AI for Service Businesses: Why Most Founders Are Using It Backwards (with Dr. Yeshwant Muthusamy)

speaker-0: Whoa.

speaker-1: Hello. ⁓

speaker-0: Hello No guest yet?

speaker-1: ⁓ yeah. His note taker's here, but not him.

speaker-0: Okay. Yeah. The note taker is always on time.

speaker-1: Yeah.

speaker-0: Crossover into a sales church that I offer training workshops and courses specifically designed for non-technical business professional. ⁓ Don't focus the hands on exercises. ⁓ huh. It's got a Yeshvik Yeshvik? That's how you say his name? No. Yeshwant

speaker-2: My name is Yeshwan. That's the name of my company is Yeshvik Solutions.

speaker-1: There he is.

speaker-0: Mut to Sami. Solutions. Okay. Hi, how are ya?

speaker-2: I I was trying to click on the meeting link, but because you had put it in parentheses, it kept taking me to maps instead of just the UI.

speaker-0: Yeah, did the same thing to me too.

speaker-1: Was that are you are you talking about in the ⁓ in the calendar invite?

speaker-0: Calendar. Yeah.

speaker-2: S be because the actual URL within the within brackets, it was treating it as a map location. So

speaker-1: Yeah.

speaker-2: Anyway, I I had to ⁓ open up the calendar invite to edit and then cut and paste it to get in.

speaker-1: Yeah.

speaker-0: Wow. Okay. Well good. Glad you're here.

speaker-1: Like why I think I'll see if I can figure that out.

speaker-0: Let's see. ⁓

speaker-2: Huh.

speaker-0: ⁓

speaker-2: Did you guys get my email?

speaker-1: Yeah, I was just gonna I was just gonna mention that. So I mean we're looking for that. We let you decide where you want to direct viewers in that. And if you had a chance to look at the the outline, the out the overview. Excuse me. Toward the end of the show, the plan what we usually do is ⁓ we have a CTA at that point.

speaker-0: T-da.

speaker-2: Yes, I did. Yeah.

speaker-1: And right right now, this is I'll I'll read it off what I have. If what Yesh, do you mind if I call you Yesh, is that okay? Okay. If what Yesh shared resonated and you think and you want to think more carefully about where AI actually fits in your business, head to yeshvik.com. He ⁓ or and then at that point we can we can talk about

speaker-2: Sure.

speaker-1: what it is you actually want them what action you want them to take. So

speaker-2: Okay.

speaker-1: co do you want me to add something in there about your workshops at that point?

speaker-2: Would be that would be great if you're able to if you're able to do that, because my my upcoming workshop is on May 29th. So that's the nearest in terms of any CTA that they might be interested in. But it is meant to be an in-person local workshop. So I don't know what the geographic reach of this podcast is.

speaker-1: ⁓ it's in person. Okay. Well, that's fine. ⁓ how about how about I just say something like ⁓ if what Yesh shared resonated with you and you want to think more carefully about where AI actually fits in your business, head to yeshbook.com. He's a person I would want in the room when these decisions are getting made. And he won and he works with companies on everything from custom AI solutions to training your team to expert advisement.

speaker-2: Right. So

speaker-1: And he's working on some workshops. Yes, you want to tell him about your upcoming workshop at that point. Yeah. And then we can do it do it that way. Okay. Yep. And then I'll include like we we can include like the the link that you provided previously to yesbook dot com and the link to the ⁓ workshop then, so that's okay.

speaker-2: That was That works.

speaker-1: Awesome. Excellent. Okay. Are you having a great day today so far?

speaker-0: Right.

speaker-2: So far.

speaker-1: Yeah.

speaker-2: Lots to do. And and we choose that.

speaker-1: Yeah. I don't I don't know if we actually address this in here, but ⁓ I know that ⁓ you're a bit of a gourmet, right? Yesh?

speaker-2: I have my own gourmet chocolate truffle business, which is a real business with an output store and a lot of satisfied customers. And we have so it it is

speaker-1: Chocolate truffles?

speaker-0: Yeah.

speaker-1: Those are not the same thing as like the mushrooms troubles.

speaker-2: The these are all made with fine Belgian dark chocolate. I usually fine Belgian dark courture in all of my truffles and most of them have a small amount of premium liqueur in them as well.

speaker-0: And should say.

speaker-1: That sounds amazing.

speaker-0: Wow. Sounds awesome.

speaker-2: Thank you.

speaker-1: Well, okay. Are you ready to get going here?

speaker-0: Ready to go? All right. We're going to be broadcasting this live on YouTube on my channel and it'll be recorded on on ⁓

speaker-1: I'm gonna we just got up on we're up on ⁓ Apple, sp ⁓ Spotify, Apple Music, or excuse me, ⁓ Amazon Music and like a bunch of littler ones after that. So

speaker-0: ⁓ wow, I forgot about that.

speaker-1: Yeah, okay.

speaker-2: So e is this actually is this actually gonna be live?

speaker-1: Yeah, we're also brought

speaker-0: Going in live on YouTube.

speaker-2: I wish I'd gotten the live link before I would have socialized the heck out of it on on LinkedIn.

speaker-1: ⁓ yeah, didn't didn't think about that. Forgot to mention it.

speaker-0: Well, I'll give it to you afterward. Alrighty. Okay, so human at the helm. ⁓ all right. Inf So you've probably heard about a hundred times that you need to be using AI in your business. Maybe you've tried. Maybe you opened Chat GPT, asked it to write a proposal, looked at the result, and thought, this is not how I would say it. So you closed the tab. Or maybe you went the other direction. You bought five tools, plugged them into your operations, and now your business runs faster. But it feels like It's not yours. Most of these are symptoms of the same problem. Most founders have been told to put AI in their business without ever being told what AI is actually for. So, today's guest has spent his career building this technology from the inside. And he is going to walk us through a very different way of thinking about it. Welcome today to the Growth Ceiling Podcast. Our guest is Dr. Yeshwant Mutusami. ⁓ we're gonna call him Yesh. And and ⁓ he is the owner of Yeshmix Solutions. So let's take it away, guys.

speaker-1: Yeah, welcome. Welcome. ⁓ so ⁓ Yesh Yeshvik Solutions. Yeah. Yeshvik Solutions is a boutique AI consulting firm in Allen, Texas, where he helps companies build customized AI solutions, trains teams on how to actually use this technology, and serves as an expert witness in AI related intellectual property cases, which is interesting. And before that, Yesh had spent years inside companies like Samsung Research America, Nokia Toyota, and Texas Instruments working on AI long before it was a headline. He holds seven patents and over 30 peer-reviewed publications. The reason ⁓ we wanted Yesh on the show is that ⁓ most of our founders, listeners out there, are most likely reading about AI. ⁓ And most of what they are reading is a b is a is written by people who have never built it from the inside out, right? So Yesh has. And ⁓ he has a perspective on how to use it that I think is gonna be ⁓ land pretty hard for anyone in our audience. By the way, it's Dr. Yeshwant, so just say no, yeah. ⁓ all right. So Yesh, you've been working ⁓ in AI, Dr. Yesh, ⁓ f ⁓ for for and for decades. Long before ⁓ any everyone else was talking about it. So when you look at the way founders and small business owners are approaching AI right now, what is the pattern that you keep seeing?

speaker-2: Sure. Thanks Nathan. ⁓ so here, I mean, the way I see it is I see a mix of of of reactions and and emotions and it kind of spans that gamut from people who have thriving small businesses who are downright scared about AI to people who have jumped in with both feet, but who have been kind of been taken in by all the hype. So it it it's a wide range that that I'm seeing, but the what I try to focus on in my coaching and my training and my consulting practice is to try and get the non technical business professional to look at AI the way I think it needs to be looked at, meaning that it is it represents at the end of the day a set of very powerful labour compression tools. AI will let you do things much faster. It will spark new ideas in you that you probably would not have thought of yourself. But at the end of the day, you still need to be accountable for its output. You cannot abdicate your responsibility and your accountability to a large language model, saying that AI told me to do it, or AI said it it was okay. Is not a valid excuse anywhere because at the at the end of the day, you are still responsible for your business. You are still responsible for taking the output of what AI gives you and then seeing whether it matches your voice and whether it matches the needs of your end customer. So that is what I try to try to impart to my clients and and my students.

speaker-0: So that makes so much sense.

