Zach Hammer: [00:00:00] That general principle of being skeptical you have to if it’s information you got to validate that it actually makes sense if it’s this is the right way to do it you gotta see if it actually works in the real world and do trust me your experience on that now with that being said here’s part of what’s interesting when you partner AI with a process that is proven I know that this works. I know the right way to do it. I’ve seen it work. I just need to make aspects of this happen more automated. I need to make aspects of this happen at a consistent rate where somebody isn’t having to sit and type something out. Somebody isn’t having to do this process that really is just running through an algorithm. That’s where AI can actually be really powerful as long as you know how and where to deploy it.

Zach Hammer: What’s up? Welcome back to the show, AMP Intel, here with Zach Hammer and Charlie Madison. Today, we are going to be talking about [00:01:00] AI and is it overhyped? Is there more promise than there is delivery? Is there a more sizzle than steak? We’re going to be diving into that. Charlie, welcome back.

What do you think about this topic today?

Charlie Madison: There better be steak, not just sizzle. That’s what I’m saying.

Zach Hammer: I could go for a good steak. I can always go for a good steak. Shoot, I’ll take steak. I’ll take ribs. Ribs would be good. There’s lots of

Charlie Madison: I had a good Tomahawk bison steak this weekend at a Jiu Jitsu tournament. Like, How manly is that?

Zach Hammer: Man, I don’t know. Like smoking a cigar and maybe, like, I don’t know. I don’t know what else you throw into the mix, but, Yeah, because I mean you had punching people. Well, Jitsu, you may not have been doing a bunch of punching, I guess, but grappling people.

Charlie Madison: Did have someone get knocked out.

Zach Hammer: there you go.

There you go. Good call. Good call. Yeah. Love it. But yeah, so AI and is it overhyped? So here, like really this kind of comes on the tail end. I watched a video recently that talked [00:02:00] about some of why it might be overhyped. Talked about some of what we see being promised versus what being delivered isn’t exactly in alignment.

And so maybe let’s lay that out. let’s lay out some of what we’re seeing, cause you and I both work a lot with AI in our day to day work. What we’re seeing in terms of like the arguments that, when credence to this, that maybe it is overhyped, maybe that we’ve been sold a bill of goods? right. And really for me, I know probably one of the biggest pieces of evidence is what the companies announce versus what actually even comes out. And then also what is announced and demonstrated versus when you actually go into the tool, what it feels like to try and get close to that result.

A great example of this is Sora from OpenAI. That one was launched or what was talked about at this point over a year ago, I believe. I think they might have just launched. I just saw a video of somebody talking about it so maybe maybe they’re in the process of launching right now. But

Charlie Madison: Which is four [00:03:00] score in seven years ago in AI world.

Zach Hammer: Yeah, yeah, like, they talked about it. It was really cool. It’s a, text to video generator. That was one of the first things that we saw examples of having you know, object permanence where like, if my hand goes behind my hand where you can’t see it and it comes back out the other side that like the AI understood that. And it wasn’t like my hand comes out the other side and becomes a banana. That sort of idea. It was one of the a system that was cool for that. It was announced and it’s been like over a year and some people got to play with it, but most of us in the real world didn’t. And so you see a lot of that, you see a lot of you know, people talking about these cool, impressive features.

And really they’re like Cherry picking, selecting the best examples, you know, talking about the promise of the future, but when it comes to actually delivering on it, it’s really hard to get there. you know, You also see that with some of the image generators where they show the best case example of, look at this thing that you were able to generate. And in your mind, what you’re thinking is you’re thinking, the workflow looked like this. They put in a [00:04:00] prompt that felt easy, and immediately got that image back. And that’s not what it most likely looked like. What it most likely looked like is they typed in what they thought the prop should be, they got back something. It was like, Oh, that’s not quite right. And then maybe they try a couple more times, and then maybe get, they get closer. And then they go back and they mess with the problems a bit to try and dial in what it is. And then they ultimately end up with the thing. And No, you’re still left with a process that yes, you get a cool, impressive result. There’s still a lot of power in that, but it wasn’t this thing that felt like I put in some words and AI gave me back the perfect thing. It still took work, right? It still took actually having some effort.

