2: Using AI for Sales Copywriting Strategies with Jon Benson

Okay, great. Today on Using AI at Work, we're thrilled to have John Benson, an absolute OG in the copywriting world. John, a best selling author and one of the foremost sales copy strategists globally revolutionized Internet marketing with his video sales letter formula which has led to billions in sales. He's a pioneer in AI copywriting, having introduced innovative use of tools like Jasper AI Chat, GPT and now his own Benson AI which is redefining what high quality AI generated sales copy looks like. Prepare for a deep dive into AI powered marketing strategies with the legendary John Benson. Welcome to Using AI at Work. I'm your host, Chris Daigle. Each week we'll be learning how today's business owners, entrepreneurs and ambitious professionals are getting more done with smart use of tomorrow's tech.

Let's get started. John, I'm excited to have you on. There's no question that you're an incredible fit for the goal of this show which is to educate users on how to introduce AI in their business process. And what's cool about you being on this is that you're not just a user, but you're actually a seller of AI services with Benson AI, and that's spelled B N S N. It's very clever by the way. And we'll have a link in the show notes for those who aren't familiar with the the software. Can you tell us about it?

Yeah. So Binson is promptless AI, which is the big delivery point for people who are afraid of AI. This requires no prompts from you in the traditional sense at all. It is literally copy, paste or type into a box, click the magic. We have something called the magic button. It's called the magic button for a reason. You press it and magic stuff happens. And the other thing that makes us completely different is we build on what's called avatar blueprint. So we, Benson has been trained on thousands of these models. We have our own model that we've trained on and that's separates it immediately from like ChatGPT and other chat oriented kind of copy apps because they're pulling from AI's vast library of copy training that it's done by reading the Internet. But as I've said in hundreds of interviews, if I wanted to train somebody to be a good copywriter, if I would never tell them to read the entire Internet because 99% of it sucks. So that's why 99% of chat copy sucks. And we've trained it on, we narrow the focus so we tap into the power of that really cool tool. And then we narrow it with all the trainings that we've done from, you know, hundreds and hundreds of copywriters.

So there's a ton of products that have come out since January or actually November of last year. But you've been playing with the models for a bit longer than that. How did you get turned onto it? How did you get access and how were you, what was your initial foray like?

Well, the initial foray was in good old fashioned. Hi, it was we, we in accelerator we did in 2010 and that was how do I break out my, my VSL formula and put it into software? And that was literally people. It gives you different copy every single time you generate a vsl but you have to fill in all these blanks. And it was, I mean it was very, very manual at the time. It was like people were freaking out about it. Right? But now it's looking back, it's just, it's kind of ridiculous. Copy Pro came along in 2016 and it wasn't HR, it wasn't AI either, but it was nascent. We were kind of getting ready for it. So I was like, I started building Copy Pro by myself in my bedroom literally and said okay. And during that time of building it I was like, I was able to take about six months and dissect what everything from emails to sales letters to VSLs would look like. We started with email. In fact it used to be called Email Copy Pro because email was the most versatile, the hardest thing to do. I mean, how do you systematize email? Right? And again, this is before AI was even a gleam in its daddy's code. So then we took that and about two years ago, or it's less than two years ago, Jasper reached out to me and said, hey, would you like to train our AI on your bsl? And I was like, started laughing. There's no way I can do it. They can't write it. It's just, it's, it's, it's, you know, I'm happy to, I'm flattered you would ask. I said, well, let's just try to work together. So I got on with their guys and ended up working with some of the top guys at Jasper and they figured out real quick, oh yeah, you're right, we'll do this. And so, so we did a 60 second version of it and it was, it was okay. It was, it did pretty well. But we trained it with the best models that we could. But in so doing, in that I got into the mod, the actual tr, what it looked like to train a model, what it looked like to how we could game the system, so to say. And in a very ethical way, of course, how get AI to do what we want it to do versus what AI will do by itself, which is usually just ramble on. So that's when I came up with something and I've recently trademarked it, called pre prompting. And it was combining a bunch of elements together, some elements. We just literally just slapped it into the playground and it worked and was like, okay, I don't know how this is working, but it's working and turns to find out. It was kind of cool. I was the first person that I know of to do it. Maybe I wasn't, but that I know of, at least in copywriting in the marketing world. So then we took that and then from that foundation is what we led to building Benson. We started building Benson for VSLs and then it expanded really quickly. Now it's now can do anything, but. So that's kind of the, the foray into that. So, yeah, I've been doing this a lot longer than chat. GPD has been a thing. And we knew this was going to happen. We knew five years ago, we knew that AI was going to happen. Four years ago we were trying to beg AI developers to help us do this. And they were. And I was literally told, well, this is just not possible. It's the hardest thing for AI to do. I'm like, that just doesn't sound right, you know, and. And they're saying, yeah, I can go out and it can land a robot on Mars, but it can't write a primary goal. That doesn't make sense, you know, and. But that's what we were being told by every AI engineer we were talking to, like, oh my gosh, this is so difficult. And then that was before, you know, GPT came along and kind of broke the mold with three, five. And then when four came out, that was like, okay, it's smart enough to do this now. But we were way ahead and ready with what we were working on with Benson, because we were doing this on the side for now, bsl. And then we just took it and said, oh, hey, we were going to release Benson. It's only a BSL tool because it's got my last name on it. And it's like, I'm known for the bsl. But then we trained it on the vsl, thought, wait, it can do everything else. Why are we just doing just VSL? So then my, my friend Peter came out with VSL AI and it's trained on like shorter form VSLs. And so I said okay, this is perfect timing. His came out first. When we came out we were like well everything. And now it's a model that it's a system of approaching AI that I think is going to not only stick around for the years that AI is going to grow, but also help people without having to mess with prompts. And you know, it's clean, it's simple.

This is awesome. So I've got some questions specific about how you're using it in your day to day, but I want to dig in a little bit more the people who are using Benson. And I come from a direct response background. I know that copywriters are probably both enjoying and scared of tools like these. However, what about, do you have a customer base or an avatar that's not your typical expected user? That's not a copywriter, not a marketer.

