482 – AI for Small Business with Kevin Surace

Foreign.

Welcome to the how of Business with.

Your host, Henry Lopez, the podcast that helps you start, run and grow your small business. And now, here is your host.

Welcome to this episode of the how of Business. This is Henry Lopez and my special guest today is Kevin Sures. Kevin, welcome to the show.

I'm so happy to be here. Henry. Thank you for having me.

I I am excited as well. This is a a big topic, obviously artificial intelligence, especially as it impacts small business owners. So Kevin is with me today today to share his knowledge and his vast experience with artificial intelligence and the current and what he foresees is the potential impact of AI on small businesses like ours. He's also going to share highlights from his incredibly and fascinating varied background. You want to receive more information about the HOWA Business, including the Show Notes page for this episode and how you can continue supporting my show and receive exclusive content and discounts through a Patreon membership. Just visit thehowwahbusiness.com I also encourage you to subscribe wherever you're listening to the show so you don't miss any episodes. Let me give you just a very brief bio on Kevin. He's accomplished so much that I could go on forever. But Kevin Sarrace is a Silicon Valley innovator, a serial entrepreneur, CEO and TV personality. Kevin has been featured by various periodicals and TV networks and has keynoted hundreds of events from Inc. 500 to TED to the US Congress. He's been widely featured and recognized in periodicals including Inc. Magazines, Entrepreneur of the Year and CNBC Top Innovator of the Decade. While he has a technical background with 94 worldwide patents to his name, he's also known as a highly dynamic speaker. And that's one of the things that he does, is he's a speaker who is a true entertainer. He's funny, he excites people, he educates, he energizes audiences to action. Kevin has led pioneering work across numerous industries including including cellular data, smartphones, semiconductors, artificial intelligence, which we're going to dive into today. Construction and building energy management system including I was very interested as I was doing the research. Kevin, your your work on window and energy retrofits on the Empire State Building and the New York stocks change, just to name a few. He's also an accomplished music director, as if he hadn't done enough. Conductor and Broadway and streaming producer and percussionist himself. Kevin lives part of the time in Sunnyvale. I think he's in upstate New York today as he chats with us. Kevin Sarrace, welcome to the show.

Well thank you for having me. My, my goodness. I'm not sure I can live up to all of that.

I suspect you can, sir, because you have accomplished so much. So I think that you can. And so give us just briefly, you know, you started your career in technology and technology sales, if I got that right. In fact, we, we may have worked for competitor slash partner back then, Back then in the 90s. Uh, this was after you graduated, graduated from the Rochester Institute of Technology with an engineering degree. So the thing I always am curious about, if you can put yourself back then in that time frame, especially when you were in school, what did you want to do when you grew up?

Yeah, absolutely. In fact, I, I, I always wanted to be an actor. I wanted to be on stage and I wanted to be a drummer and I, I still play drums, but. And so people said, well, why didn't you pursue that? Well, in some way I did pursue it. I pursued it as my second career while I had my first career. But my dad, probably rightly so, and I fortunately listened to him, there was no choice at that time, but, but he said, that's interesting. You can do acting, you can do your stage, you can do that on the side, but you're going to go get an engineering degree and because that will do you well now I loved technology and I loved engineering and I, I had a little radio repair business and TV repair business when I was a teenage major because I could fix any kind of electronics and I would build things and I really, really loved the electronic space. And interestingly enough, my, my father worked for GE Audio Electronics at the time, which made radios. It was a billion dollar business at one time with GE was number one in radios and tape recorders and clock radios and things like that that we don't even care about anymore. But at that time that was a very large business. And so, you know, he was really in the electronics business. He was in marketing and sales. But, but I, I kind of grew up around electronics of the time of, of the day, which at first had tubes and eventually had transistors, right. And eventually integrated circuits. And that was just a fascinating time and how fast things were changing. The LED came out and you could have digital clock radios now. And for most of your listeners, they, they may not remember that, but that, that was thrilling. We had digits instead of a circle, right. So I was always interested in that. But I loved my music and I love my theater work and my dad, you know, said you, you know, go, go for a real degree. So not that, not that performing Arts degrees aren't real degrees. I love performing arts. But, you know, it was like you're, you're, you know, you'll be better set on your living. And he was right. I ended up going to Silicon Valley, as you were, I think, in Silicon Valley, and, and building a career in technology. And that career has led me to all kinds of ups and downs, as technology always does.

Right, yeah. Tremendous opportunities there. Why? What drives you or what do you feed off of when you're on stage? You're doing it now, obviously, as a speaker and primarily, but because I similarly, you know, when, when we're doing sales, you get to be on a stage of sorts. Right. Small audience, but. So you get to be on stage. What is it that appeals to you about it?