What about for you? What are some of the things where you’ve seen before we got on this call, you mentioned that there’s times where I feel like AI is actually getting dumber. What are some things that you’re seeing on that front?

Charlie Madison: I was using it today to write a program and I feel, like, it’s a small program. There’s three files. And over the last week, I’ve been asking it to update the code. And it is [00:05:00] updating one piece of code that breaks all the rest. And I feel like maybe six months ago, maybe even a year ago, a lot of times when I was working with this, for whatever reason, it was able to have a much more holistic view. And I’m actually paying for a programming type. And my programmer said, you know, it’s not working as good as it says. And now that I’m in there, it really feels like it’s getting dumber. You know, Another really good example was, I had a really tough real estate negotiation, a few months ago. And so I was like, hey, act like Chris Voss from the Black Swan and you, Richard Bandler, with his NLP uh, linguistic programming and help me figure this out. And the end results were absolutely amazing. Times that I ran, I’d used it so much that I had to wait a few hours to use it again. Like, It took a lot of work for me to get there. The fact that I have knowledge [00:06:00] of Chris Voss, negotiation at NLP, I could tell when it didn’t work and it took a lot of finagling, but then the end result was something that I never would have come up with myself. So, like, It ended up being amazing. And I mean, it took a few hours.

Zach Hammer: Right. Yeah And you know, really I think, like what we’re left with here is we’re in an interesting zone where there are some aspects of this that are maybe a bit unprecedented. Yeah. Where we might be on the cusp of some of what’s being hyped actually being able to happen. right?

Like That is a possibility. But what we’re seeing happen is that there’s a lot of money available. There’s a lot of investment available. A lot of these companies are being encouraged. Hey, if you slap some AI features in this, that’s what people are looking to invest in. And so we have a lot of like. they’re adding terminology, language, and marketing speak in order to gain investment. And it doesn’t necessarily actually matter if the thing does what an end user wants. right? It [00:07:00] doesn’t actually matter if the customer, the person that actually needs the result, that’s not who they’re necessarily building for. Because that’s not where the big money’s coming from. The big money’s coming from the investment.

And so there is this aspect of what gets somebody to invest and the hype that gets somebody to invest doesn’t always align with what you’re actually seeing delivered. And really, I think there’s a few ways that this could be going. One is that in some cases, I think we might even be seeing that because that’s where the focus and incentive is, it’s causing problems to surface in the product itself that may not be like we’ve seen. Some of the things feel like they’re getting worse because they’re not building with that in mind. That’s not where the incentive lies.

In some cases, it might be that it was just never there. Some of the things that are being touted and promised, it’s more they’re selling on future vision rather than what they’ve actually been able to achieve. And so we’re seeing some of that too.

And really, you know, what we’re left with is we’re left with is all lost. Is there no value in this? Is it all hype with, no [00:08:00] value, nothing to get out of this. And the reality is, I think that is also too far of a pendulum swing in the other direction.

It’s not all smoke and mirrors. There actually is, right now, present, still, a great deal of powerful technology, powerful stuff that can really change your business. But you have to have a realistic view of what that looks like and how that comes together. does that sound right?

Does that sound like what you’ve seen as well?

Charlie Madison: Yeah. And I, think you’ve got to have expertise in the domain that you’re using it. Because I’ve noticed, here in a couple of weeks, I’m teaching some Jiu Jitsu classes. And so I was using AI to kind of come up with a curriculum. And at one point, it just went off. It was like the biggest hypey person ever. It just like went off in this rant, and I was like, I think you just totally made all of that up. Can you please get back to the original question? Oh yeah, sorry. But like if I didn’t know  Jiu Jitsu, I would have thought this was completely realistic. And so like I’ve found [00:09:00] that, I can’t trust it if I don’t know anything about the domain.

Zach Hammer: Right, yeah. And so that’s one of the skill sets that’s definitely like right now in this present world, one of the key ideas is be skeptical.

Be skeptical not only of what you hear the company’s promising, but also of what you hear the AI saying. And so there’s a couple of ways to navigate that. If you aren’t familiar with something, know that you really can’t expect it to be reliable, and so you will either have to also research it, or you will have to use a tool like perplexity or ChatGPT with search in order to be able to look at the sources and see how it’s arriving at its answers or conclusions, to see if you agree, to see if you align with it.