speaker-1: We we talk a lot on this on this podcast about the three three main pillars visibility, viability, and value in a business, right? ⁓ you want your business to be more visible to the right people, you know, so you're attracting the people and which leads to better, you know, better clients, ⁓ better leads overall. But then there's also that viability layer where you want the business to function well, you know, which touches many different areas, obviously, right? Not the least of which is customer satisfaction, their experience that they have when working with you. And so like to your point, if you're leaving it to the AI and like bl you know leaving everything up to them, well the AI said it, so it must be good. Post it, you know, that's that's a recipe for disaster. ⁓ And the way I see it, like using AI well fits into that viability portion specifically because your ⁓ internal systems and decisions are are either gonna, you know, support it, you know, or they're gonna break under the usage, right? Right. And ⁓ so there's there's a phrase ⁓ that I kind of want to start with you say that humans need to be at the helm of AI, not just in the loop. Right. So what's the difference and why does it matter so much for a small business owner?

speaker-2: Well, because once again, I mean, you are you understand your business. You've put in the blood, sweat and tears to get that business to where it is. Now you want to use AI or AI tools to ⁓ to o optimize it, to s to to s to smooth out all the wrinkles. And that's a great thought. But here but you have to approach it in a very intentional way. You have to approach it without losing sight of the end goal. What is it that you're trying to accomplish with AI? What are the pain points that you want to address with AI? Things that you or your staff are doing on a daily basis that is time consuming, labor intensive, those might be the low-hanging fruit to have AI tools address that. That frees up time to ⁓ for you to do ⁓ other tasks more. More long-term, more strategic tasks. And there is also this people management aspect of it. I mean, in a small one person, two-person firm where both the four group or both the four founders are on the same page, it's not that big of a deal. But as you begin to grow, there is this widespread ⁓ fear that AI is out to take my job. And that is true only to a very small extent. But as I like to tell my clients and my students, AI is not going to take your job. But the guy or gal who knows how to use AI effectively is most certainly going to take your job. So you need to try, I mean, you need to make it a top priority to educate your team members that this is not about replacing you. This is about making you more productive in what you're doing. I mean, if you had a junior employee whose main task was writing Excel formulas and making sure that the Excel split sheet gave out proper charts. That can be now offloaded to an AI tool. That's gonna get it done much faster. But that doesn't mean that that ⁓ employee who might be a really smart kid can't be doing something else. It's all about upskilling them. It's all about showing them how to use AI and what you can do if you are going to offload a few tasks that that that AI can now do. And there is a lot of there there are a lot of tools that will help you do that.

speaker-1: Do you find that a lot of ⁓ business owners, founders are treating AI like a magic button?

speaker-2: Some of them are, but ⁓ but the smart ones or the ones who have been burned in the past are not. But but but I see some of them treating a a a AI like, ⁓ I have I mean, i it's kind of funny for a lot of people, AI is chat GPT and chat GPT is AI. And I've I've I've even

speaker-1: Phoenix.

speaker-2: I've even had people walk up to me and say, This AI thing is so cool. It started in November of twenty twenty two. I kind of cringe when I hear s statements like that because AI has been around, I mean, since the nineteen fifties. The first known paper on AI was written by Alan Turing in nineteen fifty. And the title of his paper was Can Machines Think? Yeah. You know considered the origin.

speaker-1: During

speaker-2: ⁓ of AI. And it's a great paper. If you ever want to get a feel for how brilliant that guy was, I I would I would highly recommend go go going through that paper paper because that shows how this whole thing started. And so I always start my workshops or my courses with a quick timeline and evolution of AI from that paper to where we are now. Not because I want to bore people with a history lesson, but because I want to show people in just a few slides or or or maybe even one slide to show the progression of thinking, how we have come from that paper to now we have things like Claude and Chat GPT five point four and Grok and so on and and and and and so forth. It's been a fascinating journey to see the progression of of of of human thinking.

speaker-1: Yeah. I I could imagine that would be that could be its own series of episodes. Yeah. ⁓

speaker-0: Now that ⁓ sorry ⁓ can I ask this question real quick? ⁓ you've you've been in the AI space for for a lot of years. Now that AI is is accessible to the general public, do you feel like it's moving faster because it's accessible to everybody and now there's all this innovation around it? Or is it moving at the pace that it was moving? when it wasn't accessible to to us.

speaker-2: Yeah, no, I I I think the pace has definitely quickened because there is a lot more funding. I mean, so it it's all a cycle, right? Because more people come to know of it. And everybody, I mean, if you're a VC, you're not gonna look at a company if it doesn't have the word A AI in the pitch deck, right? I mean, it's I mean, we are in that age. But at the same time, ⁓ I I know of certain VCs who have become very smart and they are not, they don't look at companies where AI is just a wrapper because you want to be funding companies that are AI native. I mean, if you if all you're doing is ⁓ adding a AI wrapper to an existing product or a tool, then the next revision of Chat GPT is going to render your company useless because they're going to probably incorporate that functionality. And so and so you have to, I mean the pace has definitely increased. And there's a lot more innovation that's happening because a lot more money and resources is being thrown at the problem. But come but what comes along with it is a lot of hype, some of which is very dangerous. And I mean a l a lot of claims being made on LinkedIn and other social media media that haven't quite been fully vetted. And so, I mean, I I'm sure we've all have heard of ⁓ this company that laid off all almost all seven hundred of their ⁓ customer service A ⁓ agents and it may and it made a big new news flash only for them to quietly rehire back most of them about a year later when they realize that they cannot make do with just A AI tools. So you have to

speaker-1: Yeah. So

speaker-2: Yeah you have to take a step back. I mean a a a AI is not like a new IT tool. It actually is a transformational technology. So it's gonna change how your company thinks and works at a very fundamental way. And if you're the C CEO of that firm, you have to think about those things before you kind of go in and say that, ⁓ I'm gonna make my company completely A AI. ⁓ first or or or whatever the term you want to use. It has to be done in a very intentional way.

speaker-1: Yeah. So in that case, you're it you know, potentially looking at AI as the end all be all, you run the risk of losing your employees' loyalty, you know, if you go and fire most of your staff. I mean how how is how are the the remaining employees gonna view that for one thing, let alone the ones you've already fired. And then and then you know, ⁓ I I Perhaps there's market pushback for companies that are doing this, even, you know, which could end up costing the company in the end, in that case, right? So there's multiple landmines here potentially for for companies that are getting into this and and going about it the wrong way, in other words. ⁓ so on the flip side, however, you have business owners who are won't touch it with a 10 foot pole, right? They they won't go anywhere near AI. ⁓ what what are they missing out on? What are they losing out on?

speaker-2: They are losing out on a lot of efficiencies and they are losing out on the market perception of them. I mean, I actually was speaking to a guy who runs a marketing agency. I mean, he's a somewhat of an ⁓ somewhat of an older gentleman. He's in the I think mid to late sixties, and his take was I'm old. I have something that runs well. Why am I going to mess with a new tech technology? But I'm thinking to myself, that is that is exactly the wrong thing you want to be say say saying at this point, because it is okay to have a f fully functional business that is not yet using AI, but you need to have an open mind to see where you can address the human inefficiencies which you have probably taken for granted, given that you've been having this bid business. we which has been in existence for all these years. And you might find that you are giving some employee time back. Or you might have your employees do other things which you want to do, which are on your rural roadmap, but they can't get to because they are too busy do doing these tasks. So it just requires people to have a more open mind. And I would approach it not from the view of you need to use AI because this tool I mean it's Claude today and it might be FUBAR to next week, or it could be ⁓ XYZ the the the the the week after. But rather approach it from what can AI do to me to for me to increase my ROI? And ⁓ in that regard, the a a shameless plug for the tagline of my consulting firm, we maximize your ROI from AI. I mean, that's is literally the tagline for my practice. And that I believe is really how you need to approach it. What can AI do for me to maximize the ROI? I have sunk all this money into this business. If I use AI, am I going to be able to get money out of my business faster, more efficiently? And with a smaller number of high value employees.

speaker-1: Yeah. ⁓

speaker-2: than having people do things like date data entry and th and things like that.

speaker-1: Yeah. And again, though, doing it well is the key here because otherwise there's you know, there's there's risks on both sides. You run the risk of of ⁓ trusting it completely ⁓ and blindly and not trusting it at all. So you know the happy medium is the is is the way to go, right? ⁓ w okay, so when you walk into a small business when you start working with a client, yes, what are there common operational messes that you encounter along the way.