Well, look, I like to say I'm an edutainer. Edutainer. So that's education and entertainment combined. I have had a lot of experiences. I've been working in and around applied AI. Now that I'll separate that from core AI, where you're developing new AI algorithms, to using those algorithms to build a product that does something unique for customers. Right. For users. And so that's the applied side. And I've been focused on the applied AI side since the mid-90s and developed what later became Siri, later got licensed to become Siri. It was called, it was a General Magic. It was called Portico and my talk and Magic Talk, eventually General Motors, OnStar, we built that system for GM. So this was fascinating work, of course. And, and, and so for me, getting on a stage is to be able to share some of those experiences. And since I did not get to be on stage, acting, singing and dancing, that's about as close as I can get, right? That and these wonderful podcasts and some TV work. But I really do love, love sharing those experiences and love sharing them in an entertaining way. Yeah, in a very entertaining way.

You love communicating and doing so entertainingly so that it. So that people listen.

Hey, we've all seen keynote speakers and I do 40 or 50 keynotes a year, mostly for corporate events, and we've all seen keynote speakers. They are 100 times smarter than I am, I can guarantee it. You know, they've come from a university and they stand behind the podium with that piece of paper. Hi, I am. Oh, let me look. John Smith. I'm going to talk about cancer detection from, you know, whatever very important topic. But, you know, half an hour in, everyone's on their phones or asleep, right? One or the other. And so I think when you Have a large audience. It's different when you got 20 people. When you have a large audience, you have really got to project. You have really got to project excitement. And you got to be funny and, you know, you got to keep it flowing. Because I've spoken audiences of over 5,000 people, I can tell you that I gotta. I've gotta entertain and engage the person all the way in the back row, which might be row 50 or 80 or something like that, Right. And I think just a lot of speakers don't think that way. They go, here's my material, I'm going to educate. You know, it's not a lecture. Keynote speaking isn't a lecture. This is. These are. These are corporate people who've come to some event, right? Some conference, and you're standing in the way probably of either their breakfast, lunch, dinner or wine. So get off the stage, say what you're going to say, get off and, you know, and leave them with two or three nuggets. Right.

So at what point or did you from early days aspire to or have the idea that you would be an entrepreneur?

Well, there's a great question. No, I didn't know that I would be. I started working in a large guy who started out with IBM for a while and National Semiconductor and then a division of Epson called smos, Seiko, Epson, that was all in semiconductors. And then I went to a small company and I got a taste of what it was like to be in a company with 20 people instead of 20,000 or 100,000 or 500,000. And that was really, really interesting. And then a friend of mine had an idea about starting a company where you would have a cellular phone and it would actually have a screen and send and receive data, not just voice.

Interesting concept.

Yeah. And that was in the mid-90s, so we're talking about way before there's an iPhone. And so we built a product called Air Communicator. We started a company, built a product, got it funded. And it turns out at that time, only the FBI wanted to use things like that for a lot of very special reasons. And the other thing were these what we call, road warriors. At the time, road warriors were heavy travelers that actually had a need to access their data in some way or get their email or things like this. Right. And in calendar and et cetera. So. So we developed all of that and we had the first product out doing that around, you know, 93, 94.

And so that was. So then that gave you a taste for entrepreneurship and that put you down that path. Although you've done entrepreneurship, you've gone also. And I mean, I don't know your current organization. Are you a founder or serving as CTO there?

What is the situation now at Fance? A founder? I was CEO for a long. I hired a CEO for those who have not. You know, you talk to a lot of small businesses on this program and anyone who has been a CEO, whether it's a tech business, whether it's a venture funded business, or whether it's just small business, family business, whatever, you know, the CEO job is, is, is, is the one everyone aspires to. And then when you get there, you wake up every morning and go, how did I get this job? Right? And often what happens is an awful lot of your time is spent on people issues, but a lot of your time is spent on dealing with unhappy customers. And you know, remember, all the bad things end up on your desk because you're the one who has to solve them as the CEO. Right? So you really do less and less of the fun thing, whatever's fun for you, less and less of the fun things. Now there's a lot of cachet and well, I'm a CEO. I've been a CEO so many times. May I never be CEO again, right? But what happens is something happens at some company and people drag me in and say, well, we need you, you need to fix this. Can you be the CEO and I'll be the CEO for a little while while I find a CEO. Because, you know, at this point, I, I don't want to prove, I'm not going to prove that I'm the best CEO in the world and I'm not the best CEO in the world are far better CEOs. But I really enjoy the technology and I really dive into the technology or the problem. Sometimes it's not technology or the business problem or the channel problem or whatever it is, I want to go solve those. Right? And as CEO, you're just too busy with financial and people and problems and so politics do that, all that. That's right. That's right. So it's not all it's cracked up to be, folks. That's what I would say. So a lot of, a lot of advantage being in strategy, technology, cto, cio, cso, perhaps on the board. I'm on several boards, but in the case of Advance, I founded it. In the case of Token, which is a cyber security company, Token Ring, I help helped to get it sort of relaunched in a new direction a couple of years Back and in the cyber security ransomware direction. And that is just going to be gangbusters.