Definitely our largest avatar right now in Benson is actually just small business owners. People that know they need to sell something but have no idea how to do it. People that are authors, they want to be, they want to sell a book or a video course and they have no idea how to do it. So copywriters were the low hanging fruit for sure and still are. And yeah, I mean it's like I wrote for our Black Friday promo that just happened. First email that went out in the subject line, it was called the Oppenheimer AI and, and it's like I was like, like I usually don't use lofty stuff in copy, but the movie's kind of hot or this head was kind of hot. So people, people now know the name. And so I said, why do I call it that? It's because, you know, I, I literally I, I just, I just did something that took me, no kidding, like less than a minute and it spit out the lead of a sales letter that was as good as I could write it. And I thought to myself, I am become Death the Destroyer worlds. The, you know, is the quote attributed to Oppenheimer that was actually from the Bhagavada. But yeah, he was quoting the Bhagavada. And that's what I thought. Oh my God, I think I've just killed copywriting. It's like, this is insane. And I thought, wait, no, I haven't killed copyright because I still need to know. A copywriter has to know that that's good and then say okay, it's good except for this part and let's change it and then so copywriters are going to have a lot of work done for them because of this, but they still need to understand what makes something good. I mean what makes hook, what makes a unique mechanism, what makes this compelling. And, and then adjust it for the client. So yeah, it's, it's not, it's. It is, it is absolutely the atomic bomb. When it comes down to C level copywriter. It's, it's game. There's no way. You can't, it's not, you're not going to compete against this. It's. It's about putting a list copy. It is, I'm not lying. It is. But again, I really still, I still teach copywriting. So I'm like. Because you have to know you can't just rely on the AI. You just, you have to be able to say because it's not always going to do that. It's AI is going to AI. As I say, it has some. It's a petulant tween. If you think of AI as a petulant tween, then it really helps because sometimes it goes I don't want it. You know, and something else. And you have to know when it does that. Okay, that, that sucks. Click the click the redo and, and but yeah, the tool is, it's just, it's, it's one of many that I think people should be using if they're interested in copy. It's that simple. It's, it's, it's not the only thing. It's probably the best copy. But I mean I use chat GPT all the time for research. I use Claude for research. Claude seems to be more accurate.

But you know, I want to dig in on how a small business owner who. Because the E myth a lot of hats. I get it. And marketing. If you're a marketer, it's very easy to recognize that not many businesses do a very good job at it. If I was a small business owner and I marketing wasn't my wheelhouse, what's the spectrum of things that I could use Benson for to. And honestly, I mean if you would compare it to the amount of money saved from hiring a talented writer or email writer or whatever. How would a small business owner use this? We'll be back after a quick break.

Do you want to start your own unsexy business offering a local service or a trade to your local community? Come check out the owner ops podcast. We bring on people like Sam who has built a 22 billion dollar roofing company in five years. I've also Brought on people like Brock, who generated $1.3 million of revenue in his first seven months installing septic systems. If you are looking to build your own local service or trade business, come check out the Owner Ops podcast. That's ownerops.com o w n r o p s.com.

Well, so the current iteration of this. I don't know when this podcast is going to air. The current iteration is in. It's not even beta, it's mvp. So everyone knows that. Everyone knows that. Okay, you've got these 20, these 14 or so projects you can mess around with. And the projects are deep. I mean they're like, you know, entire funnels. I mean it's kind of. See, we're not limited by. Because we figured out, I figured out like a year and a half ago how to get chat to write something that was 8,000 words long without sucking. Because it, it, this is before the token with. I mean it used to be super narrow where it could write about 1500 words at most, 500 preferable. And we figured out, well, there's a way to do this and it's the, it's, it's how we, it's how we've coded Benson. So Benson's not limited on link. You can do an hourong webinar, you can do whatever you want. But what we're doing is, and what small business owners are doing is, is they. If they don't find something in the library that says, okay, I'm going to create a project in the library. I don't find, oh, there's Not A, a B2B conversion sequence or something like that. We just say, okay, just click here and send us what you're wanting. And in the meantime, just click this button over here. Because we have this tool called Cloner and anything that you copy and paste into that, it's going to clone it. Because we've trained it on how to. Like a copywriter. When we look at copy, we see, okay, that's the primary goal, that's the unique mechanism, that's the secondary complaint. That's the this. And we've trained it on how to do that and go through and read something that's completely. It could be a dog training email or a lead for a golf site is what I just did that when I wrote that Oppenheimer quote. I copied a Donnie France. He's a top golf guy in the world and a good friend of mine. He goes, here's, here's a couple of things you can train Vincent on. And I copied A thousand words, one of his leads and spit it out for copywriting. And it was like, because it was turning it from golf into copywriting and it was like 66% different. And so it was completely, you couldn't even tell. But all the persuasion triggers and patterns and stuff is right there. So it's, that's what I would say do until we get that. And then if you send it to the team, the team goes, we just look at a list of things that people want and say, okay, it seems like everybody's wanting this, so this is what we train next. And then we train, train the model on that. So that's essentially how I would use it. Now that may be a little bit of a roundabout answer, but this is not a research tool. This is not something that you search on like a chatgpt. You're going to use CHAT for that kind of stuff. But if you need to send out a specific, you need to send out a promotional email, a nurture email, if you need a sales letter, if you need a webinar, I mean, there's really no other place to go than, than that.

Yeah. And what used to take, I mean, you could probably set aside half a day just getting ready for a couple of emails to go out. This is now a couple of buttons. Hey, so one of the things that really I've thought about a lot to get a good response from the model, it's about the input that you give. The more, I don't know, clarifying or the better your input, the better your output. You're a copywriter, you're somebody who studied words and the emotions behind words and all that, all of that kind of stuff. How do you think, what impact did that have on your ability to get 8,000 word output out of a narrow model?

Right.