speaker-2: ⁓ there is a lot of

speaker-1: Tried they tried to in implement it. Sure.

speaker-2: Yeah, yeah. I mean, people work with a lot of legacy data and they take the legacy data that they have, which is in different forms for granted. And we as human beings are very good at di discerning or being able to process these different d data that comes through in different modalities, like from emails, from chat messages, from p PDF documents, and so on and so forth. And so they have this perception of, ⁓ I've heard that a ⁓ AI can extract, help me extract more value. So I have data, there is AI here, and the next box has to be the that the value which I extract. What they tend to miss out, that is just the tip of the iceberg. That's what you see on the surface. Under the data is a whole thing about normalizing the data, putting the data in the form. That the that the AI ⁓ model can understand. You and I can are able to synthesize information from a Slack message and email and a text and a PDF document into a coherent whole. An AI model will not necessarily be able to do that unless it is given proper training or guidance. And so so and and then the the there is a big issue of. Bad data. I mean the old the old adage which was first coined back in the sixties still holds garbage in, garbage out.

speaker-1: You you go.

speaker-2: have the best AI model in the world, the most expensive mob model, but if you feed it garbage, it's gonna give you garbage as output. ⁓ And then there's a whole thing about the model. If you are a company that's training an AI model, you have to be very careful about the inherent biases in the data that you feed your model. Because otherwise you're going to end up having an AI model that displays signs of bias or discrimination against certain ethnic groups.

speaker-1: ⁓ man, this is a tangent we could go down 'cause

speaker-2: Right.

speaker-0: We and we've done it before. We've done it before.

speaker-2: So I mean a AI is a capital intensive investment at the end of the day. You have to make sure that you are that that you have crossed all the T's and dotted all your I's before the output of the model gets to your end customers because then the cost of bad PR is way more expensive than spending a few extra days. Trying to clean up your day d data.

speaker-1: Right.

speaker-2: That beard is can be monumentally expensive to fix. ⁓

speaker-1: Yeah. ⁓

speaker-0: I heard that there's also ⁓ there's also a problem too with ⁓ feeding it too much data. I was ⁓ I'm working working with a friend of mine and he was telling me about how he had you know he's just getting off the ground starting his business and I think he was looking for advice from several different like marketing experts, you know, and so he fed it, he put together a Word document. ⁓ that had ⁓ advice from Myron Golden and and Alex Mormozy and Russell Brunson and all these different people. You know, and and I said to him, and he was like, you know, okay, so I I put in all this stuff and, you know, and ask it to give me advice ⁓ for for whatever it is I want to do. Like, you know, how to put together a webinar or how to do how to put together this or that or the other thing. And I'm thinking, and I told him, I said, yeah, but ⁓ Myron Golden and Alex Hormozy have different philosophies on certain things, you know, and Russell Brunson may have a completely different philosophy than the other two on certain things. So I was thinking, okay, how does the AI not get confused when you're giving it so much and then asking for a concise some, you know, concise advice out of it?

speaker-2: R right. I mean it depends on the AI mo model too, right? I mean, because there are some newer models which are actually quite good. I mean, Cloud Sonnet four point six or Cloud O Opus four point seven, not trying to plug any sp specific large language mob mob model here, but those I mean Cloud has been getting a lot of press recently, ⁓ but but but but but I also use ⁓ chat GPT five point four and all those things. I mean the k the best thing you can do is try to play one chatbot against the other, give the same prompt to both of them and see what the output is. But it also requires you to have some amount of ⁓ in introspection and critical thinking yourself to see which one makes more sense. And I and always insist on providing on the ⁓ on the on the chat bot providing you verifiable links. Yeah. Always. I mean that is always part of my prompt. Provide verifiable links for all your assertions. Do not make things up. And so because I have reams and reams of examples from Chat GPT where it has just blitzly made things up. But when I go back in and say, I don't think what you're saying is true. Can you Can you please recheck that it immediately backs off and says, I'm sorry, I made a mistake.

speaker-0: Yeah. I found that too. It'll it'll you know, I'm asking it to, you know, write write something up for some marketing project that I'm doing. And it will put in testimonials, even though I fed it my stories, my case studies, my testimonials, and it's still making up stuff. And I said, Well, where did you get this? 'Cause I don't remember this being in my list of stories. ⁓ yeah, I'm sorry. ⁓ it I I made it up because, you know, I thought it was a more perfect story for this scenario than this other thing that you had.

speaker-2: Yes. I mean so you have to be I mean it it it is quite literally like you hiring a young, bright eyed, bushy-tailed young assistant to help you with your work. Are you gonna throw a piece of paper at him on at him or her on the first day on their job and say, Now go do this and then you go off and do something else? It's unlikely. You're probably gonna sit with them for their entire first day. You're at least gonna go And check in on them once an hour to see if they have gotten the if they feel for the ropes yet, if they are having any issues. Or so you're gonna try to help them to succeed because it's their first day on the job. Using an AI chatbot is literally like that. You have to have a dialogue with them. You need to have them work for you and not you working for them. So

speaker-1: ⁓ I I just read recently, Mark Cuban said that the the biggest challenge ⁓ as far as like using AI in business is that the AI tools will often provide different answers to the same question. Yep. And so the consistency is a big issue. ⁓ and but you know, there are some mitigating things that you can do there, with especially what if it comes to

speaker-2: Yes.

speaker-1: Holding it by the hand and training it, you know, and being very literal and specific about what you're looking for. Yeah. Yeah.

speaker-2: And and and and and that that and that's the reason I all I kind of refer to these large language models as nothing more than nothing more than probabilistic parents. All they are doing is picking the most probable next word in the sequence based on their training data. That is all they are doing. They do not have a conscience, they are not sentient, they don't know you from Adam or Eve.

speaker-1: Yeah. Yeah.

speaker-2: they they are they are basically a they are basically a probabilistic parrots. So they're just picking the most most probable next word in the sequence and that's how they build their output. So to impart some kind of anthropomorphic re ⁓ c character to them is not quite right. Now there is a lot of there there are l lot of models that employ reasoning engines During inference time. That is not part of the that is not part of the training data of the large language model. That all that reasoning is something that's put on top of the output to make sure that it makes sense. So it helps to understand what it is that you're actually dealing with.

speaker-0: Well does this mean we don't have to say please and thank you to our bots?

speaker-1: You

speaker-2: Will be I mean that th that's an interesting question. That makes you feel good, but what you're doing is you are actually expending more tokens. Because the bottom water

speaker-1: More water.

speaker-2: And now a lot of the c companies are moving to a usage based model. And so and and if they start charging you at API rates, then that can be quite expensive.

speaker-0: Right. Because the I noticed like the the AI, they respond to everything that you you tell it. Like it's it feels like it always has to has a have the last word. So if I upload ⁓ I have an AI bot like ⁓ responding to my text messages, if somebody sends a smiley face, it'll send a smiley face back and and another message. Thank you so much. You know, please feel free to reach out again if you have another and I'm thinking, just leave it at the smiley face. Like, why are you you it's so verbose sometimes. But yeah, every message that you that you ⁓ give it, it's it's gonna respond. And so yeah, i if you're paying by the message. Yeah, it gets expensive if you're if it's you're doing all the niceties, all the pleas and thank yous and the And all the rest of it. Although I do hear people say that they do that because well, first of all, they're afraid of the bots taking over later. And so and if the bots are are taking over, then they want the bot to remember that they were kind.

speaker-2: Unfortunately, the b like I said, the bot doesn't know you from Adam or Eve because none of these large language models currently have what is known as adaptive training. What most people attribute to these bots is the notion that the more you use it, the more it learns about you. But what is really happening is someone, either you or someone else, has turned on the ⁓

speaker-1: Yeah.

speaker-2: memory context flag in the bot setting. So what that does is it remembers your past conversation.

speaker-0: It's pattern recognition. Right.

speaker-2: It is using that as context. So for example, if you went in and said, build a I mean generate a nice birthday card for my niece, for my niece named Joanne, it now knows that Joanne is the name of your niece. And if you happen to give the age of Joanne, it knows that too. And if you turned on the memory context flag on, A few months later you go in and ask something about Joanne, it remembers based on your past conversation that Joanne was your niece who's eleven years old. That is it. It hasn't it hasn't adaptively learned about you because it doesn't have that capability yet.