Yeah, that's fantastic. Okay, so there's a common theme, obviously, with technology, but as I was doing the research and you've done so many things and been a part of so many different organizations, I was trying to identify another common theme. But it, but it seems like you've taken an opportunistic approach to what do you do next? But my question is, then, how do you decide what you do next?

Well, here's the interesting thing, at least for me. And I, and I think for a lot of people, what drives me is I wake up one day and go, wow, there is a problem. There's a pain point, real pain point. And that pain point, I wonder if that could be solved. Let me think about that. Right? And I may think on it for an hour or I may think on it for a year, but when you find a real pain point, that's where there's real business. And I think that so many of our businesses, we come up with an idea that's looking for a home or a feature, that's looking for a product or whatever. Right. And so you do not want to be this technology that you develop that's looking for the pain point to solve. It doesn't solve anything. And we see a lot of tech companies, a lot of funded companies like that. We see a lot of mom and pop companies. We see everything. Right? And they're looking for, you know, they've got a solution, looking for a problem.

Yeah, they've got to, you got to evangelize to try to find people who uncover a latent pain. Perhaps. At best.

Yes, right, at best. So if you find a pain, if you're out there and you see a real pain point and you see nobody's got a solution to it, then you say, well, okay, do I need to break the law of physics to fix this problem? And if, if the answer is yes, then you shouldn't do it. But if you don't need to break the law of physics, and then you should step back and say, okay, do I have the intellect or do I know people who have the intellect? Or can I, which is much of what I've done in my career, go read other patents in and around this space to learn how other people might have targeted it even though they never solved the problem. Right. So. So there's so much to learn. It used to be if you needed to read a set of patents in a particular field and maybe apply them to A new field, you'd have to go to a patent library, and God help you search that because it's all on paper. Well, of course, for the last decade or so or 15 years, we could just search online. We do a Google search, find every patent that has, you know, this kind of technology in it, find out how that was utilized, maybe in other places to solve problems. And then you go, I've got a new application for this nobody's ever thought of. And so you can really. I don't want to be an expert in a field, but you can be a near expert. You may, in fact, be able to be an expert in the application of that technology to your field in a matter of weeks with enough reading.

Got it. Yeah.

So.

And so. All right. So how do you determine, though, when it's time to move on to the next thing?

Well, that's a good question. You know, usually have to drag me kicking and screaming, I suppose, because usually I have a lot of things going in parallel. Right. Again, you know, you sort of come up with something and. But that's nascent and you're still working on something else. So. So, you know, I tend to be sort of, you know, working at a few companies at the same time, but. But some are in a very different stage, so it's an hour a week, you know, and others are like they're 40 hours a week. Right, okay. And so there's this, this. This sort of overlap, and then as one goes public or as one gets sold or as you bring more people in and they don't need you as much, which is, you know, at Token Ring, which is here in Rochester, New York. You know, we started when I restarted it and refunded it, we had about 12 people and me. And, and. And over time, it didn't need me anymore. It needs me very little. A few hours a week. And the reason is we brought in way smarter people than me in this particular field, in the cybersecurity field. And so there's still a lot I can add, but there is. There's immense brain power at the company now. And so over time, over a couple of years, I kind of replaced myself, which was the. What's was the idea. I'm still on the board. I'm still chairman of the board. I'm still very active in the company, but I don't need to do certain things anymore. I can check in on it because there is a team that does it. And in fact, the best thing you can do is any level, cso, cto, CEO, best thing you can do is replace yourself with people way smarter than you and still have your percentage of the company, still maybe be on the board, still whatever. But absolutely, because otherwise you're going to just be doing that. You can't get out of it and you want to love it, but there's a lot of other people. So you want to get so successful that the money's coming in, things are happening, and you replace yourself with people better than you.

That's when you become an entrepreneur. This is Henry Lopez, briefly pausing this episode to invite you to schedule a free coaching consultation with me. I welcome the opportunity to chat with you about your business plans and offer the guidance and accountability that we all need to achieve success. As an experienced small business owner myself, I understand the challenges you're experiencing, and often it's about helping you ask the right questions to help you make progress towards achieving your goals. Whether it's getting started with your first business or growing and maybe exiting your existing small business. I can help you get there. To find out more about my business coaching services and to schedule your free coaching consultation, please visit the howobusiness.com take that next step today towards finally realizing your business ownership dreams. I look forward to speaking with you soon. I'd like to. Let's start diving a little bit more now into I could ask you questions about that forever, but we'll, we'll focus now on AI. I'd love to get your definition, especially as we're talking about generative AI or as you have called it, applied AI. Don't know if those are the same thing necessarily, but I'd love to get your explanation or definition at a high level from a business owner's perspective, if you will. What is generative AI?