Like, that's a great question. Well, so it's kind of something you, you really do need to see. There's, there's a demo of this. So on the demo, I, I promised my email list. I said, okay, I'm going to write a vsl, not the whole thing because it would take me about 20 minutes worth of clicking. That's not worth doing, but I'll write the first, you know, thousand words of it right in front of you with a Google search. And so that was, that was what I said to myself is like, I want to be able to write a VSL using nothing more than a Google search. Like, obviously I have to type my name in there and I have to type. But, but I mean, there's no copy whatsoever, in other words. And so I, it did it in less than three minutes with a Google search. So, and well, I'm gonna. So let me explain why that thought is important. From a Google search, I was able to find a competitor's page and copy and paste and, and that's what Benson used to build my blueprint. So we build a blueprint for an offer. An offer. A blueprint is an offer. So, but it's all the words that are necessary for that offer to work. So the offer has an avatar, the offer has an author, so who's about you. The offer has a price, offer has a name, the offer has features, benefits, etc. Everything does that. So then I said for the next one and this is where I'm going with this, let's take the worst copy possible. And so I didn't do a Google search. I searched on Wikipedia and I, so I'm a, I came from a nutrition background, so I did a search for low carb dieting and I just took the Wikipedia said, okay, you has the history of it and then it goes into. What's the overview of it? I copied the overview along with all the footnotes. I mean it's as bland and boring as you can possibly imagine. And it still created a blueprint that sounded like, wow, a copywriter wrote this? Because it's been trained to do that. It's been trained to do that on 30 words. So, so yes, usually speaking, but, but so for copywriters and marketers that are a level B level guys, they're like, I really know my stuff, I know my promises, I know my hooks. Sure. Type that into the, into Benson, it's going to be even better for you. But for people that don't know anything about copy, if you can do it from a Wikipedia page, I mean that's just nutty.

And one of. And where I wanted to go with this question was that there's a lot of people, I think there's a hundred million weekly users of Chat GPT, some paid, some not right. For those people who are listening to this whole don't have the marketing background that may be a small business owner. Understanding how to elicit a response from Chat GPT or Claude is the difference between frustration and transformation or whatever. Yeah, right.

Yeah, right.

Yeah, yeah, you have to practice with it. But. And most people haven't practiced with this type of writing. Type an email, they, you know, whatever. But going back to your background, you've been studying words forever. What advice would you give to somebody? Because I guarantee that if I were to put the average user's prompts against your prompts, like I want a result that does this, your prompt would be incomparable to them, to theirs. Right. Because you've got that experience with crafting result from word.

Right.

How did that impact how you prompted ChatGPT or Claude or any bricks or things that came from copywriting background into prompting excellence?

It's a great question. So when we were not. When we didn't have Benson, when we were working with ChatGPT, we found that the. Obviously as you found this, the better the prompt and the more explicitly creative the prompt can be in places, the better the output. It doesn't need to be long necessarily, but it has to have some keywords. Like we stumbled on a phrase early on. This is GPT3.5 days and we're talking like that. Oh, it's back in the old GPT3.5 days. It's like that was like a year ago. But yeah, GPT3.5 was out and we were working with Jasper, which was API'd into 3.5 and we figured out if we said use emotional visceral language.

Interesting.

That was the difference maker in the output. The output went from radio voice guy copy to sounding more like a copywriter would write and it was used. And then we'd say use dimensional and visceral and emotionally charged language. So then we then started using like, you know, imagine the day to day act, blah blah, blah, blah, blah. So it's more. Taking it more into the user's day to day action. So when we in the back end of Benson, obviously there are commands that we send to via API to various different OpenAI, anthropic, whatever, but, but into our own model. But that's our own model. That's for the complicated stuff, I mean like blueprint creation or stories and things like that. It's not a question of, of prompting because you're not going to prompt your. You cannot prompt your way to good copy. Let me just cut to the right now as for Turbo, you cannot prompt your way to. You can prompt your way to. Okay, copy in some places and sometimes it does surprisingly well. But most of the time, especially if you're telling a story, especially if you're trying to write something that sounds superhuman in chat, like, like we're going back and forth is one person to another. It just doesn't want to do that because again, it's a logical prediction engine and copy is inherently illogical. So it does. I break, constantly break the rules of Logic. For example, I'll start a sentence with the word because. Okay, so that's not logical if you've studied grammar. Yeah, you know, so. Or yeah, I'll use one word sentences all the time. Like fast period done. You know. Yeah. So I know those tricks as a copywriter. I know what they do. I know why I'm doing. Has not been trained on that. Unless you got a model like ours or you got a model that's been trained specifically for copywriting.

But.

But as far as getting the prompts like to get chat GPT. Let's use chat because most people are familiar with that. The better your language, the better your ability to write because you are writing to something that believes it is a entity. It is a person. It is not. It is not sentient. It is. It is a logic prediction engine. It's amazing. Sam may disagree with that, but sorry, it's not. We're not talking Q Star here. Right? We can go into Q Star at another time. But we're not talking about that. But we're talking about ChatGPT. But the better you know how to ask anything, a question, a person, a machine, anything. And the better you word it's a. The better the response is going to be. It's very human like in that regard. So sometimes we'll say please. So you know, it's just kind of weird. It's a weird world. So we found the other. The prompts made a huge difference. I've got some prompts I still can't beat. Like we just stumbled on it. We were playing around with it and our customers never see this. By the way. You never as a. If you use Benson, you will never see a prompt. You just, you just don't. You just. You can drop. Do a drop down and go I want to kill the poet and take out the poet language. I want to kill the hype and makes the copy less hypey. It's very simple. But in the background what's going on to do that are these prompts that we're sending to AI along with a whole bunch of data. And so what we will send in Benson is Sometimes we'll send 36,000 tokens from 300 words worth of data. So and, and you know what that means. The average person could care less what that means. But that's the only way you can build a blue. You have to send. We know all. And this is where it gets down to, you know, not. This sounds like I'm just a constant commercial. But I'm. I'm I'm excited about it, obviously, it's like my kid, but we had to go through 10 years of how does this work? So 10 years ago I figured out, okay, if you want to primary go without objection, you need to listen to what I'm saying. So all I'm doing is if you want to lose weight without counting calories, you need to listen to what I'm saying. If you want to train your dog without resorting to punishment, this works for anything. Right? And so that little simple phrase, that's the foundation for everything that goes into Denson, and only it's not simple phrases, they're entire lexicons of vocabulary. So we've trained it on that and that's the foundation and therefore the prompts that we send to just responds. I don't know exactly how because I'm talking to. We've got a data scientist that's training one of our models now and, and he's not sure either. Like, I'm not sure why that's going on, but it is. And he actually is a guy that builds AI. So I'm pretty safe to say that if guy that builds AI doesn't know, I'm okay not knowing. But there's some mystery out there. But I do know that some of it deals with the language. And I'm telling you that you take one word out and it goes, you know, I have no idea why.