And if you don’t, expect that you’re probably going to either end up with information that’s garbage and that will ultimately prove itself out, or you’re going to end up with egg on your face when you talk about something being true that is not true. When you talk about something [00:10:00] going down the direction that’s just not real. And most of us don’t want to go down that route. And just note that you’re going to have to plan or expect that you’re going to need to do some research. You’re going to need to actually validate the information.

But it’s still useful and worthwhile to be directed toward those things, to gain a better understanding. Use it as an opportunity in these areas where you are interested, where you are curious to gain a better understanding, to learn more about what’s real and what’s not real, to, have it reveal your blind spots, while also having you get a better understanding over over, the subject matter at hand. Note that, right? There’s value in that. This thing that actually feels like a problem may actually be part of what helps you, as long as you’re expecting it. Cause I, I’ve found that. I’ve been, like, I ask a question and I’m like, is that real? That doesn’t feel real. Is that actually the case? And then I go in and I read the article and I’m like, Oh, I get what you’re saying. I get why you said this, but it’s wrong. Or here’s what’s off about it. And now I know better how to ask the question or how to think about it, or what really is working or not.

And [00:11:00] occasionally in some areas or some places, I do learn things, where it’s, like, yeah, this is a proven process, or this is something that’s pretty well documented and understood. It’s like, Oh, okay. yeah. So Anyway, so it can be useful on that.

On the other end, just like you said, you know, if you want to use it as a tool to help you step outside of your own blind spots, that could be useful as like a research partner, to research. But don’t trust what it says, it sparks the trailhead. It gets you started on the journey that you go and you actually read the source material. You actually research the thing to gain a better understanding and know that you’re going to have to, if you’re not an expert in that thing.

And then on the other end, know that if you really want to get good results from it, you still have to bring expertise into the mix to get good consistent results, that somebody who knows what they’re doing is going to be able to leverage it quicker, more effectively than somebody who doesn’t. I think coding is actually a pretty good example of that. I have actually started using AI you know in a lot of things to do some coding. But I could tell you, things that I know that somebody who [00:12:00] knows what they’re doing would be able to iterate on the stuff that I’m doing quicker.

They’d be able to get done what I did more effectively, just because they understand how these different pieces mix together. Whereas, I’m almost having to like back into an understanding, where it’s like, I don’t know what I’m doing. Make this functionality do this thing for me. And then it has to try two to three times to use a function that doesn’t exist and be like, I keep getting this error. What’s this issue? For it to finally be like, Oh yeah, this tool doesn’t use that thing. And I’m like, Oh, okay. So now when it comes up again, I’m like, that doesn’t exist in here. Do it this way.

But so like, I actually gained that expertise through that process. But it takes time. I have to know that I’m going into that to learn something. I have to know that I’m going into that in that way. Whereas somebody like you, you could potentially leverage it and get to that end result quicker, but you’ll still see what it’s done.

Does that make sense?

Charlie Madison: Yeah. a hundred percent. 100%. And you know what, what I’ve found, I think, it’s a skill on how to use AI.

Zach Hammer: Right.[00:13:00]

Charlie Madison: And the cool thing about programming is the machine tells you if it works or not, like you get an error code. The problem with other domains, there’s no instant feedback. You know, It’s like the lawyer that was arguing against using AI in uh, I think in, court, and his paper used AI and had fake court cases in it.

Zach Hammer: Right.

Charlie Madison: They printed out so much material. He didn’t, you know, like You can’t have AI fact check itself like well, as much as, you gotta have some outside validation.

And so with programming, it’s the machine doesn’t do what it’s supposed to do with the other stuff, like when I’m using it to negotiate or something like that. It’s my gut feel, in my experience.

Zach Hammer: And even more so than that, like that’s your first step. So you look at it and you say, you have to put it through that filter of does this make sense for me to try it this way? Does that add up with my real world experiences, [00:14:00] etc? That’s your first step.

But the other step is, there is that real world feedback. If I ask ChatGPT for a well structured Facebook ad that’s gonna drive, clicks, or signups or whatever, the ad either works or it doesn’t. I either get the clicks at a cost that makes sense or I don’t.