Sure, generative AI. Well, generative AI is a type of artificial intelligence. We'll get right into this. That's based generally, but not always, but generally on a transformer model. And this model is a translational model, meaning that instead of training on words alone, it trained on phrases. And this came out of work at Google in 2017, and it came out of work at Google. That said, when we translate English to French, we're doing it word by word. And in France or in Spanish or some other language, it doesn't make any sense. Right? Right.

I mean, what we call literal translation often comes out completely wrong.

You know, it's wrong. But you know what they were saying?

Sure, but.

But it's like you're not speaking my language, you're speaking junk. So they had an Idea that said, well, it will be hard, but if we trained a model on phrases and got lots of them, then we would wait and hear a phrase and we would convert the phrase to French or Spanish or whatever it is. And it turns out that was a really good idea. It works. And all modern translators, today, most modern translators are based on that method. You type a whole phrase or you speak a phrase and it will translate the phrase rather than word by word by word. Okay, great. Once you could learn phrases, well, then if you learned billions or maybe trillions of those phrases and stayed in the same language, you could start to put together logical words that form reasonable phrases that answer the question. So if I, if I asked a large language model, of which there's many, but most people know about ChatGPT, but there are a hundred others, right? Many, many others, including one released today in open source, a larger lambda model from, from Facebook. So what's interesting here is that if you stay in the same language, I can start to ask it questions like, you know, who was the first president, United States or whatever, and it will automatically, through its deep learning network, and we could talk about how neural nets work in a second, it will start to find things that match essentially the answer to that question right now. That's incredible because instead of using a rules based system like I had to do when we built Portico, and Portico was a virtual assistant, an AI assistant, and she would talk to you, Mary would talk to you like, like we're talking. Except she had a rules engine and only a rules engine. And so if you said, mary, I, please give me my email or read my email, she'd go, okay, I've got your email, let's go. And you'd say, mary, I'm really in love with you. Oh, I, you know, I'm taking, you know, whatever. She had all, all kinds of responses. But in the end, if you, if you tested that many, many times, you'd run out of rules and run out of rules itself. Yeah, so, so that was a rules engine. We used those and they were to some extent artificially intelligent. It did use machine learning, but only on the speech recognition part, called a hidden Markov model. But here's the interesting thing I want everyone to take away with generative AI people, it's just math. It's literally math. In fact, if you were taking courses at any university today in any kind of AI, including generative AI, it's probably in the math department. Why? It's math. It's not science, it's not physics, it's not taking over the world. It's math. And it's just a probability on one word coming after another to form a final phrase that makes sense in the English language in this case. Right.

So, and then over time, that model is learning quote unquote, and, and, and having its probability be more accurate as.

A result of that could. And it can, it is. Not all of them do that because many of them already learned from almost, you know, a trillion tokens, a trillion phrases, and in a trillion phrases, me adding yours to that may or may not help.

Got it.

But, but if there's supervised learning where I can go back and tell it that was a good answer or bad answer, and most of these will do that, it will take that and say I can learn from that. Okay, they told me this was a good answer or it was a bad answer, but I need that supervision and to tell me otherwise, I don't know because I gave you something, you didn't tell me if it was good or bad. I have no way of knowing. Right.

Right.

Now this is, this is absolutely fascinating and it's blowing people's minds, but those of us who have been in and around the space now for 70 plus years that AI has been being developed, we could see it coming like, this isn't a, you know, this isn't a shock. And there was a GPT one and there were large language models before that. And so for the last many years, people have been toying and it's been getting better and better and it's. So it's not some shock, but it is a shock to most humans.

And I have a question for you there. From a, you know, from a business marketing perspective, what do you think became the tipping point? Why, why did chatgpt all of a sudden catch fire?

Sure. Because they released a public version in chat GPT3 and then 3.5 and then 4 and before that it was only involved like I was on 1 and 2, but it was for researchers and it was for, you know, understanding what these language models could do or it.

Was being of course embedded into other tools that we use. The other software solutions that we use. It was being embedded there. It's just, you know, for two years.

Into Copilot with Microsoft and GitHub, so, so coders for two years have been getting suggested code and that was based originally on GPT 3 and then 3.5 and now 4 for, you know, for the same reason. So people have been using it in certain ways, but, but most of the public didn't then the public got access to it at some level and something like 16 of the public has tried it. It's not a large percentage tried it. And, and they love it. Now they don't know what to do with it after playing.

