59in labor costs in the company right now.
- 40:02If that would return an answer that blows people's minds, and
- 40:05you can do that right now. It's not like you have to build
- 40:08a, you know, the sequel language in place.
- 40:10You can, you can AIUI on top of this data.
- 40:13So if you can unlock the data, it becomes really valuable.
- 40:15I was talking to somebody about this.
- 40:16It's almost like fracking. Remember how fracking was this
- 40:19new way to extract oil out of the ground?
- 40:21So you're no longer just going deep, but you're going like
- 40:23spreading out. To me, that's what AI is.
- 40:25It's like you, you're fracking, you're leveraging that data that
- 40:28was unaccessible before and doing it in a way that is much
- 40:31cheaper and much more accessible than it's ever been.
- 40:34So I think that piece in just coming into an organization, you
- 40:38could just do it with one. Like are you a healthcare expert
- 40:41like I am cool. Hey, I will show you how to get
- 40:45every single one of your quality reports for every one of your
- 40:48locations in the next six weeks. I wouldn't say like in a
- 40:51weekend, you know, give a reasonable time period, but then
- 40:54you have an opportunity to actually learn and implement it.
- 40:56Does that make sense? Yes.
- 40:58I can't tell if you're quiet because it sucks or if you're
- 41:00quiet because like you're thinking about or.
- 41:02No, I already told you it's a banger.
- 41:03Nick, what do you want? I had a headache.
- 41:06This is good. This is really good.
- 41:08I'm just thinking all these things.
- 41:10I'm like, what will the audience think?
- 41:11What should I do right after this?
- 41:13Like, how quickly can I implement this on my computer?
- 41:16I'm just thinking my mind is just going nuts.
- 41:18It's like for you cold, like I was talking to somebody about
- 41:22this because I was at this executive off site and one of
- 41:24the kids I was, I had mentioned you and they're like, is that
- 41:27guy on TikTok? I was like, yeah, I was like, is
- 41:29that the Kerner office? Anyways, I think for somebody
- 41:32like you, if you have open cloth, you could be sending cold
- 41:35e-mail campaigns 24/7 because there's this window of time
- 41:39right now where people aren't sick of too much AI.
- 41:42They're getting there. But like pretty soon,
- 41:45everybody's going to catch up, everybody's going to be doing
- 41:47the same cold e-mail outreach, and all of a sudden that Channel
- 41:49is going to be flooded and you have no longer have arbitrage in
- 41:51that Channel. Right now, you have arbitrage in
- 41:53those channels. If you set up a clawbot, you're
- 41:56very clear with who your customer is and you know the
- 41:58distribution channels and then hit it, there's arbitrage.
- 42:01You are going to find people in the next 6 months when once
- 42:04everybody kind of figures that out, those channels are flooded
- 42:06and there's going to be a new opportunity.
- 42:07I don't know what that is, but like right now there is leverage
- 42:10if you know something. So in my mind it's like, what is
- 42:13your secret sauce? And now with Clawbot I can
- 42:17unlock it because I could hire like two or three people to be
- 42:19my minions. Go all in on it.
- 42:21Go learn, just go play. You will figure something out
- 42:24and it will be incredibly valuable.
- 42:26When you're talking to Gary, what model do you normally use?
- 42:29Opus 4.6. So good.
- 42:32It's really good. Is it time to leave chat GB
- 42:34behind? Dude, yeah, so I like I use
- 42:38ChatGPT for what I would call like the Honda Accord things.
- 42:43You run into the grocery store. It's amazing at what it does and
- 42:47it's reliable and I know what I'm going to get every single
- 42:49time. And so if I have large data sets
- 42:50I need to extract stuff from, I'm going chat TPT.
- 42:54But Frontier models like Claude, it's pretty incredible.
- 42:58Gemini's new 3.1 model, it's pretty incredible.
- 43:02Like it's weird to say because it doesn't feel like it was that
- 43:05long ago that there's like weirdness with some of these
- 43:07models, but Claude feels like I have an expert in every topic
- 43:11known to man at my fingertips all the time.
- 43:13It costs a lot, but I, yeah, I use Claude all the time.
- 43:16Here's my stack though. I'll tell you my stack.
- 43:18So I use Claude to plan 4.6. And when I'm building software
- 43:24now, I'm like, OK, this is my idea.
- 43:28Help me write sort of my PRD, my product, product requirement
- 43:33document. There we go.
- 43:35So it writes my PRD of like what I'm hoping to accomplish, but
- 43:38then I will send it out to like Gemini and Codex to say, I tell
- 43:41it, go do an adversarial audit, have them tell us what we're
- 43:44missing. And I go through like 4 rounds
- 43:46of that. And then after those 4 rounds,
- 43:48I've got something that's pretty good.
- 43:50And I start now the planning phase.
- 43:51All right, let's plan something. Go do another adversarial audit
- 43:54on the plan implementation, not just like the build spec, but
- 43:57the plan implementation. And it goes, you know, it goes
- 43:59through all the steps. So I'll use other models as a
- 44:02way to sort of glean other insights that I might have
- 44:05missed. But once I have all of the data
- 44:08and I just need good analysis, Opus, I mean, I just that Opus
- 44:11is the one that's like incredible when it comes to
- 44:13analysis. So what I'm about to show you is
- 44:15like super, super low tech compared to what you're doing.