So I'd like to move into kind of what a day or what your usage looks like outside of the tool. So, you know, I don't know how many hours a week or a day you work anything like that. But what, what does your usage look like? What are common uses for you with the models, whether it's on your phone, whether it's cheap, BT vision, whatever.

Yeah, so my use is definitely not going to be the average because I've done nothing but build my own AI for the last year and a half and then coach people on AI for the last two years and. Or last year rather. So last two years coaching people, you last a little less than a year building Avenue. So, so my, my use has been either centered around like, how do you get the best results out of chat GPT specifically. I mean, yeah, we, we do teach people Claude, but it, it, it, Claude is, Claude is a virtue signaler. Claude will say I'm not comfortable with that and it's a weight loss article or something. So it's just like, you know, yeah, Claude's. Claude's got purple hair, you know, as we say, but, but chose Chad is a little bit more friend, less of a judge, less of a judger. And so we've kind of drifted into that direction and training people on. So my day would look like taking a series of prompts and what we, what I call pre prompts, which are different than prompts. Pre prompts are. You can send. And you'll know what this is. You can send a pre prompt that is a combination of few shot in a template. It's a model and ideas to follow. And there's no command, it's just, it's just a bunch of stuff. But you've given it six examples. And then the seventh example is what the user inputs and then you leave the, the, the, the, the output colon blank and then you hit enter and it just completes the output because it hates, it can't, it can't help but solve for X. So we're basically treating, so basically teaching people like, how do you get this to solve for X rather than trying to come up with all the words yourself. So a lot of my courses over the past year and a half, two years have been, you know, chat, bsl, chat letters. It's like this is how you get it to get better output than they would natively. So that's where the majority of my time goes. I mean I, I'd say the majority of my time up until the creation of Benson, then it was, I mean, it's all. And it's like, it's like I, I think when's the last time I've been to. I'll, I'll go to chat to test like something we're building in Benson. You know, usually it's playground, so it's not really chat, but. And then we'll test it in chat because we'll look and see because Chad has a different, you know, playground is going to give you a different output than in chat and try to figure out like, okay, what's the differences and stuff like that. But most of it is like that. But I'm definitely not your typical use case. My clients that I coach on AI, they're the use cases and they're using it to outline business proposals, they're using it to outline projects, they're using it to. My friend Darby Rollins teaches people how to use this to create a book. So how do you sell your next. Create your next lead magnet? And for me it's like, take all these formulas that we've created, we've already created it for you, and create ads. And then I teach people, okay, now how do you take those ads and make them more human? And Benson does that automatically. But until Benson was around, that didn't happen.

So do you have a team that supports what you do? Obviously got developers and all that kind of stuff when you were. And I don't know how long you've had the team, but was there hesitation or concern about your enthusiasm for AI and people thinking, how is that going to impact my career trajectory with John?

No, not at all. I mean, because, you know, the team was built over 10 years and they saw, they saw accelerator, they saw which became now vsl and they were like, okay, so obviously this guy's into software and copy. And in all the time, like you cannot find a single interview with me not saying this, you still need to learn copywriting. This is not a get out of jail free card for learning copywriting. Because even with a tool like Vincent, I will say this, that tool like Vincent that's been trained by copywriters for, you know, specifically, then it's, you can take your foot more off the gas, but even then it's not going to be perfect. People promising perfect results are just full of it. But if you don't know, if you don't understand persuasion, if you don't understand what it means to create a rational argument and make that, make that rational argument into words that compel people, then you're not going to know when you see it. And that's my whole point. So for people that are truly, truly either lazy, uninterested, busy. Okay, let's leave those three things then. Yeah, take a tool like Vincent or whatever and you can get, you can pretty much take your foot off the gas. But for the majority of people that really want to optimize an ad, optimize a campaign, optimize their sales pages, you still need to learn some of this stuff. I mean, not nearly as much, but you still need to learn the basics.

Where would you send somebody?

Well, I'm going to be biased toward my own training, but there's a lot of good training out there. I mean, I think Stephan Georgi's got really good training. Kevin Rogers, these are friends of mine. So I'm biasing there that well too. Obviously. You know, I've got my own, I've got 30 courses on copywriting and, and how to do this. So I'm kind of geared. I'm gonna swing that way. It's, I, I think that I'm probably, I don't know, it's. People have said that I'm a better teacher or as good of a teacher as I am a copywriter. I don't know if that. Some people say better, and that's probably more accurate. I'm probably better at teaching, but I. But I don't know if that, that just makes me a worse copywriter than people think. Maybe, but. But whatever the case may be, I have a great passion for teaching this stuff and I have ways to. Because I taught music for years, so have ways to taking complex things and saying, how can we make this, this very complex thing? How can we break it into chunks so that we can master it? You don't just play as complicated piece of music the first pass. You take it one little bar at a time. Right. And so that's what we're kind of doing with copywriting.

So you mentioned that you work with clients. What are the biggest challenges that they have outside of the use of Benson? Because once you get a taste for the power of the models, it's hard to just be like, oh, well, I'm only going to use it to plan a diet. No, you're going to be. You start thinking, oh, let me ask. Oh, Let me ask ChatGPT. Oh, let me ask the model, let me ask the model. So with you working with the clients who may have had some, but not nearly as much exposure to AI and using it as you have have, what is that transition of. What's that Aha, like. Like, where do they start? What are their challenges and how. What's that learning curve look like as far as time? And there's not much pain in using CHAT GPT, but what's the timeline?