Exactly.

Usage has gone down. Right. But.

Right. And that's. And so that's where I want to get some of, some of your thoughts in from a small business perspective. I'm using it almost daily, by the way, to generate content primarily, yes, of course, of different types, whether it's marketing content or internal content or training content. But I think that what's happening here, and this is why I'm starting to address this on my show, this is actually the first time I've done any kind of a deep dive on it. Kevin, is it's. I compare it for small business owners similar to websites, although it's improved. Not long ago, only about 50% of small businesses had a website. And why is. Because they were paralyzed by not knowing how to do it or what to do it or they got their money taken, you know, so they were paralyzed by the technology. Is that part of what you're seeing here is that they don't know how to use it?

That's right. You know, small business. And by the way, when we talk about use it, you actually have to go to build something on top of it, right? For the most part, yes. You can go and type something and it'll give you an answer and that's cute. And then you go back to your business. But let me give you some ways that I think small businesses can and are using this. Right. One is customer support. If you integrate GPT or any large language model in your customer support system to handle basic customer inquiries, provide automated responses, assist with frequently asked questions, things like that, you now have a customer support person that you don't have to pay or you're paying, you know, a penny a response. That's amazing. Now that's amazing. Why is it amazing, by the way, is we can't hire enough people. And that's not going to change. I hope everyone listening to this hears those words. Right? The birth rate in the United States has gone down for 30 years. The number of graduates coming out of high schools is going down, and the number of people retiring is, you know, starting to skyrocket as baby boomers get to that age and going through that age now. So now you have a real problem because every small business I talk to says, oh, hiring is my number one problem, but I have a secret weapon. I'm advertising Here I'm doing. I said, yeah, you and everybody else, you're all using the same thing. But in the end, the pool is getting smaller. Right? So if you want to double your business, even in the next 10 years, the old way to do it was we just double the number of employees. I used to have 30 employees. I guess I'll have 60. Well, maybe. But no, you're not going to find another 30. It's going to be extraordinarily expensive. So you need to amplify the employees you have. And the way you amplify them is leveraging AI, customer support, content generation. You're generating content. I'm generating content. Anyone who is smart is amplifying their own brain by 10 or 100x. By what? By generating content. Not that I couldn't write it, except what I find is it gives me so many good ideas. I still edit it and clean it and this and that, but it's. Instead of spending maybe three hours, I spend like 10 minutes.

That's the time that I would have taken. I just did something the other day that would have taken me for particular the outline and kind of the ideas, the framework of it. Then moments I have it, and now I'll take it and massage it and make it mine. But the time that that has saved me is brilliant.

Yeah, look. Look at artwork for PowerPoints, right? Or artwork for websites or whatever. It's not that we didn't. It's not what they don't want to go to artists who will draw this stuff and create these things. But for me to go outside and go to someone, and it's not even the money, it might be 150 or 200 to get someone to draw up some, some. Some comps for me. And we'll take a look. But the cycle of that is days or weeks. Can I get on the calendar? Can I explain what I want? Can they draw five, you know, ideas, then come back, give me the outline, then I go, finally, you know, at the end of all that, maybe I have the perfect drawing for that slide or for that advertisement or whatever. Okay. If I use one of the image generators that are out there now, mid Journey or others, I, I or Dali, I can have, you know, five drawings in front of me just from a description I want. I need a drawing of a cat on Mars eating a mouse. Boom. There's a cat on Mars eating a mouse in a. You know, dressed like an astronaut, et cetera. Now I can't draw that. So what this has done is it's democratized drawing, democratized, actually, even photographs. I can, I can now take some photos of a person like a female model that, that says, hey, I need new model photographs. I can put that into a particular engine and within a few minutes I've got photorealistic versions of her in totally different places that she never went to that democratized photography because no photos were taken. I just took whatever photos I had of her face. Right? So all of these are interesting for small business. Generating ads, generating content, generating content for the web market research, personalized recommendations for your customers. That's fascinating because that's hard to do. Language translation, content editing and proofreading. So you can actually go back and say, please proofread this and make it better. So I could take something from a GPT model or a large language model, fix it, edit it, put it back in, literally, copy and paste. Here's another one. Customer writes to you screaming, this happens that every small business. They've sent you a terrible text, a terrible email. Copy and paste that into a large language model and say, look, I need a reasonable, thoughtful, caring response that hopefully will drive the customer to actually buy more from me rather than write bad things online about me. Okay? Frankly, a large language model is way better at that than we are in the moment. Because in the moment we're angry and we are going to want to use our emotions in a way that are not going to bring the customer closer to us. It's going to get the customer to give us a bad Yelp review. Right? So we want to change that. Well, let this assistant, right, this additive thing, this thing that multiplies our intelligence, let it do it for us. So in workflow, automation, things like that. So if you look at those eight or 10 things that I listed, it's as if I hired 10 different people with 10 different skills and I amplified my intelligence because there's still only one of me. I am still the robot overlord, except I have 10 other people working for me that didn't work for me a day ago.