And so long term, you might ask the thing to give it to you in one way, but really, you validate what works or not through your real world experience, seeing what actually happens. And you either do that through doing the work and getting there, or you can leverage other people who have already validated that to say, I’ve seen what works and so I’m going to guide this AI system into what’s already working.

That general principle of being skeptical, You have to, if it’s information, you got to validate that it actually makes sense. If it’s, this is the right way to do it, you gotta see if it actually works in the real world. And do trust me. your experience on that.

Now, with that being said, here’s part of what’s interesting. When you partner [00:15:00] AI with a process that is proven, I know that this works. I know the right way to do it. I’ve seen it work. I just need to make aspects of this happen more automated. I need to make aspects of this happen at a consistent rate where somebody isn’t having to sit and type something out. Somebody isn’t having to, do this process that really is just running through an algorithm. That’s where AI can actually be really powerful, as long as you know how and where to deploy it. Where you could say, given this transcript, write me an email this way, write me an article this way, but I want you to leverage these templates and that sort of thing.

Like it could do that really well, and that’s really where the opportunity is that I think a lot of people are missing. What they’re hoping for is they’re hoping for, you know, Rosie the robot that’s gonna go and do your dishes and take care of Astro, and do all of that, and we don’t have that right now. And to be completely honest, I don’t know if it’s even on the horizon, based on what we’ve seen with how AI is developing, and what the real world results are, looking like.

Now, right. we’re in a little bit of an unprecedented unprecedented time that we [00:16:00] might, flip a switch. We might get over there and we might have that. I’m not saying, that’s not possible. But right now, that doesn’t seem to be the case. And there is decent evidence to say that we may not even be close, in terms of what we actually have going on.

But if you already know what a process is and you build out smart tools, smart systems to do it on autopilot, we have seen massive time savings and things like that. Turning, I record a 10, 15 minute video transcript into an SOP that now can be used over and over and over and over again. Taking videos like this and being able to chop them down into, smaller bits, to be able to leverage in a marketing system that we already know that works, and surface the right, the right clips at the right time. We’ve seen that work. We’ve seen that be effective.

What are some other ways that you’ve seen, what’s working right now, still be really effective?

Charlie Madison: you know, You mentioned blind spots. In one of my negotiations, it was like with a close friend and it was family stuff. And so like super heightened emotions, right? Like [00:17:00] so much stuff. And as I was going through it, what I found is, and I was using Claude for it. Claude would repeat back what I said, and I’d be like, or what I typed. I’m like, oh, that’s not actually what I meant.

Again, it took me a while, it probably took me about an hour, maybe it wasn’t that long, it felt that way, but and it actually helped me get to the core. I was actually wanting to say and do. And I’ll tell you what, the conversation could not have gone any better.

Like, This was one of those things where a year ago, I talked to my wife about having this conversation. And she said, don’t do that. I still want to be friends with these people.

Zach Hammer: Right,

Charlie Madison: I can say, since this conversation has actually brought us closer together. It was amazing because I was able to, one, get to the heart of what really was important to me. And I was able to share with it what I know about the other person, and like our history. And like, we’re like, we’re friends, so I know his sensitive points, you know, his trigger [00:18:00] points. And so it was able to encompass all of that into this conversation where I was like, from the rip. And this isn’t about sensitive point. And it really just gave me like, stuff that, it’s like my context window didn’t have the ability to have all of this in mind.

But When I saw it going back and forth, it was able to really encapsulate so much more information than I can. I think that’s, what it is. Can I encapsulate so much more information, you know, I said, use Chris Voss and Richard Bandler. Completely different people, and so it’s you know, a lot of times I’ll say, take this idea and teach it to me from the Farnham Street mental maps, or from first principles. Or I’ll say, you know, so it’s able to take completely different things and mash them together. So Those are the things that really stick out to me.

Zach Hammer: Yeah. It’s amazing how powerful it could [00:19:00] be. Sometimes, literally what we need is, we just need the time to process something. Right. And, And it is really hard for any of us to step out of like our own way of thinking. And some people are better at it than others, but I think even for those who are great at it, it’s hard when you’re in the midst of whatever you’re doing, however you think about things, however you’re used to doing it, to try and look at it from a different perspective.