speaker-1: It's still just accessing the ⁓ what do you call it? The the context. Right. The the the ⁓ the memory module is just its own context that it creates, yeah.

speaker-2: A and now, I mean, and if for some reason one day you just go and turn off that flag, that's it. It loses all the

speaker-1: Yeah. I I heard ⁓ speaking of ⁓ disasters ⁓ here w with ⁓ using AI, I heard recently about someone who ⁓ had a a ⁓ re a repeat I can't remember what it was a loop set up ⁓ with claw cowork. And you know, talking about the API charges, it was charging him, it would send through the prompt. And then have a 30-minute delay and then send the next prompt, which was a compilation of both of those, and it would just repeat. So what started at like you know, 50 tokens ended up at 800,000 after 26 hours. And it was like you lost six grand on this. So it's yeah, yeah. Yeah. Yesh, have you ever encountered, you know, and you can anon anonymize the story, but have you ever encountered where a company was thinking that they were using it productively? But actually there was a problem that was going on. Is that

speaker-2: Well, I mean, I I was recently talking with the CEO of a of a of a construction firm. I will not say what aspect of c construction, but he had deep enough pockets to give f f ⁓ the max cloud max seats to quite a few of his d the developers to build agents and he's got a lot of agents in operation for his business.

speaker-1: Yeah.

speaker-2: But he wasn't able to answer very simple questions that I was asking in during the discovery call, that if he knew which version of Claude all those agents that were making use of. He had no idea. And I said that you don't want to be taking, you don't want to be using a sledgehammer to swat a fly. For example, Claude Opus four point seven.

speaker-1: Yeah. Mm.

speaker-2: is the most expensive in terms of tokens.

speaker-1: Yep. That's what the guy that I the on the loop he was using Claude Opus. That was part of what I was calling.

speaker-2: And and for people in the know, it's well established that you want to use Claude Opus as a strategic advisor if you're building agents, because for a lot of simpler tasks, Claude Haiku four point five is more than enough. If you want to go up one step above in terms of reasoning, go to Claude Sonnet four point six. Haiku is five times cheaper than Claude Opus five point four four point seven. Claude Sonnet is 40% cheaper than Opus 4.7. It makes you feel good that you're using the latest and greatest, but it's like buying a Ferrari or a or a Lamborghini and take it grocery shopping, which is about a mile away from where you live. It makes you look good. But two things, you probably can't fit a bag of groceries in the trunk of that.

speaker-1: Yeah. Right, yeah. Yeah. All right.

speaker-2: ⁓ of the cannon and you don't need a lambo to go if all ho if all you're doing is going to grocery shopping and the grocery store is ⁓ about one mile away. So it's like that. You do not need to use the biggest and the greatest and the latest just because it's there.

speaker-1: Right, yeah. Awesome. All right. So let's get into kind of like your approach on how you're actually thinking about these things. I know we're kinda we've already started going down that path, but when a business owner comes to you and says, I know I should be doing more with AI, but I don't know where to start, what is your approach in that case? I

speaker-2: I mean, I always offer a free 30 minute discovery call with all potential clients, and the per and the purpose of the call is for me to try and understand what it is that they are dealing with in terms of their data, what is the problem that he or she is trying to solve, and what is their pain point. I mean, I asked them to describe to me three or four tasks or the top N tasks, whatever end they feel. comfortable with, that is the most time consuming, that that that that actually represents their pain point. And that gives me valuable insights to where I can bring in real value. And I'm I'm a big fan of doing things step by step to try and pluck the low-hanging fruit first and to show immediate immediate ROI. And then we can go go go go go go on to the next task. So for me, as part of my AI c consultancy, I'm not pushing a specific platform or a product. I want to know what is it that you're tr trying to do and how can AI help you? There might be k cases, and thankfully for me, there haven't been too many of those. That your problem is not an AI problem at all. It might be a data wrangling problem. And AI might help, but you don't have to use AI. But most of the time, if you are a if you are a firm that's dealing with data and these days, who isn't, right? You get so much of data from several different sources sources, there is usually a way to use AI to make your life a b ⁓ a lot easier. So that that's what I tr try to focus on. And so you you end up with a solution that is custom made.

speaker-1: ⁓ yeah.

speaker-2: for that client, for their business vertical, for their data. So

speaker-1: So you you're d you describe AI as a labor compression tool. So what can you walk us through what that actually means? What what kind of labor compresses well and what does not?

speaker-2: Right. ⁓ tasks which are which are which are repetitive, tasks which require you to go and get data from multiple sources, that that can be very time c and consuming. Just to give you a quick example, I get an AI daily news digest every day in my folder by seven AM thanks to Claude Kobe. And that is actually collating news in various aspects of AI, like AI research, AI AI AI policy, any news about the latest ⁓ business developments pertaining to AI, any stock market, any stock market movement. pertaining to AI. So I have a about of five or six A areas that that I want that that report to basically break down the things for me. But the here is a fun part. It is doing all this work while I'm sleeping. Because that report is done and in my folder by seven AM every morning. And I make sure

speaker-1: Never gets sick, never takes vacation. Yeah.

speaker-2: You know, it it it it it it never catches a cold and I ask it and I ask it to give me verifiable links for it. So I can go and see for myself if I find something really mind blowing. Is it really saying what I t think it's saying? But there is a link I can go and I can go and see for myself if that is really or if this is making stuff up. Now that that is just one example. I mean, can now in theory, I could go and get that information myself, but it would probably require me to make eight or nine or ten separate searches. The advantage of me having Claude Cowork unleash all these agents in parallel on my behalf is that I can get a I can get a I can get a good night's sleep if my objective is to have everything done by 7 a.m. in the morning.

speaker-1: Yeah.

speaker-2: So that's an example of labor impression. And the other one which I gave already is that if you're trying to wrangle a spreadsheet that has one million rows and hundred columns and get a bunch of p pivot charts and graphs, you could do it with Excel pop pop formulas. I mean, I have I I know how to run piv pivot charts and pivot tables. But it's a it's a real pain in the red to do. Right. Have AI get those things done for you. Because they have gotten really good about it. But you need someone to make sense of those pivot tables and those pivot pivot charts. You need someone who's smart enough to extract actionable insights from those graphs. So it doesn't take the human out of the loop.

speaker-1: Yeah.

speaker-2: You are actually, yeah, and the and the and the human is not just in the loop. The human gets to decide what to do with the data. And that's why I keep saying human at the helm. Because that is really where you need to be. You need you need to be the ultimate orchestrator of all of your AR agents. I mean, if you guys, I mean, as you guys are probably know, with the with with the whole notion of agent TKI, there is usually an orchestration a age a an agent, an agent that spawns multiple multiple sub agents as and when hi is is needed. But you need to be the ultimate orchestrator of all of the agents. Because you get to di decide if that information that that you got from your brand new, shiny new agent TKI s system is worth going and presenting in front of the board or in front front of your VC or or whatever.

speaker-1: So okay. How how do you go about helping b ⁓ founders, business owners who figure out which decisions and tasks belong to AI? Is it just rep anything repetitive, systematic belongs to AI, generally speaking? Or is there kind of like a heuristic that you take them through or

speaker-2: I mean, ⁓ anything that pertains to the long term vision of the business, unless you have done a brain dump of everything that you want the business to be doing 10 years from now, five years from now, that is something that exists with you. And for those kinds of things, you can ask for in input, but you have to make sure that you provide proper safeguards and you provide the proper c proper. context, but you still want to be the final guy. I and if you ask it for five options and it gives you five options and none of them make sense to you because they don't conform with your long term vision of where you want to take the business, feel free to ignore all five options. It's just a growth and I mean you want to go with your gut. I mean o oftentimes

speaker-1: Yeah, yeah, yeah. It's okay not to listen to the AI. ⁓

speaker-0: Okay.

speaker-2: Going with your gut makes the most sense because it is who you are. It is what you want to do.