Yeah, well said. Thank you for sharing all those experiences. Funny or ironic that you shared that last one, because I have a client, a coaching client, who did just that. They had a customer service issue, had Chap GPT create a response, and both he and the other, he asked his other service person, what do you think about this? And they both agreed it was better than what we would have come up with. Right.

I had a. A friend of ours text us. She came out and visited and stayed with us for a week and then went back and she's, you know, very senior C level exec. Right. She wrote the most beautiful text to me, to me and my wife, but thanking us. Right. The most beautiful text. And the prose was so stunningly gorgeous. Just, it's like she must have thought about it for a week, you know what I mean? Just every word and the placement and it was just gorgeous. And so I took that, put it in Chat GPT in this case, and said, what's the best way I can respond to this?

Ah, based on.

And the response was in this. And I said, respond to this in the same prose. Right, right. And, and ChatGPT gave me a response that was stunningly gorgeous. And so I pasted it and I, and, and at the bottom I said, written with help from ChatGPT and sent it, which made her even more joyful. So the response was gorgeous. It were. There were words that I could not have formulated. And, and, and, and when I read it, I said, this is from my heart. I wish I could have put those words in exactly that way, but I'm going to take it. And, but I was honest. I said, hey, look, you know, this was done with help from Chat GPT. I think I changed three or four words. That was it. So, you know, interesting, great. Not always good in school, because everybody's using this in college, even in high school, scares the teachers. But, but, but let me say something about that. It's not going to be asked, but it's interesting because what that does is all of the instructors, all the professors said all the kids are cheating. They're just using ChatGPT. Okay, so then some of the instructors have said, you can't use it and it's barred and this and that. Look, the truth is you're now in a world that has language models the way we've had math models, the way we've had math automation. So we've had math automation since the 70s. We had calculators and then we had Excel. So we've had math solved for us for decades. And we don't think anything of going to that tool, do we?

Yeah. Calculators now are allowed in testing environments. Right, of course. That's okay. Yep.

So, so that's what I'm recommending. I'm on the board of RIT as well, and we talked about this in the board meeting. And certainly my strong recommendation is embrace it. It's going to look, kids are going to use it anyway. And worse yet, well, better yet, whatever, when they get out of school, they're going to be expected to know how to use it. So you want to now embrace it in every class. And yes, there's a time when all computers are off and people may have to write out a few sentences, just like you'd say in math. I need you to do some long division in fifth grade. After that, you'll never do it again. But just do it once. Right. So you might make them do some things once or twice, but then after that, I want you to understand prompt engineering on these models. Right. I want you to understand reinforcement learning on the models. I want you to understand how you get the best answers and the best result and how you edit that, how you. How you use your editing and proofreading skills. All of those things are things we've got to be teaching these students. And, you know, but right in the style of Shakespeare. Sorry, that one solved. That's like multiply, you know, three by 5278, you know, that's solved. It's a button. It's just a button. So we now have language automation the way we have had math automation for decades.

Agreed. Agreed. You made a statement in one of your speeches that was recorded related to this, that I want to read here. You said, quote, your job will not be replaced by AI, but you could be replaced by other leaders that leverage AI to do their job. That.

That is exactly right. That is exactly right. It's at this point, either including in every small business, either you are embracing these new tools which consider them a calculator for language. Okay? Either you're embracing them or you're dead.

Or your competition is. And they're going to kill you with it.

That's right. Or your competition. So you're either embracing this or you're gone.

And it's a democratization of this tool now that's accessible to me. Yes.

Yes, it's. And that tool democratizes all of these things you used to have professionals for. I need a content writer. Well, no, you don't. I need, by the way, you couldn't afford one. You didn't have one. Right. Because if you have a content writer, you're going to empower them. And now they're 10x the content writer they used to be. Right. So everyone asks, is this going to eliminate jobs? What I always say in my talks, and I've said for whatever a decade, this is going to eliminate tasks, because some tasks you no longer need to do, but if you're an expert in that particular field, you're going to use your expertise to control the machine. Right. So if you are an expert in images and the way they look. Yes. You used to take those through a camera and you may still take some with a camera. Camera's not over. But for some cases you're going to take some other things that you've got. You're going to upload them and you're going to say, place that person at Niagara Falls and, and make them look at this angle and do this and do that and it's going to give you 20 of those back as if you flew with the person to Niagara Falls and took that picture. But you no longer have to. Now you could have always done some level of that in Photoshop. You could have even changed the clothing and the look and you know, but it, but, but that took real expertise, a lot of expertise, sometimes days and days of work. Now we've democratized that. Give me another one. I generated, and there's some others who generated film level content. It's not quite at the level of Hollywood, but generated film like content just from text. You know, I want this to happen and I want this to happen. I want them to be scared of this and I want, you know, generate a one minute film, okay, frame by frame. And it generates, it generates the video, generates the audio, generates the speaking. Everything's generated. Now people go, I can't believe it. This is unbelievable. Well, hang on, Hollywood's been doing that for 25 years. You ever heard of Pixar? Right? This is not new. Right? They've been able to do this for a very long time, right? For a hundred million dollars and then for $10 million, but not for 10 cents. So this is democratizing the kinds of tools that Hollywood has had to use for decades. Democratizing it so we can use it.