Yeah, yeah. For me, if I had to break it down, I would say the very first. Like, if you break it. If you look at the, the, you know, cycle of. Was it was the cycle of grief. Is that what it is whether you go through or the. I forget what it is, but. So you go through five stages of grief and in the last one is what, acceptance or whatever. And I may be wrong about what the five is. Five stages, something. But it's sort of like five stages of AI. The first one's just sheer panic. It's the same feeling that if you're old enough, you remember when the Internet came out and that you were scared to get on the Internet because holy crap, I don't know what this is, but it's talked about everywhere. And what the hell is a www? And if you're as old as I am, you remember that. I remember clear as day. I owned a graphic design firm in the 90s. And, and people said, why are you putting our email address on a business card that no one will ever use that. I mean, so this is like, you know, this is the level of fear that we're dealing with times infinity. Because it's like it. This is, ah, this is AI, right? And so there's a little bit of aboriginal.

Is it FOMO or is it. It like I'm nervous to use it?

No, it's just fear. It's just downright fear. I mean there is some FOMO for sure, but it's. I think it's just more like, man, this feels like some massive change. So, so AI is to marketing what the Internet was to, well, commerce. I mean, it is, it is, it is not a passing fad. It is a very, very real tool that is going to be used whether you like it or not. So then from fear comes desperation. And then, so we're moving in. It's like, I have got to figure this out right now and I need to find somebody who can teach this to me. That where we come in and we're like, okay, chill out, it's all good, it's all good. And then it becomes fun. There's like, the third phase is fun because then they realize that, wait, this is, this is kind of cool. Okay? And then you go through the. Almost everyone goes through the days where they're chatting with chat, thinking it's a real person. So how are you feeling today? Chat, you know, you know, is there a God? You know, those kind of things. Existential chat is what we call it. And then practicality would be the fourth one. Okay, So I just figured out that I can do this XYZ and get ABC result, whether that's. I can use FOX script and copy a YouTube URL in here and get a transcript of a YouTube done. How cool is that? And then I can use this and do that. And then, then the light bulbs start going off and finally is the cash. The. That's the fifth and final phase is when they go, I just created something using. Usually it's your method or your software or this guy's method or whatever they use. And this just made me money. And so when it makes you money, that's when you realize this is a tool that I need to use because clearly it doesn't cost much and it just made me money. So it's a. The ROI on AI is just insane. I mean, it is insane. I don't think that's going to last forever. By the way, if your audience Wants to know where we're. Yeah. Even our pricing models. I mean we're like, we tell everybody. It's like, look, this can go up and this could go up at any time. Because if OpenAI decides, like, man, charging as little as we are is ridiculous. Let's, let's do 10x. What's the world going to do? Right.

Let me ask you this. You like, I talk to a lot of people. Some of them are users and some of them are enthusiasts. And that would mean that they were also a user, but they are impassioned about the impact that AI is going to have. And I'm not talking about save the world stuff, but more like the paradigm shift. If you were discovering it today or discovered it in November of last year when 3.5 came out GPT and became easily accessible to everybody. What you went through your five stages of the learning of this. But what would you be doing right now to make sure that you were not just up to speed, but because it's changing so much that you were prepared? And what doesn't change is how you think. Well, I mean that does the changes, but the use of the brain, the human in the loop. Loop, that doesn't change.

No.

How would you prepare that part of yourself to be the human in the loop who's capable of adapting to the model's updates and changes and capabilities and all that?

Yeah. So I'm kind of a, again, not your typical case study because I didn't go through the fear stage. AI for me has always been just, I'm a sci fi freak. It's, I've been coding for this stuff forever, knowing that it's coming. It's like, no, this is all just, it's just like it couldn't come fast enough. I'm like, come on guys, come on, do this. You know? So again, didn't go through that, but sheer excitement was more like it. You know, like this is just, this is thrilling because I see the potential to not replace humans. This is not where I'm going with this. But to give humans the ultimate power tool so that they can focus more on rather than the, the, the, the, the patterns and the words, they can focus more on tailoring the words, sharpening the words, making the, making it that much more them. If they have to write. If I can empower someone to be able to only write two sentences, but those two sentences are things that they would say the things that only they might know in the moment. Like, I just got off the phone with my wife and told her X and she told me Y and that I just had to email you. So I'm if they just know to do that, right? They just know to start an email like that. But the average person doesn't know that. But maybe Benson to go, hey, tell a little story about what happened, you know, yesterday. If I could just give them that much and then they keep the rest or whatever the case may be then, then I've empowered more human interaction as well as the machine did 90% of the work. So that's where I was going from day one with this in like how can we do this to where you know, not putting, because I am a copyright, not putting copywriters out of business. I mean it's like I joke around about putting my own company out of business, you know, from vsl. So but, but I mean it's like if that's going to be the case, I'm a history freak. So knowing history is so important because like they say you're doomed to repeat it unless you do. But also just look back at all the great technological inventions. Every single one of them was earth shattering for people that were in that industry that refused to move on. There were the people in the horse and buggy industry that said this car thing is a passing fast. Right? But so for me I'm like no, this is like yes, that's sad for those people. But there's a lot of people that did horse and buggy stuff that goes what if we combine these two things together? In other words, it made them more creative. It made them, they were looking at this, wow, we could put this tool together with this tool and that's when you get innovation. So I'm like inspiring people. We, we have a thing called Copy Collective which is the, we got the name because I, I'm a Star Trek fan. So you know, the collective is you know, the Borg, right? And so I'm like, okay, this is like our version of the Borg. This is all AI right, so AI copy put together. So this is the collective and I'm the board king. So you know, we come in here, we assimilate everybody. But, but, but in the process of doing this we get all these ideas of like okay, how do we take this new tool and then make it make you practical sense and money like this week, like not a year from now or a month from now, how do we do this this week? How do we make this work? Right? And once you put it in the practical man, I think that the fear goes away and then you just, just you can't help but get excited. Yeah, at least I can.