speaker-1: Okay. Okay, so if you want AI to to actually, you know, stick and and work in the business, what has to be true about the business strategy, the the sales process, operations? Are are are you concerned so much about all those different components? ⁓ I would imagine that that would actually come up when it comes to trying to figure out where AI can actually help a business, but

speaker-2: Yeah. I mean you want to I mean it also comes down to getting buy-in from all the people who are part of your b business, right? I mean, so if you have just four or five employees or if you have 100 employees, it's the same issue. It just the scale has become much larger. You want buy-in from everybody else and you want them to be educated so everybody is on the same level at some basic level. of what AAI is and what it can do for you. I mean, I cringe when I hear stories about, ⁓ yeah, we are a small c company and each one of my employees has their own personal subscription to Chat GPT or that is a disaster waiting to happen.

speaker-1: Yeah. Yeah. A lot of proprietary information getting leaked there potentially.

speaker-2: Business data like you wouldn't believe ⁓ because you have no control over what they are p putting in, and you think you're saving five times 20 bucks a month by having them have their own personal subscription. But what you're losing is visibility and losing control over your business data. It is not about you breathing down the neck of your employees, it is about You making sure that they are not leaking your valuable IP. No not because they are bad people, but because they didn't know any better.

speaker-1: Mm-hmm. You heard it here, folks, making sure that you have your AI subscription shared and it's like business level amongst your employees so you maintain invisibility so you can see what's going on. Right. And what what is being shared.

speaker-2: That is I I think is so important because if you do not know what your employees are doing or what they are feeding it to ⁓ to AI and

speaker-0: That's a thing too. Like they all need to be working from the same from the same song sheet, right? Like everybody's you know. ⁓ Yeah. So all of the what's coming out of it is is consistent. They're not all feeding it different things and ⁓

speaker-1: Right. Back to the consistency thing, right?

speaker-2: And and not not many people know this, but Chat GPT, the free version, Chat GPT plus for which you pay t twenty bucks a month, even Chat GPT Pro for which you can pay up to two hundred bucks a month, none of them have any data security.

speaker-1: Fine.

speaker-2: And for all of them, you can put in a you can you can turn a switch off saying that please don't train on my data. But that is not legally defensible in court as far as we know. You have to go to Chat GPT team or chat GPT business to get SOC2 type two data security, and where open AI gives you a legally binding agreement saying that they will not touch. your data for training.

speaker-1: Well.

speaker-2: That is what you want. You want either Chat GPT team or Chat GPT business. I mean, most most small businesses won't cock, won't qualify for Chat GPT and enterprise because that needs ⁓ ho ho a a I ch I think at least a minimum of ⁓ sixty seats or but at least look at team or bid b or b business.

speaker-1: Yeah. Mm.

speaker-2: And the cost for those are not that much higher than the chat GPT. And so you also want role-based access. Only certain people who are authorized and high enough in the food chain need to be accessing an external chatbot, which is hooked up to external servers, where once you give it your company data as part of a prompt, there is no way of getting it back.

speaker-1: Right? Mm. Hmm. Yeah.

speaker-2: So the whole thing about role based access, you don't want the junior intern who's there just for the summer to have the same level of access to Chat GPT as you have the CFO or the CMO.

speaker-1: Hmm. So we need to be strategic about how we're deploying our chat bots, ⁓ using AI inside the business, making sure that we're controlling who has access to it. you know, for visibility sake and security and proprietary information sake, ⁓ and all that. And for consistency sake, so that the messaging that's coming out, you know, is unified and, you know. We're not ⁓ saying ten different things, you know. Right. So yeah, all very important things to consider. ⁓ all right. So let's get into like the the multiplier effect ⁓ segment here. When ⁓ we're talking about like when we're doing this well, what the what does it look like? So when a business gets their imp AI implementation right, they have humans at the helm, you know, ⁓ Th we have all the ⁓ data security ⁓ you know th considerations ta you know in in in place and implemented. What changes about the business that they may me may not have expected?

speaker-2: Th they they will find that things are operating smoothly. Their employees feel feel seen and heard because they are now doing real strategic work as opposed to being just ⁓ doing b busy work and they and everybody feels that they're making a contribution, things that can be done ⁓ that that You used to take hours or days, now get done in a few minutes. But that doesn't mean that you just take it out of the output and out it goes. You have somebody looking over it. And so there is an aspect of training. The the external perception of your company as an AI forward and enterprise is going to cause more customers to flock to your company because. They now see that ⁓ these guys are doing something right.

speaker-1: Is is that a is sorry to interject, is that a polarizing aspect, you know?

speaker-2: It it d depends. I mean on I mean it's ⁓ if you're in a service business, then you might I mean it might it might actually be a good thing be because you're providing you're able to provide service to your end c customers where you very where you are able to come across as having your finger on their pulse because you have used AI to do sentiment analysis on their so on their social media posts and you know what are the trigger points that cause your end customers to get upset and you have addressed them because you have you are constantly monitoring your the your Facebook and Twitter feeds about your company. And you don't have to have a young kid there staring at the screen blary eyed. You you can have two that actually go and look at the posts and say, okay, this guy is really upset. Why? Because of the words that he used. or because of the word combinations that he used. He also is the same guy who made a post last week. So he is an unhappy customer. Give him a call. Ask someone from customer service, give him a call and see what is his issue. How can we fix it?

speaker-1: Mm-hmm. Yeah. Could head off a disaster later on, yeah.

speaker-2: Yeah. I mean, ⁓ it's great to be on social media, but y you have to make social media work for you. And there are so many tools that you can have social media work for you to make sure that you are actually staying on top of these things. Because as they say, one bad c one d disgruntled customer is all you need for yes for

speaker-1: Yeah. That could cost you a lot of money.

speaker-2: Things to go south. So

speaker-1: Yeah. So you you've seen this from both sides. ⁓ yes, y the technology side and the business side. So what do you know, a founder who implements this in their business, what does their role start to look like a year after they've done so? And it's done properly.

speaker-2: Well, he's still the ch chief executive. He's he or she the buck still s stops with them. Nothing changes from that perspective. But he needs to he or she needs to surround themselves with people who are able to advise them on what is the next big thing. Because with AI, I mean each day I open up my LinkedIn feed, if I'm not careful, I already feel like I'm one week behind. Then you kind of take a step back and then you kind of sift through the hype and then you realize it's not so bad after all. A lot of the posts on LinkedIn, it helps to understand, are generated mainly for clicks, mainly for clicks. ⁓ I built the I I had Claude Claude Cowork clean up my entire email box. I had Claude Co Cowork re reorganize my downloads folder.

speaker-1: Right.

speaker-2: What they never tell you in that post is you have now given unfettered access to your mailbox to a to a desktop app that is still to anthropic servers. You don't know what data is being shared because you're not sitting there packet sniffing. So you don't know what data is going out of your PC to anthropic servers.

speaker-1: Yeah. Yeah. Yep. Good point.

speaker-2: It it is like inviting a stranger off the street and giving him your safe combination or sh sh or sh showing him where you store your expensive jewelry. Jewelry and your

speaker-1: Can you help me organize this?

speaker-2: Why would you do that?

speaker-1: Yeah, yeah.

speaker-2: So that that that's what I am plan I'm going to teach in my upcoming workshop how to extract maximum value from something like Cloud Cowork, which is a no code agentic platform from Cloud. You want to get in on the age agencai wagon, but you don't want to go programming real agents you using cloud code or or c or you know or cursor or whatever it is. Then Cloud Co Claud Cloud Covar is your bag. But you need to know how to use it without compromising your data privacy. And a lot of people have heard about agentic AI, at least in the small to medium business circle, but they don't know what Cloud Cov is. And for me, that was a big surprise. I'm like, you don't know what Cloud Covar is? But he just told me that you knew.

speaker-1: ⁓ yeah. It's the rage.

speaker-2: You've heard of you heard of you've heard of H and TKI.

speaker-1: Yeah. Yeah. That whole terminology, what is an agent? You know, like we could that could be its own episode, I think, too, but

speaker-2: ⁓ so yeah.

speaker-1: We'll we'll include a a link at ⁓ in the show notes and ⁓ about Yesh's upcoming ⁓ workshop. I I want to ⁓ make sure we're mindful of the time here. ⁓ so r real quick, ⁓ if a business is implementing this properly again, and what what ch what changes about the business value would you say as far as the AI is concerned?