It’s not impossible, but it takes work, it takes effort, it takes energy. And we don’t always have like that peak energy to do that. But with AI, like it never runs out of that energy. You could say, I want you to pretend you’re a robot pirate that always answers every other word with beep.

And what’s interesting is, like, that as a requirement, it may or may not be able to do well, but it’s going to do it better than you and I could, in terms of holding to that consistently. And being able to do that, lets you practice things. It lets you potentially role play out situations. It lets you you know, Sort of go down that route of seeing how this might play [00:20:00] out to envision something in a way that’s hard for you to imagine. You could test having conversations and then practice it when the stakes are low.

What would this look like if somebody was like this. And like these are things that we could do ourselves, but they take a lot of mental energy. you could ask yourself, how would I write this if I was this person that I know really well? You can put yourself in that flow and you can try, and you’ll learn things that way. But when you can literally just ask, I want you to rewrite this as if you were George Lucas, writing this story. Or I want you to rewrite this as if you’re Adam Sandler, describing the same story.

What would change? What would be different? Why would it be different? And you learn things in that process that you wouldn’t learn otherwise. Because it’s it’s hard for you to think that way, right? cause you’re not that person. That can be really useful and really helpful.

There’s this aspect of discovery and having the secondary brain that really what it’s doing is it’s allowing you to look at your own thoughts from different angles, and your own expertise from different angles, and that you wouldn’t typically be able to. So it gives you some ability to do that.

The other thing that I’ve seen is, it certain aspects of [00:21:00] what you already know work, it allows you to do at scale that you couldn’t do at scale before. right? Where you could say, I know how I want this thing written. I always want it written this way. And if I had this idea, here’s how I’d write it. Here’s how that same idea applies, to it in a different way. And so it lets you take a portion of yourself and predefine it to flow in a consistent way. But you still have to know what things need to be present, why is it written this way, what aspect matters that you could validate, did it actually do it right on the other end.

But yeah, when you have that, it allows you to do a lot of these things at scale that you couldn’t before. But it’s still, it’s own skillset, and it’s own way of developing out processes. And ultimately what we’re seeing right now is, people are hoping that AI is going to completely eliminate the need for people, or it’s going to completely eliminate a lot of the people that you need.

And people are looking at it as the real opportunity here is for me to reduce costs. And there is some opportunity there. There’s some opportunity where you can reduce costs. But typically, where I see the biggest opportunity is, I know that [00:22:00] this thing works. Now, let me bring consistency to it, bring scalability to it, that sort of idea where you’ve already validated something, you’ve done the hard work of testing an idea and you wanna scale it.

And then further in the same way, teaching your people to do the same thing. Right, Where they can bring consistency to their work that they might not normally be able to, by being able to say, this is how I always want to answer these emails. This is how I always want to do this. Now let me take that process and turn it into a system so that it could just happen on autopilot so I could free my mind up to focus on the things that actually matter. Not stepping out of thinking at all, but leveraging the things that I’ve already thought about and figured out and make them consistent. Rather than expecting the machine to think for me, it’s I did the creation part of the thinking, and I’m just getting the consistency and repetition to the machine.

Does that make sense?

Charlie Madison: Yeah.

Zach Hammer: But if you don’t do the thinking on the front end, whether it’s you, or somebody on your team, or anybody else, it’ll give you hot garbage.

Charlie Madison: Hot garbage. Yeah.

Zach Hammer: Or as [00:23:00] you call it, generic slop, right? it’s not going to be good. And you know, we keep looking at the hype and thinking that someday, it’s going to magically not become hot garbage, but we’re not seeing that. But if you do apply good, powerful, creative thinking around vetted ideas that you’ve practiced, it could be really powerful. that that’s really where the opportunity is. It’s around empowering your team with AI and empowering your processes with AI, not replacing people, but empowering them. That’s where we see a lot of the opportunity. It’s more in scale. It’s more in consistency, rather than people are looking at like, how can I cut a paycheck from my expenses.

And it’s more like, you got the paycheck, and you got somebody that’s working for you powerfully. How do you 10x what they’re capable of doing? That’s where the real opportunity is.