- 44:18But I, I posted this interview with this woman as a a snail
- 44:23mail subscription business and the interview was doing really
- 44:26well. So I thought, how could I
- 44:27monetize this video even further?
- 44:29How can I like create a super hyper detailed business plan and
- 44:33then just make a Stripe link that goes to a Google doc of the
- 44:36business plan and I can sell it for five to 10 bucks?
- 44:39And so I'm going to show you how bad this prompt was.
- 44:41This is my prompt. I pasted the YouTube transcript
- 44:45of my own video. This is a transcript of an
- 44:47interview. I want to use this data plus
- 44:49other data from online to make a very comprehensive business plan
- 44:52for someone that wants to start a snail mail subscription
- 44:54business similar to hers. Find other people doing a
- 44:57similar business and their reported number of subscribers,
- 44:59revenue or profit. Include charts or graphs as
- 45:02appropriate. Make it very thorough and
- 45:04tactical. How can people get stamps,
- 45:06stationary, custom envelopes printed, etcetera.
- 45:08Marketing very comprehensive. Take your time.
- 45:12I'll be selling your output for $5 in the pinned comment.
- 45:15For people that are serious about starting a business in
- 45:17this space, why do you want them to know what I'm doing with?
- 45:20I don't know, it's just all right.
- 45:22No. So it's just all right context,
- 45:26right? And so here's what it gave me.
- 45:30This was one prompt. I did not edit the prompt at
- 45:33all. So it's got charts, it's got
- 45:36crossed. It's 30, it's 30 pages and it's
- 45:396500 words. It took like 10 minutes.
- 45:42OK, OK, cool, right? Pretty good budget.
- 45:45Let me show you what's what's really cool.
- 45:47Here's I went in the stripe and I said it's loading.
- 45:51It's loading. I went to the description of the
- 45:53YouTube video, clicked edit and I said hey, if you want to learn
- 45:56more, you know, I made a 6500 word business plan and then I
- 46:00went to the pin comment, copy and pasted the exact same
- 46:03message in three clicks. I made a Stripe payment link.
- 46:05See all those $9 transactions every 10?
- 46:0915 minutes. Shut the freak up.
- 46:11OK. Next, are you?
- 46:13Kidding me? Yesterday it was $5
- 46:15transactions. You raised it OK, and then I
- 46:17thought oh let's AB test it. So I made this simple Google.
- 46:21Sheet you crazy man. I made a simple Google sheet
- 46:25.24% of people that view the video paid $5.
- 46:29Point two, 3% of people who watch the video paid $9.
- 46:32Same conversion rate for almost twice the price.
- 46:35Oh my gosh. So what that did was my revenue
- 46:39per 1000 views went from 1355 to 3528.
- 46:44It almost tripled with one prompt in Opus 4.6.
- 46:49Almost tripled Nick. I'm going to up level this.
- 46:53I'm going to up. Should we zoom out?
- 46:54Should we circle back? Drill down on this?
- 46:55Yeah, yeah, like. If you had an open claw, you
- 46:58could you could now say, hey, I know I want you.
- 47:02I want you to have an automation where every time I have a video
- 47:04that gets above this many views, you go and create a business
- 47:07plan, create the stripe link, create like it can create that
- 47:11people sheeting right. It's like I want to know the
- 47:13analysis and then every day message me because I want to see
- 47:17what the update is, which blows my mind.
- 47:20And I know that sounds so simple because you were like it was so
- 47:22easy, dude, I could just freaking create the sheet.
- 47:24I could just freaking do XY and Z.
- 47:25Yeah, you could you add up all of those over I cost you.
- 47:29I sound like Ed Mylet. You add that up over a day, I'm
- 47:33going to kick your butt. You add that up over a week.
- 47:35I'm way ahead of you, bro. You're just mad.
- 47:38I'm living in the future and you're living in The Cave trying
- 47:40to do it 300 years ago. No but seriously, like it's all
- 47:43of those little time saving things that allow you then to go
- 47:46focus on the highest leverage use of your time, which is
- 47:48creating content, being creative.
- 47:50It's not the administrative stuff that's the unlock of using
- 47:53something like Open Claw. That's the same thought I had
- 47:55today was I could use open claw for this.
- 47:57You would be the belle of the ball with open claw, my
- 48:01frightened. Dude, I'm at Max capacity in my
- 48:04brain right now. You got to.
- 48:05Call it all right, love you too. Where can people find you, Nick?
- 48:11Twitter go founder's Nick. I have a YouTube channel called
- 48:14nickonomics. I'm firing it back up I'm I'm
- 48:16back in the game. This was nice.
- 48:18Thank you. It's.
- 48:19The most resigned thing I've ever heard you say?
- 48:21Twitter. You knew that like it was the
- 48:23lowest and worst value of your call to action.
- 48:26You're like Twitter, because what else is there nowadays?
- 48:29You can find me at Nick Consulting $5000 an hour.
- 48:35I probably should actually. I mean, seriously, what you
- 48:38think? Please share it with a friend
- 48:39and we'll see you next time on the Kerner office.