speaker-2: I think the the investors in the b business are happy, the end customer is happy, and the CEO can actually go and brag about how well his b b b b b business is doing. ⁓ I I I think the value, it might take a few months for you to see the value, but the upfront work that you put in to make sure that AI is integrated in a proper way into your workflow and not just as a bolt-on. I think you're go you're going to see real value. It might take maybe a few months longer than if you just kind of went and spent money on the latest AR two. And it might be worth a few posts on on LinkedIn. But outside of that, once ⁓ all your LinkedIn engage engagement kind of pe peters away, you still have a real b b business you wanna run. You still have a real value that you wanna extract out of all that mum. money and resources that you put in. So

speaker-1: And going back to your tagline, right? We help you maximize your ⁓ ROI with AI, right? Did I get that right? Yeah, okay. Yeah, I mean, that's it, right? We we want that's the whole that's the whole thing right there. We want to make sure that we're getting something out of it. So ⁓ all right, let's do a a rapid clarity around here. So, what's one question that every business owner, every founder should be asking about their AI? in their business this quarter.

speaker-2: Are there any repetitive drone type of tasks which people in my business are doing that I can offload to AI? And if there are, then let's talk, because that is where you need to start. How can I maximize my efficiency? And to at the at the risk of sounding like a broken record, AI at the end of the day is a set of very powerful labor compression tools. If you at least start off making use of it in that way, I think you'll be fine.

speaker-1: Yeah. Yeah. Okay. What is the single first step for a founder who has never seriously implemented AI?

speaker-2: Get educated. And I don't mean to sound self-serving here at all. Although if he wants to show up for one of my workshops, that's great. But my point is this: try to understand how AI can help your business at a very fundamental le le level. And you might reach the conclusion that AI is not going to do too much for my business. If I have a stone masonry b business, AI's use in my stone masonry b b b business is not as high as someone who is who's going to be who has a who has a digital ma marketing agency. They're going to use a lot more of AI than me if all I have is a stone min masonry business. I can use AI to get really nice mock-ups of how the particular project is go is going to look after we are done with it. It'll help for my end customer to see what we can do for for for him on that project, but I need to still have skilled workers who actually go and do the actual pri ⁓ actual project. No AI in the world is going to help them there.

speaker-1: Yeah. Yeah. Not yet anyway. Yeah. ⁓ all right. So what's ⁓ what's one thing AI should never be allowed to do in a service business?

speaker-2: Interact with your end customers without proper supervision.

speaker-1: Yeah. Yeah. Send email. Okay.

speaker-2: On proper.

speaker-0: So all these voice agents out here that people are deploying, they have to have ⁓ supervision before you ⁓ before you deploy it. Better great training and all of that. Right.

speaker-2: I mean, it's perfectly fine to have a voice agent answer the call for you while while you're sleeping or you happen to be on the beach, but make sure that it has been properly there are proper guardrails around what it can say and what it should not say and have a graceful handoff to a real agent. Set perfect.

speaker-1: And then last one, one what's one signal that tells you that a founder has crossed the line from human at the helm to the human is checked out.

speaker-2: ⁓ boy, very much that well one one obvious signal would be that he if his company makes the news for all the wrong reasons.

speaker-1: Okay.

speaker-2: But the ⁓ other one might be that people at his own company don't know what's going on or that they don't know that their AI has managed to piss off a whole bunch of customers because they said the wrong thing at the wrong time. So

speaker-1: Yeah. All right. Perfect. So if there's one thing I want the listeners to take away from this conversation, it's that AI itself is not a strategy. Okay. it's a tool that, as as Dr. Yesh said, it compresses labor, makes it simplifies things, right? And frees up your headspace. But like any powerful tool, you know, it amplifies whatever's already happening in the business. Yeah. I think one analogy I like to use is ⁓ you know, a a hammer is also a powerful tool, right, potentially. But if you, you know, use it the wrong way you can cause a lot of damage, right? So

speaker-2: It it's it's it it's interesting the that the that you mentioned hammer. The one the one analogy that that I use is it's like giving a chainsaw to a to a twelve year old. If you train that if you train that twelve year old how to use the chainsaw, he's gonna do a fantastic job trimming all your backyard head hedges while you're wh wh while wh while you're enjoying a tall glass of iced tea.

speaker-1: He's

speaker-2: It's gonna probably lose a limb or two and hurt the

speaker-1: Yeah, i and you'd probably be lucky if that's all that happened. ⁓ yeah. But if if your if your operations are already running well, then AI amplifies that, makes it run faster, potentially, even more efficient, right? But if they're not, obviously the opposite is true. ⁓ and so the the business owners, the founders who are implementing AI

speaker-0: Flames.

speaker-1: And who are winning with this technology are going to be the ones that are at the helm, as you say, and not the ones hoping it's going to steer for them. If what you if what Yesha said here and shared has resonated with you and you want to think more carefully about where AI actually fits in your business, head to yeshvik.com. ⁓ he's the person I'd want in the room when these decisions are getting made. And he works with companies on everything from custom AI solutions to training your team to expert advisement. And he as we mentioned earlier, ⁓ he has an upcoming ⁓ workshop in was it May twenty ninth?

speaker-2: Yes, and it is an in-person well workshop. So it makes sense only if you are in the Dallas Four Four Fort Worth area. I s I s I suffer from severe zoom fatigue when it comes to holding our workshops. So I'd like I'd like to have an in-person interaction with my students. And it's only for it's a half a day w work ⁓ workshop and the one-line di di description ⁓ which is also in the name. Stop prompting, start delegating. No authentic AI for your business. And it's basically going to be using Cloud Cowork. And it's going to be you, you I'll show you how to use Cloud Cowork in a privacy first way. Meaning, at no point will the privacy of your data be compromised if you follow what I am going to teach you in the workshop.

speaker-1: Well, you heard it here, folks. If you want to j g join if you and if you can make it down to Dallas area, join ⁓ Dr. Yesh on his workshop. Do you have any online workshops?

speaker-2: Not yet, no. I do have ⁓ I do have videos on my on my on my website of my talks and my ⁓ and my appearances on various panels, but I don't have quite any courses.

speaker-1: Gotcha. Okay. All right. Simote, you wanna take us out?

speaker-0: Well thank you so much, Dr. Yesh, for being here with us today. We appreciate the the insights that you've given us. And ⁓ and I am sure that we are all going to start thinking about ⁓ AI and agentic AI in a different way because because of ⁓ what you've shared with us here today. ⁓ and if you watching are ⁓ you've learned something, please put it in the comments. Let us know. What did you learn? What are you going to do differently because of this conversation? ⁓ and what's your biggest takeaway? What's your biggest aha? And ⁓ also share this with others. And if you want to hear more from more ⁓ business leaders like Dr. Yesh, please subscribe to this channel and ⁓ you know boost us in the in the algorithm so more more of your friends can see us and and More people will say it's okay.

speaker-1: And you can find us on Apple Music, Spotify, ⁓ Amazon Music now. So no no there's no excuse, in other words.

speaker-0: Look up the growth ceiling podcast. If you are looking to to get past that that ceiling in your business, if you've hit a plateau, we are here to help. ⁓ and yeah, and visit, click the links in the in the description. We're gonna give you a link to an assessment if you need help, help, ⁓ if you've hit that hit that growth ceiling, then ⁓ we can definitely help you. No obligation. Just meet with us. We'll be we're happy to meet with you. Anywho, we're done for today. So we will see you next time.

speaker-1: Thanks everybody. Thanks, Doctors. Appreciate it. Great conversation. We might we'll probably have to have have you come back for a part two and three, I think. But

speaker-2: I I love talking fast. I I can talk all day as you guys have been

speaker-1: ⁓ man, there's just so many so many tangents we could have gone down first. Sorry. Yeah, that's great. All right. Excellent. We'll we'll let you go. Thank you.

speaker-2: And if you can if you can if you can send me the link, I would appreciate it.

speaker-1: I will send you all the links and everything, don't worry. I'll get ya.

speaker-2: Yeah. Thanks, Simon.

speaker-0: Thank you. Thank you. Bye. Bye bye.

speaker-1: Bye bye, talk soon. Yep.

speaker-0: Wow.

speaker-1: right. Before I forget, I want to add something to I was consulting with Claude, speaking of ⁓ and I was ⁓ by the way, ⁓ really cool. I I wanted to bring this up, but I felt like it was gonna take us down a another tangent. But I f I I got I came across this really cool have you used Claude at all?

speaker-0: Little bit. I use it for like writing.

speaker-1: You know about the skills, right, that and Claude?

speaker-0: I have heard about it, yeah. I haven't really used them.