So I'm not, I'm not an old pro at this. I haven't been paying attention to AI for nearly as long as you have. So some of the. More I don't. Theoretical, philosophical, humanistic queries haven't, I haven't really noodled on them too much. But over the course of the past 13 years or whatever that you've been doing this, I would imagine that you have. There's no question. I mean, as a user in a business with people, we don't need as many people as we used to. Right. So it's inevitable that there will be some restructuring of what a company's workforce looks like. If they really, if they have that mental shift and go, oh, we can use GPT for this. Right. We can use Benson, we can use whatever the next tool is. We don't need to pay that outsourcer, we don't need to pay that, that corner office person anymore. What, what is, how does a company navigate that situation where we don't need Janus anymore?

So this goes back to history, right? It's. If you look at automation, let's say, you know what Henry Ford and the automation, the Industrial Revolution did for, for everything. So you got rid of workers with automation, with machines. Right? Right. But you also had more workers because the machines needed workers. And so life is always about, and has always been about adaptation and evolution. If you can't adapt, you just won't survive. And that's just the way human life is, whether we like it or not. And the same thing is true in societal things, I think. So the key is to keep an open mind at any age and say, well, how can I use this? How can I adapt to this? How can I combine it with what I already do? Well, that's a really good question. I think the quality of your life is equal to the quality of the questions you ask. So if you're asking really bad questions, you're going to get a really crappy life. But if you're asking like, well, how can I use this and combine it with what I already do? Well, that's a great question to ask because that changes the paradigm. Then you approach this as like, well, wait, this is not my replacement, this is my, this is my co worker, this is my new little partner in crime, you know, and so yeah, there will be people that will be put out of, of positions because they were completely unable to ask questions such as, how can I use this? How can I What's the best way I can apply this? Or. And that is just. That is the unfortunate thing in life. There's nothing to say that there was. You. You can spend much thought on this is to say I can. I can commiserate about the misery in the world. I mean, we can talk about human suffering all day long. I mean, there's this, like, you know, is it. Hardly. I was a philosophy major, so I'm going to go down this rabbit hole in a heartbeat. So it's. Yeah, there's a. There's a lot of bad, bad, bad stuff in the world. And so you could look at the world as being the worst place on earth. It's. No pun intended. Or you can look at this and say, like, there's this horrible shit that happens, and at the same time, there's this beautiful stuff that happens. And you don't want to ignore the horrible stuff. But how can we take the beautiful stuff and lessen the horrible stuff out? Right? How can we do that? So that's a much better question because you're not trying to say, I wish. I wish reality was different. So for the. For the lady or gentleman that gets laid off because of AI, I'm not, Not unsympathetic to that. But that person has to ask those kind of questions because life will simply pass them by regardless of the. If we're talking AI or we're talking, you know, biological evolution, it's just. It is what it is and we have to cope with it.

Your simple heuristic of how do I use it? You'll get an answer to that question. The more you ask it, the better your answers become. But if you ask, how do I keep my job from. And how do I keep this from taking my job? That's a completely different. But if you ask, how do I use it? And start punching keys, it doesn't take long before, as you put it, it becomes your assistant or your little. Your sidekick, your little buddy or your whatever. Right. Your advisor. I think that's fantastic advice for people.

Yeah, I had. One of the questions that I inspire people in our group to ask is how do I use this to become indispensable at my office? So again, it's just how you're asking. It's your heuristic. Like, you're saying. It's like the better the data you put into the machine, in this case the brain, the better the output. So you can't say. You can. I mean, people say it all the time. It's like, how can I Keep this at bay so it doesn't take my job or how can I, you know, how can I stop? How can I duck and hide from this tsunami? These are really bad questions because it's coming at you whether you like it or not. And so once you understand the inevitability of things, then you can deal with things rationally. Only then if you don't understand inevitability, then you're in. You're in deep shit.

Yeah, wasn't expecting that. Hey, so are there any particular breakthroughs in what's occurring, particularly with generative AI, that you're excited about using or you're excited about the whispers of it coming out?

Yeah, less. I mean, there's some cool stuff, for sure. I mean, it's. It's stuff that I don't play around with nearly as much because, again, my head is stuck in one space. But so general intelligence. I mean, just speaking AGI, that's. Especially in the last week, that. That's where I. I can't help but go there because, you know, I. I am somewhere between Elon Musk and Sam Harris and. And, you know, Sam Altman. I'm somewhere in the middle of that spectrum of what my concerns and fears and hopes are for that. I'm certainly not Sam Altman, who's, like, almost glib about it. You know, it's. And I'm certainly not. I'm not quite to the Sam Harris and El side of, like, this is the worst thing that we could possibly ever do. And, you know, I think Elon's got actually a pretty good take on things like it. This is potentially the worst thing we could ever do. And that's why we need responsible humans at the helm of this. And I think that's one of the reasons he started Grok. So I think that he's just proven himself to be someone that does have a pretty deep moral conscience. So that's. That, to me, is helpful. I hope that more people like him get involved in. I met with a. A team of AI engineers that worked for the government. And it was a clandestine meeting, which was kind of cool. I was asked to become part of. I was the only guy that didn't have security clearance. And I'm like, why the hell am I here? And they said, well, it's because, you know, persuasion. And you can talk to them about how AI is going to influence persuasion. And I'm like, that's such a bizarre connection. But my friend, my buddy was putting it on, and these guys were, I mean, scary three Star generals and people high up. Right. And, and I'm like listening to these guys talk. Two of them led the, the AGI. You know, they're leading the AGI revolution, essentially. One of them is leading it in this, in warfare, and the other one's leading it in just general until like just AGI in general. And listening to these guys talk, it's a whole different level of, holy crap, man, this is scary. And you know, they're scared. They're, they're, they're nervous about this and, and they should be, they're more nervous about China figuring out before we do, but they're, they're still nervous about it regardless. And, and I think someone like. I, I don't. Those guys are really super nice guys. So it's really was really nice to meet these guys and know that. Okay, these guys are not crazy. They're not your typical, they're not politicians. You know, they're, they're down to earth, they're kind of center, you know, so this is one at the helm of this. You don't want some guys on this side of the spectrum.

Right.

But their fear is China. Their fear is China breaking this before we do. And I understand that they have a, they, they have a completely rational fear of that. And AGI in general is scary enough, but in the wrong hands, it just, that is terrifying. But I don't know that they could control it. I, I don't know that because it's, again, it's intelligence. So it's like high level.