Yeah, yeah. Another area I always think of, Kevin, as, you know, what we used to call big data. And it seemed like only large corporations were able to analyze big data because they had the computing power and the analysts and the people. Well, now as a small business owner, I can apply AI also to help me interpret that data, can't I?

Literally, you can take an Excel spreadsheet, upload it and say, you know, what's the best interpretation of this or which based on this data, which customer should I go after? Whatever, it will analyze it for you and give you an answer. It's fascinating, right? It's unbelievable. Unbelievable. So you're democratizing that kind of thing for small businesses and doing that kind of analysis where again, what have you done? You said you might have been able to do that with, with formulas in Excel or Visual Basic in Excel. But you probably don't know that you're running a business, you don't have time for that. But just upload the stuff. Let the, let, let the machine come back with some recommendations.

Yeah, I just want to just briefly come back into, to your. One of your current ventures advanced that AI having been a developer. I started as a COBOL programmer back in the 80s and then, you know, we were in the software industry. And now I look at this. What a wonderful place to apply AI. This, this area of application testing.

Yeah, absolutely. So all applications, all programs have bugs. We know that everything we use online has bugs. We find them, we see them. Well, these applications are getting more and more complicated. And when we started this company, the thinking was, and that's about a decade ago, is that we had to get to a point where the machine will find the bugs for you. The machine will find the bugs for you. And we've been using a type of generative AI since 2017, not using a transformer, but a very different model that would generate test scripts and, and go through the application to find the bugs for you after some training. So you might train it for a day. And after that it goes and does its thing and it learns at each new build. That is miraculous. And every one of those customers, you know, millions and millions of tests, sometimes that AI comes back and says, warning, I found this. And they go, we would have I100 customers while. We would have never found that. Yeah, we would have never found that. We wouldn't have written a test script to go find that. Our customers would have found it. It would have been, we would have had a revert. It would have been bad. So leveraging AI in that way to go find software bugs is unbelievable. It is not marketing wise, but technically true. About 100,000 times faster than humans writing test scripts to find their bugs. But it's a machine. Of course it is, right? I mean, it's not. It shouldn't surprise anyone yet. The, you know, 99.999% of, you know, companies out there are still writing scripts or manually testing their software. And mostly because they didn't trust the AI. I had more customers come to us, we give them a demo and they go, well, I just watched it generate 5000 scripts in a few minutes. How, how can that be? I, you need to tell me what each of them does. So while they're in a folder, you can look at that. Well, well, how am I going to see 5,000 scripts? I don't know how I can. You know, I got. Everybody calm down. It's AI. It's going to go find your bugs for you. I don't believe it. Okay, well, let's run it. And of course it finds all these bugs. It goes, well, they're not all real. No, no, no, they're real. It's a machine. It doesn't lie. I mean, literally, Henry, I can't lie. It just says, this is what I found. Might be intermittent, but I found it. So the hard part, up until I think a year ago, was getting people to say, I can deploy AI in this way, even though we have a system that does it. And now with ChatGPT, it opened people's eyes to the application of AI within their businesses. And now we're seeing really hordes of companies showing up and saying, actually, we want to apply AI. You guys have all the patents in this field. You've been doing it the longest. What can you do for me? Right. So sometimes the market has to catch up to the vision, and the vision was right back in 2012. It just took a long time for people to believe that a machine can do this kind of work.

Understood. Let me ask you this, Kevin. If I'm a small business owner, and obviously this is all new to me, where do you recommend people start with their learning about AI and how to apply it to their small business?

Yeah, so, so I think a good place, and I'm, I'm, you know, just started reading it, but Reid Hoffman's a friend and, and, and Reed just wrote, I'm going to look at this. It's called Impromptu Amplifying your humanity through AI. And basically it's a book that is half written by AI, mostly because there is example after example after example of, hey, I'm just going to ask at this, I'm going to ask at this. And so I think when you give people 300 pages of examples, they will walk away with, hey, I've got some ideas how this might apply.