speaker-1: So there's this guy that created apparently he was ex-Google. I think he's ex-Google. He created, he's like, you know, you can't you can't trust the first output that AI gives you. Right. And so he created what he called ⁓ the council skill. Mm-hmm. It's four different perspectives. No, five, five different perspectives and a chairman. So it's it's four unique and then a chairman. and it

speaker-0: Okay.

speaker-1: They take it, they take whatever it is that you're giving them and look at it from different perspectives and bring toget come together in a consensus output. Which is and so anyway. The the council, I ran it, I ran the the con the idea, the question, like what is my differentiator for the for my podcast? I ran it through the council. The council came back. The first thing I said, you don't have one. Universal

speaker-0: Okay, interesting. Okay, thanks. Wow. Okay. ⁓

speaker-1: And so ⁓ the suggestion that they gave was okay, let's think about the whole thing, right? It's the growth ceiling podcast. So what you could do, what is you know, interesting, unique, would be to have a segment at the beginning of every episode with thirty, maybe ninety seconds where you're just talking about a recent growth clarity call. Okay.

speaker-2: Okay.

speaker-1: where you know, a diagnosis has been made or a conversation has b has kind of, you know, circ ⁓ kind of centered on things that are impeding the growth of a business. And you know, just talk about the results from that. What what are your thoughts on that?

speaker-0: think it's cool. ⁓ except I think I would want the person on the call ⁓ kind of in the room and they kinda talk and they talk about their own

speaker-1: Yeah, the c the council suggested not to do that because it's actually, you know, cutting into if if they're a guest, right? And and or

speaker-0: Yeah.

speaker-1: Yeah. It's it kinda like cuts up the th the whole rhythm thing. So that's that was their point, but

speaker-0: I I do like the idea of ⁓ of doing a maybe doing a ⁓ doing a call live with a person who where you're we're helping them with their business and coming up with like like the ladies one of the ladies today in the in the in the networking group. She was like I wanna talk to you about this whole, you know, surprise, delight, and wow. Like you've never really heard of that before. Yeah. ⁓ okay, cool. Good. Because you know, it's like it a lot of people think think that way when when we say surprise, delight and wow. ⁓ well, I serve my clients at a high level and you know, and it's like, okay, they're only thinking

speaker-1: She signed up, right? I sent you an invite to the call next Friday. What about the other lady was saying That's exactly what the other lady was saying. Like we we give great service, we you know, we're we we give them exactly what they're come they came for, you know.

speaker-0: Not everybody's gonna do that. Like everybody's gonna say, well, this is what we do. You know, no no other like competitor is gonna say, Well, no, ⁓ well, I haven't thought about giving good service. Hmm. You know? Nobody thinks about that.

speaker-1: Yeah. And so I I you know, the the logistics of this still need to be kind of worked out, I think. But part of the problem is my concern is that there's not going to be a growth clarity call every week.

speaker-0: Right. Right. Yeah, that's true.

speaker-1: I hope to get to that point, but it's not it's not there right now. I do have some backlogged that we could use, but I would imagine that you've also had conversations that we could

speaker-0: Yeah Yeah, I've had

speaker-1: So it doesn't necessarily have to be the growth clarity call specifically, but it could be you know, a time where we diagnose somebody and help them figure something out, right? About

speaker-0: Yeah, that's true. Okay. Well yeah, we could talk we can do that. ⁓

speaker-1: And and and say so the so as far as the logistics of that would would go, there's the figuring out what exactly we're gonna say about and about whom. Like ⁓ but then there's also like the actual doing of the thing. And the again the council suggested just recording several of them in a row. Take like and like thirty minutes or an hour and just doing several, banging them out, just get ⁓ all done and then just tack them on as, you know.

speaker-0: Yeah.

speaker-1: to the to the start of the podcast before we get into the conversation. And just have it that's where the actual

speaker-0: it as a recording at the beginning of the conversation. ⁓ that's interesting.

speaker-1: And it's just you and I bantering back and forth about it. That's all it is. B thirty ninety seconds max.

speaker-0: Okay. ⁓ okay. So we have a clarity call and then it's you and I talking about it after

speaker-1: Afterwards. It would be anonymized unless you know they're I again, so this is getting into the logistics, like maybe they want to be, you know.

speaker-0: 'Cause people do but ⁓ I found that 'cause like I've I've done a few interviews with people and and I'll tell them like, Hey, you know, ⁓ you're gonna be on my YouTube channel and ⁓ and we're gonna like talk about your business and talk about like even even things that you may not be aware of or you know, I'm just gonna run through what I call my customer journey mapping session, right? And and I do kind of a a a quick You know, and I try to make it conversational, but we go through it and a lot of times when people are like they're like, people are willing to do it to do it because they want the exposure. I'm like, okay, I'm gonna give you a chance to talk about your business and what you do, but I'm also I also want to take, you know, what you say, like, okay, you're you're great at this, ⁓ and this is your mission and this is what you want to do. Okay, great. Let's go through it. How can people and then let them know? Like, ⁓ and then they get to tell people how can you con how can ⁓ they contact you and what's your lead magnet, blah blah blah and you know, all that stuff and they're willing to do it. Not everybody, but some people.

speaker-1: Yes. So what ⁓ this just kinda came to me from what you just said. Where where I could see this going is we start it as like a ninety second intro to the current podcast. Mm-hmm. And then leading up to an eventual spinoff where that's the only thing we're discussing and it's actual we're ac taping potential clients. Well in a you know, essentially a a discovery session of sorts where we're diagnosing their current situation and giving them a prescription for moving. That becomes a separate

speaker-0: Does people get it becomes a whole yeah. ⁓

speaker-1: Then and then what we can do real quick before I forget this is ⁓ completes my thought. For if that is the case then, then we have a constant flow of ninety second segments for the other podcast and we can do cross promotions very easily. Sure. So anyway.

speaker-0: Yeah. thinking can we can't we do we have to spin off a separate podcast? Can we do them all in one podcast? I mean what I'm saying

speaker-1: It does does make sense, doesn't it? Yeah.

speaker-0: Can we do like ⁓ so we do s ⁓ some episodes that are that are, you know, we have a guest and the guest is talking about, you know, their expertise and what they do. And then we have some episodes that are, you know, like assessments or or whatever we want to call ⁓ you know, to like, you know, we have a guest today, blah, blah, blah. ⁓ our guest is gonna is an expert in blah, blah, blah. ⁓ but this part this episode is today is we're doing a we're doing a a session with so and so, you know. and that person for us would be a ideal client, you know.

speaker-1: Yeah, yeah, in that case then like the ninety second segments become like a prelude. Like make sure you tune back in.

speaker-0: Yes, yes. It become like a a nice hook to to keep watching. Yeah.

speaker-1: Yeah. I like it. I I'm excited about that. So we just need to figure out, you know, a time for us then to get together and record a bunch of these just to get it going and then tack them onto the front. And then we can work it out.

speaker-0: We can well, ⁓ I always say Tuesday's my podcast day, so I'm willing to do it. I'm willing to do it on Tuesdays. Like usually ⁓ like today, I have nothing from now until seven. Actually, do I? Okay. Actually I may have I don't have it on my calendar, but I'm supposed to be meeting with s with somebody today. But ⁓ but generally, like this is This is the thing for Tuesdays. ⁓ you know, and I don't have I know. I'm saying like I don't have too many meetings on my calendar. But ⁓

speaker-1: I don't have like what I wanna what I wanna do is like have a bunch from me, from my side, like so you know, lined up and we can just get ⁓ bang ⁓ out. I don't have that right now, but ⁓

speaker-0: Yeah, Tuesday, Wednesday. Yeah, Tuesday, Wednesday. I'm good. let's see. I don't have we don't have a whole lot. You know, I've been getting I haven't been getting too many people who have hit a growth ceiling. I have been been attracting people who are brand spanking new. That's you know those are the people. It is.

speaker-1: Or Wednesday? Self, that's its own feeling.

speaker-0: Y well, yes, that's true. I mean, and they're definitely in a phase where they wanna grow. Absolutely.

speaker-1: Yeah. And and we can kinda, you know, that can be kind of the framing. This this is for all of all of the the new business owners out there.

speaker-0: Yeah, that's true. We can frame it.

speaker-1: This is or this is business owners that have been around for a little bit, you know. I don't know. We

speaker-0: Yeah. 'Cause I always I do think that when people are are just getting started, ⁓ they should think about this whole this you know, putting systems in place and putting thinking about systems and thinking about automations and that kind of thing.