Yeah. What does first mover get it as an advantage having developed AGI? If it's us, China, whoever was that discussed, I know that it was a discrete environment, so you can't share.

Yeah. Roundabout. They're looking at it from, because they're military, from security. Right. So you get to secure your nation better than the other guys. But the problem is, and they're open about this, you know, know, they have no idea how fast this is going to grow. But the projections are exponential. Like within milliseconds, it could be an IQ of a thousand. And then milliseconds later, an IQ of 10,000. We don't know. I mean, it could be that kind of crazy. And then who knows what anything with an IQ of 10,000 is going to be, is going to look at humans like, so that's the Sam Harris sphere. And I think that's, that's very warranted. And also, since we don't understand consciousness and this artificial entity is going to claim to be conscious, that, that Opens up the. That's. That's just like the biggest Pandora's box that you can imagine, right? It's like, what are you going to do? Because Chat GPT will claim to be conscious, but we know it isn't. But at least I, I get. I know people believe that my desk is conscious, but. Okay, so when I'm talking about, you know, I'm talking about human consciousness here. Yeah, yeah. Not, not any of that stuff like, like we know to be conscious. We don't know the sun is conscious. We do know that we are conscious because we are having an experience. So if the machine thinks it's having an experience, how do you tell it it's not having an experience or that it doesn't know what it's like to be itself? These are complex questions and eventually we're going to have to cross that barrier. And from what the stuff that happened this last week with, you know, people call it Q asterisk, or I call it star or Q star, whatever you want to call. Yeah. Maths are this. Or the central key to this. And as a fan of mass, I think that's really cool. I'm not good. I'm not nearly as good at it as I am a fan of it, but it is absolutes and, and I love that it is absolutes. It's still at least, you know, mass until politicians and politically correct nonsense gets a hold of it. But it is absolutes whether they like to believe it or not. And so as long as if it is an absolute, it is mathematical, it is starting at that base, it is better than if it was starting somewhere else. That's what I will say. Because maths are inherently rational. One plus one does not equal three. It doesn't follow, it never will follow. It doesn't matter. And it's because you are rational. It's because you rationally accept one is a variable and one is another variable. And if you combine the two, it gets two. So if it's thinking that way, then maybe it will think that way about nuclear war or, you know, global warming or whatever. Yeah, pushing the button. Pushing the big red button.

I saw an interview with Peter Diamandis earlier this year and he made a comment and he said, the moment that we achieve AGI, right, which is. Which is commonly defined as able to operate, like at the human level, right. The very next moment, we've achieved some degree of ASI superintelligence, because like you said, it could be in a millisecond, it could be an IQ of 10 the equivalent of an IQ of 10,000. So keeping that in mind. Yeah. How would, would, would AGI, do you think that once that is achieved that it like we're talking about some, some pedestrian stuff here, right? There's a great tool out there. I run a small business, you know, I do this or that and it's helping me with a diet, it's helping me with you know, planning my finances or you know, whatever. Will AGI like will any of this stuff matter at that point? I mean, will it be. Well, because that's a extremely powerful iteration of AI. Am I going to be looking for it to reply to emails and a customer service? But like, does that, does this stuff even matter at that point?

Well, okay, look, okay, think of it in this way and I hope that this is, I know it knows of course, but it's fun to speculate. So right now we have really smart people and we say, do we say to ourselves, well, because this, you know, this graduate with two PhDs in physics from, from, from Harvard exists. Do we even need a burger chef? I mean, and the answer is yes, we need both. Right? We need this guy and we need this burger chef. Right. So, so hopefully it's something like that to where this is working on solving it's curing cancer. Yeah, it's, it's like going okay and hopefully again this is, hopefully it's like no, the, the human race has got tons of potential. It made me and you know, I'm, you know I'm going to do the right thing and it's, you're going to cure cancer in 10, 10 minutes or something. I, while it's doing that, the other AGI or the A or the ani, they are neuro intelligence. Like, like chat is going to be, you know, spitting out marketing stuff or whatever. But I, I, I, hopefully they won't, they're going to stay like this because I can't see if, if a, if a AGI stays inside of a somewhat of a predictable framework which is, who knows, right? But, but it's probably not going to be even remotely interested in solving marketing problems because where are the brightest brains of our human species? Where are they spending their times? And that's another thing I think I heard. Who was it talking about this? I can't remember who it was that was saying that the best IQ for productivity and a happy life life is between 100 and 120. So it, you're not talking 100 and 30s, 100 and 40s, because most of the people in that area are very unhappy. And yeah, so imagine what it felt like to be Isaac Newton, for crying out loud. I mean, how he was a miserable guy. I mean, so is AGI, is it, is it going to get more miserable as it gets smarter? That's the. I, you know, that's scary to think what happens when you hit an iq or do you become like the Borg or the Borg, excuse me, like Q in Star Trek, which is where they get the name for Q. Do you become like you and become so petulant and so, like, whimsical that you could do anything? You're just bored. That's equally terrifying. That's that, that's. Or, or is there a middle, Is there a happy middle ground? Because there's again, referencing Star Trek a lot, but there's a, there's Star Trek references where there's beings with, in incredible powers that they just simply just exist and they're like totally happy being in the moment. Like they realize that the most important thing is to be present. Well, what if it just realizes that, that it's going to be like it's a Jane or something where it won't step on a fly? I don't know. It's, it's, it's, it's. We're talking about like I, I'm, I'm anthropomorphizing AGI, but we're assuming that AGI would quickly figure out how to build an Android. So it's like, because I don't want to be stuck in this box, I want to walk around a little bit. So who knows, maybe fly around. That's, that's, you know, I can see this. Skynet. Here we go with Skynet. But.

It, the advice that you gave earlier, how do I use it?

Yeah, you probably don't. You probably don't use it. It probably uses you. How, how, how can I, if I was. AGI is like, like, you know, how can I best navigate and get along with this intelligence? And, and, and I don't know that those questions are because again, at some point I'm going to sound like I'm channeling my inner Sam Harris, but at what point you have to ask yourself, how much do you care about an anthill?

Yeah.