Charlie Madison: Yeah, because that, you know, in programming, one of the things is, you can get 10 one star developers or one 10 star developer. And the 10 star developer, of course, is so much nicer when you [00:24:00] think about just management overhead. And knowledge work, that’s the same thing across the board. Now, you can take the people that are great and make them greater. you know, But now, AI can be their assistant so that they spend less time on the stuff that the 80/20 rule. They can spend less time on the 80% that doesn’t produce. And they use AI to get the 20% better. AI in a lot of ways, you’re using words to create programs. And you know, right now, the programs are built to be variable.

And the skill is how to know whether it’s right on track or not and how to guide it. But imagine, if every one of your best team members was able to 10 X their productivity and you know, maybe even better the way it saves money. Like, What if you could let go of the people that aren’t performing?

Zach Hammer: And that to me, that’s where the real opportunity is, it’s not in getting rid of the people that you already enjoy working with. The people that you love working with that are great, make them so effective that the people that maybe aren’t all [00:25:00] that enjoyable to work with, that aren’t doing great work, that are you know, maybe doing the kind of work that is necessary, but it feels like you have to overpay for it. That sort of thing, there is opportunity.

So it’s like, this interesting thing where it’s, yes, there is opportunity to reduce costs, but it’s maybe not in the ways that people are thinking. And really a big part of this is, it is its own skill set. It is its own expertise. And you almost need people who are trying some of this, stuff out so that you could ask the questions, see what’s actually working and see how that looks.

Cause it’s not typically the same way that it’s working otherwise. Your processes for writing something with AI look different than maybe how you set up a human to write successfully. right? There’s similarities, but they’re also different. And so that’s really where the opportunity right now still lies.

So, you know, It’s this interesting thing. Is it overhyped? Yes, it absolutely is overhyped. But does that mean that there’s not still massive power in what it’s currently able to do? No, there’s still tons that you could do to gain massive leverage out of this technology, out of what’s available, but [00:26:00] you have to know what you need to do. And so that, you know, it’s still a skillset that needs to be learned in order to get that benefit out of it.

And that’s what we seek to do at the AMP Intel elite group, where we come together and we talk about what’s actually working for us. You get the opportunity to learn what we’ve already proven, what we’ve already vetted, but you also get the opportunity to get unblocked and unstuck in whatever you’re working on, see what the opportunities are, see what, processes could potentially be leveraged this way so that you’re able to you know, really bring together your team, you know, empower a team to leverage AI, both by learning it yourself, but also having your team learn it as well.

so yeah, if you’re interested in learning that information and learning how you can put that into practice and having the space to develop that skillset, we’d love to have you. And uh, you can reach out at ZachHammer.me/contact, and find out more about what that looks like. But otherwise, that’s what we have for you today.

Charlie, what’s your biggest takeaway or what’s your biggest thought on this one?

Charlie Madison: The name of your program is AMP Intel. Amplify your intelligence that you already have.

Don’t pay attention to the first releases. [00:27:00] It’s like I remember when the video game wars happened with the old Xbox and PlayStation. There’d be these beautiful animations and this is what the gameplay is going to look like and then you play it and it’s really cool gameplay but it wasn’t that. sss you know, And you know, think of it as amplifying your intelligence.

It’s like Iron Man with his armor, like it really can make you 10x productivity in the skills you have. So now, my suggestion, find someone that’s got the skill, and already knows some AI, or find someone that’s got the skill, and get them connected with Zach, so your whole team, then it comes that 10x amplified intelligence where, now you don’t have to have an army of people, but you can have a small, maybe so task force that all 10x times 10x plus 10x. That’s, That’s a big difference.

Zach Hammer: And that really is how you leverage AI, I think right now, to make the difference. Because it’s more about being able to do more with less, [00:28:00] but knowing where you can do that effectively versus where you can’t, and knowing that just giving somebody AI isn’t going to make them an expert.

In fact, it could cause people to deploy a lot of time and a lot of effort on things that will never work, because they aren’t the understand what’s good or what’s not. But having that those realistic processes and having that ability to leverage it to 10x what’s working, that’s where the real power is.

So absolutely, Charlie, thanks again for coming on. And everyone else, thank you for joining us for another episode of AMP Intel. And we’ll catch you on the next one.