speaker-1: It's like David said, the problem is is they don't have s enough sophistication to appreciate it and to

speaker-0: Appreciate it. Yeah, that's true. I always think that they need to ⁓ they they do need to think about it early but b before they they get into a situation where ⁓ now I have to backtrack and undo a whole lot of stuff because I've done it wrong, you know? I'm like, let's think about this right at the beginning, you know? ⁓

speaker-1: This is an exact situation with a client of mine. Same situation. Like he is convinced that he he he's using a ⁓ what do you call it? ⁓ he has like this licensing ⁓ agreement with this ⁓ sales framework called sales QB. Have you heard of this? Sales QB? It's somewhat similar to you know the ⁓ predictable revenue system. Kinda, sort of. It touches on some of the facets but not nearly as deep in some of it. Anyway. He is convinced that he needs to ⁓ have his branding, even though he's he wants to eventually have his own separate personal brand, and he and he doesn't really want to you know hitch his hitch his wagon to the sales QB, but he feels that he's he's early enough on that he feels like he needs it, right? It's almost like he's he's trying to derive authority from this established licensed Thing, right?

speaker-0: Yeah.

speaker-1: And and what I'm telling him is no, no, no. Right now is the time when you should s establish your personal brand. You know, get their branding, your colors, everything set. and do it now. And then we can just include a logo that says powered by Sale QB. That's all you need. That's all you need to do. But it's so much easier to do it now than trying to extra ex extricate yourself later on from the branding that is attached to that phase.

speaker-0: Sales QB. Right. True. That's true. That's true. Everybody knows you. Whenever they see you coming, ⁓ that's the the Anway person, you know, and like you have no personal branding of your own. And then when if Anway messes up, they go left. And you're like, okay, I don't want any part of that. Then you go off and you're you join some new company or you go do something else. Yeah. People are still thinking of you as that anime person. And now that branding is all over you. And it's like, no, create your own branding. Yeah. Yeah. I was telling ⁓ I was telling a guy a while back and I need to get back in touch with him. This guy Rod in Texas. ⁓ he's in prepaid legal. And I was telling him, start get

speaker-1: Mass.

speaker-0: Get some workshops going. Talk to people and and talk to people about he had he had space where he can host people and serve coffee and cookies and whatever. Talk to them about like ⁓ setting their businesses up for success ⁓ and mitigating loss and legal problems and all that kind of stuff. And when you're ready, this is the solution, this is the thing you need. get prepaid legal, right? In your business. It's low cost, subscription based, you know, you don't have to spend, you know, a big chunk of money on a lawyer for a conversation, right? You've got access. ⁓ but you are you are the brand. You are the the the small business solutions guy, right? You are the person that they come to in, you know, in the s the service that you're providing is is ⁓ you know, powered by prepaid legal, right? Yeah. You know.

speaker-1: Then you become seen as a resource of various different things, yeah.

speaker-0: Yeah, you're the resource and you know, and you're not just, ⁓ that's the guy that's pushing prepaid legal. Yeah. Yeah.

speaker-1: Yeah. Yeah. anyways, so I don't know. when when do you want to bang those out? When would be good for you? I know you said today works but I'm just not really

speaker-0: I've got a full day. Yeah you do. ⁓ today today ⁓ well, ⁓ tomorrow is I've got a free day. I've got free you know. I'm free after like nine o'clock in the morning.

speaker-1: What? Really? For like the whole day? Or

speaker-0: Yeah, pretty much the whole day. I don't know. I don't have any ⁓

speaker-1: We can do it tomorrow. I can I can work on putting stuff together then today. ⁓ 'cause again, I'm like you said, I'm gonna be working on podcasts today, that's it. So I'll I'll put I'll put some of those together. I'll I'll I'll go through like my transcripts, you know, from from my calls that I've had and you know

speaker-0: Yeah. Yeah, that's a good idea. So ⁓ so what are we talking about? Are we talking about like okay, like the the different problems that people are having or like

speaker-1: What do we feel like there's a reason why

speaker-0: Insights that they're getting.

speaker-1: Well, yeah, could be. But there's a reason why they're talking to you, you know. What is that reason? How did you how did you help them achieve a certain level of you know, clarity? ⁓ figure out a path forward, something along those lines, right?

speaker-0: So maybe like a like a transition.

speaker-1: Their situation. The time where you're you're looking at what they have, right? You're asking them where they're currently at, getting a diagon so that you can form a diagnosis, right? Mm-hmm. Well, I see that you have this, this, and this. ⁓ you know, in my opinion, this is what's going on and this is what you should be working on. That's essentially what I'm talking about.

speaker-0: Okay. Okay.

speaker-1: Do you think it you can pull stuff like that together from conversations that you've had? ⁓ It I think so. It doesn't have to be like it can be anonymized, I think. I I Yeah. And what and what I'm thinking is like to get us both in on it, you know, we could have like a Q and A of sorts, like A AI can come up with the questions, you know, but ⁓ based on the transcript. Okay. Seems yeah, I would I'm thinking low friction. e make it as easy as possible on on on ourselves here.

speaker-0: I feel like I need I need to get into N8 and and have it go through my Fathom notes. ⁓ because what I've been doing is like taking my Fathom thing and copy the transcripts into Notebook L ⁓ but ⁓ I need a more ⁓ a more automated way of doing that. So I would love it to like I've had some meetings with people. Go through all these meetings.

speaker-1: Okay. Use Google Workspace.

speaker-0: I don't know.

speaker-1: Alright, 'cause they turned on Google Intelligence now. Have you sort have you heard about this? Mm-mm. Basically, if you use Google Drive, it's now trained on the Google A Google AI is trained on everything in your drive. And you can just ask the AI questions from anywhere and and it will give you answers about things. Including if you open up Gemini and you're just t you know, if you've turned on access to workspace in Gemini. So What somebody had shown me was like what you can do is you can use part of your context. You could also include Notebook LM, but then it has all everything else in contacts as well that you've stored in in Drive. Anyways, so for me, I'm I'm putting in all of like I have a ⁓ a za a zap that that takes all of my fathom transcripts, puts them into Google Drive. Yeah. And so all I have to do is ask.

speaker-0: Yeah, well.

speaker-1: chat or ⁓ Gemini, you know, about about a conversation I had with this person. It it'll bring up everything about it. Wow. I don't even have to go searching for it. It's magical

speaker-0: ⁓ that's awesome. Okay.

speaker-1: Yeah, it's really kind of frightening, but

speaker-0: Because I don't put all of my ⁓ well I have a few, but I I don't put all of my ⁓ conversations into notebook LM, just just some of them. Like people who come back to me multiple times or decide like, okay, yeah, I want to work with you on something, then you know, I'll put them into notebook LM, but everybody else, it's like one off conversation. All right, I'll have to see. See what I can how I can do that.

speaker-1: Yeah.

speaker-0: Yeah. Maybe I'll just go through the ones that are already in my notebook LM. I could do that.

speaker-1: Fathom also has are you paying for Fathom, the Pro?

speaker-0: No, I've got the free.

speaker-1: I don't know if it's on the free version, but on the pro, you can ask, you can now chat with Fathom about all of your conversations. Did you know that?

speaker-0: Yeah, I didn't know that. No. Okay. ⁓

speaker-1: ⁓ You can ask it across and it'll apparently look at all the all the recordings that you have stored. So you can ask it about themes and you know, concepts that you've discussed and that kind of stuff and like, you know, pr presumably like ⁓ customer analysis, sentiment analysis and that kind of stuff. Like I haven't really dug into everything that you can do, but That's a another potential solution situation. But I think the pro version is a hundred and or two hundred and something a a year. It's like twelve it's like twelve bucks a month, I think. Two hundred and forty four. Remember.

speaker-0: Okay. ⁓ yeah. Yeah. Okay. Okay.

speaker-1: Anyways. All right. So do you want to try to get together tomorrow then?

speaker-0: Yeah. Yeah. When ⁓ afternoon.

speaker-1: What what I'm thinking like after lunch, yeah.

speaker-0: Okay. What are they like one o'clock or something or

speaker-1: Yeah, I'll just send I'll send an invite for one. Okay. I'll send. Thanks, bye bye.

speaker-0: Okay. All righty. See ya then.