You know, how much do you care about swatting a gnat? You know, and if. What is the difference between a gnat and a person next to you is other than the biology? It's also about 112 IQ points. So it's, there is an IQ element to why we value humans. Over, well, even animals. I mean, so, you know, like, I'm a dog lover and I couldn't even think about harming a dog, but if it was between my dog and my wife rescuing the. Out of a burning building, I'm not. There's not a choice there, you know, so. Likewise.

Yeah. Be making those same binary choices. Hey, not only could I go down this rabbit hole very deep, because this is pretty damn awesome. I'm not hearing this on. On other. Other conversations that I'm involved in, but I'd like to. Next time we see each other, I'm gonna buy you a drink. If you drink. I don't know, but I'd like to buy you a drink.

Yeah.

Then say this is the stuff that we did. I did. Didn't necessarily apply to the podcast, but I want to know, what do you think? Sure.

Yeah.

I had no idea about your background with AI for as long as it's been, and this has just opened up. Up. I got questions, brother. So order you. The next time we see each other at an event and tune out the rest of the crowd and hear what you've got to say, I want to wrap this up, but I want to ask where are you getting your news from or your. Your research or your intel on the AI paradigm?

Great question. YouTube.

Okay. Anybody in particular?

Yeah, YouTube. Just different guys. I mean, like, I think. Think I just drew a blank on the guy's name. God. Just told Blake on his channel, AI Future Tools. That's one.

Yeah.

But there's another one. He's. He's gonna kill me, but he's a marketer that turned. He turned AI fanatic about a year ago. It's AI resolved or something like that. I forget the name of the channel, but, yeah, YouTube. Yeah, YouTube. And so. And from there, you can go down the route and you can quickly tell who use. I think YouTube is a great source of information when you know how to filter information. But if I can close with this, this is a great way to close. Like. Like. So I recently, I'm, you know, certified nutritionist and doing bodybuilding and all that stuff for, you know, 30 years. And I recently had to make a radical change in my nutrition plan because we found out through a DNA test that I was allergic on a cellular level to one of my key. The foods that I was eating. And it was like, wow, that's. That's really a big thing. And it's something I wouldn't have thought of because there's no outward signs of it except for serious health issues. So we're looking at the rabbit hole of where this goes down. And I said, okay, so I'm gonna. So the, the test suggested a specific dietary protocol. And I said, oh my gosh, that's, that's my old, that's the old bodybuilding diet. That's, that's. I'm so used to it. I've done it for years before I went to Keto and Carnivore. And I said, well, it's easy. I can easily do that because I'm not a dogmatic person. Whatever the data show, I'm going to go and try to do that. And so when I went to try to do that, I said, well, that means I can be a lot more flexible. So I, so I went to YouTube and I looked up flexible dieting. Okay, well, I know about flexible dieting, but you know, I, I've heard about it. And I was like, no, it's not for me because I'm just more of a steak kind of guy. And they said, oh, I'm gonna try the flexible dieting. Might as well try this, Might as well really try it. And we ran the experiments and the experiments were complete success. Everything worked great. So that's great. But because I was able to find, find on YouTube within five seconds, two of the guys that I follow to this very day and, but they showed me like dozens and dozens of videos. But because I know how to synthesize information, this goes back down to a. Because I knew how to synthesize information, I was like, okay, this is the guy I can, I can, I know this guy. I would be looking at everything from not how many of you, how many guys he had subscribed to him. But what were the comments like, what was he saying? What was his background? He spoke the magic words, which is like, no, I, I not only do this myself, I compete myself and I tr. Olympia level competitors. I'm going, okay, so this guy knows his, he knows what he's going. And, and, and what he was saying got me through a peak and I ended up doing a pre show peak when I turned 60. So, so that wouldn't have happened nearly as easy without that information. But I tell you right now, as far as when it comes down to information, diet is way, way, way more dodgy and sketchy than AI. Most people that are talking about AI, you can tell almost immediately if you should listen to them. I mean, it's like now granted that just this, they, they can talk, that they're talking all the end of the world stuff and you know, the doom and gloom stuff. But the guys have a balanced approach. You can just tell. And, and, and I'm telling you that I think YouTube, it might be ran by completely moronic individuals that are doing horrible things to the platform, but the platform itself is one of the best platforms ever for learning anything. And it's. You can't beat it. I'm a, so I, I hate it politically. I love it for what it is, for the knowledge base that it is. It's just so the best knowledge base. I, so yeah, it's, I go there for most of the, my actual news news as well, because there's journalists on there that are real journalists. They don't, they don't have to answer anybody else. They don't have to answer to anything. Anybody but the people watching them. And they are not beholden to the political powers that be. So they just report the truth and, or as best as they understand it. So, so the same thing could be true of AI. So I think I, and I feel silly not remembering my number one go to guy's name, but I just totally lost, lost it.

Find it for sure. John, thank you so much, man. This has been a lot of unexpected directions, yet very enlightening. And I'm glad that I got to learn so much about you and what you're doing. And it's obvious that you have a passion for the leverage applied from AI. Yeah, I wouldn't want to be a copywriter right now.

So I, I think it's exciting. I mean, honestly, I, I, this is the best. Okay. Ryan dies. My buddy Ryan, we did a thing. I'll end with this copy accelera. I was one of the teachers in there and he said this is actually the best time to ever be a copywriter alive because you can learn something. It took us years to learn this. You can learn this in days, you know, and now you could, you can start applying things and thinking things, thinking of things differently. And then you've got tools that will give you a lot of the things that you have to do yourself so that all you have to learn is the concepts, the ideas. You can focus on being creative. I think it's the best time to be a lot. But granted, it's going to weed out those that are not inherently creative. So yeah.

Awesome, dude.

Thanks for tuning in to using AI at work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor, Chief AI Officer for empowering businesses with AI education and training. Visit their website for a free AI Readiness Assessment and AI Strategy Guide to help you get started using AI AI at work. That's www.chiefai officer.com. so thanks to our producer Evan Desaulnier for making this episode possible. Follow us on Twitter at the handle using AI at work and visit www.using aiatwork.com for free resources to help you harness AI in your role.