One or two of those is going to resonate right away and say, let me run with that.

Right? And you can learn how to do it and you learn how to log in all these things. Right? And so again, amplifying your intelligence is what this about. I want to take, I want to take Henry and make 100 Henry's, because 100 Henry's is at least 100 times more productive than 1 Henry. Same with Kevin, same with Jim. Same with sue, everybody.

Right, Understood. Excellent, Excellent. All right, obviously, as we mentioned, speaking engagements is one of the big things that you do to, to share and to communicate. And so if somebody's interested in that, just about any of the speaking agencies they can go to and find you there.

Yes, all of them. You say, type Kevin, you know, Google Kevin Sarace and speaking. I don't know, it'll be page after page after, you know, of, of speaker agencies. And. And I'm booked all the time. I love entertaining audiences. I, you know, again, trying to bring AI down to a level that people go, oh, I can put this to work as I leave here, right as I'm leaving this conference. I've got some ideas, I've got some places to go, I've got some links, et cetera. So that's what we want to do is make everyone more productive, because more productive means you're going to have more top line and more bottom line, and you're going to be happier.

Absolutely, absolutely. All right. Obviously we talked about impromptu. I think there was another book that you might recommend for us, I think, for startups.

And it doesn't always have to be a tech startup, but I like also Reid Hoffman's Blitzscaling that was written with my friend Chris Ye, who's fantastic. And that's just taking. When we've looked at some startups that went from 0 to 100 million in two years, whatever, and you start to model what it is that they did correctly, how did they advertise, how did they garner the social media attention, et cetera. There's a lot to learn in there about growing a business. Even if you don't want to grow from 0 to 100 million, maybe you want to grow from 0 to 1 million. You go, what can I take from this book? There's a hundred ideas and maybe two of them are useful to you. But. But it's. It's just worth learning. What did these companies do to go 0 to 100? Some of them went 0 to a billion in a few years. You go, wow, how do you go from zero to a billion? It's very fascinating.

And that's by Hoffman is one of the authors his podcast. I think he's still doing episodes, but really in podcast as well.

Yes, it is.

Listen to. Excellent.

As good as yours, Henry.

Well, now that you've been on here, of course.

Yeah, there you go.

All right, let's wrap it up. Kevin, what's one thing you want us to take away from this conversation we've had about, as a small business owner, beginning to apply AI in my business?

Well, look, AI is Not going to kill you. None of this AI is going to kill anyone right now. Look, don't hook the darn thing up to our nuclear stockpile, okay? Let's be clear. Don't do stupid things. But all it does is spit out text or images, okay? So everybody get a hold of themselves because everyone's reading all these headlines. You know, Elon Musk says AI is going to take over the world. It's over. It's okay. It's not going to do that. The second thing is, is if we all want to grow the way we want to grow, we need twice as many workers as we have now. And the worker pool is getting smaller, not larger. So we had better amplify our intelligence. And number three, and probably number one, your competition right down the street is already doing this. They're leveraging it in some way. So Jimmy sitting at the front desk now there's two. Jimmy's there, you know, Sue, Erica, you know, writing, writing content. While now Erica's Times five. And they didn't have to hire those people. They're paying almost nothing for it. So if you don't do it, they're going to have better content, better advertising, better responses to their customers, you know, better social media ads. How are you going to beat that? So you got to get AI working for you. It's a great time. It's the best time to be alive. When you think of it, we've never had access to this kind of.

So many accessible technologies and things that we can apply in our, in our business that are, they're affordable to us.

Yes, yes. Literally dollars a week.

Excellent. Fantastic stuff, Kevin. Again, if we want to find out more about you, we can search your name, which is Kevin Serrace. S U R a c e. LinkedIn. I'm assuming as well we can find you there.

LinkedIn. Absolutely.

Excellent. All right, I want to thank you for taking the time for sharing all this knowledge. Inspirational, of course, and so educational for us as small business owners. Thanks for being with me today.

Thanks, Henry. Really enjoyed it.

This is Henry Lopez and thanks for joining me on this episode of the Howard Business. My guest again today was Kevin Serrace. I release new episodes every Monday morning. You can find the show anywhere. You listen to podcasts, including on my website, thehowofbusiness.com thanks again for listening. Thank you for listening to the how of Business.

For more information about our coaching programs, online courses, show notes, pages, links and other resources, please visit TheHowOfBusiness.com.

This episode is brought to you by Progressive Insurance do you ever think about switching insurance companies to see if you could save some cash? Progressive makes it easy to see if you could save when you bundle your home and auto policies. Try it@progressive.com Progressive Casualty Insurance Company and affiliates. Potential savings will vary. Not available in all states.