2335 - Smart Ways to Leverage Technology and Keep Real Connections with Kevin Surace

0:04Well, hey, Thoughtful Leader. Welcome to the Thoughtful Entrepreneur. On this show, I interview purpose-driven founders, experts, and change makers who are making a real impact in the world. Want to be a guest? Simple. Go to upmyinfluence.com and click podcast. I'd love to promote your story to our audience of over 100 ,000 and across over 2 ,300 daily episodes and counting. Want to get booked on other great shows like this? Simple. Go to podverified.com and get your free guest authority score. Thousands of hosts are looking for high quality guests and Podverified connects you the right way. Now let's dive into today's episode.

0:58With us right now, it's Kevin Serais. Kevin, you are a keynote speaker. You're a futurist. You're the founder of multiple companies. Your personal website is kevinserace.com, which we have all linked up to our friend that's listening to us. Kevin, it's great to have you. Thanks for having me, Josh. Absolutely. Listen, before we get going and talking about your superpowers and talking about the future, as someone who's really focused on the future, you also have a keen interest in the past, particularly around something that you and I started talking about before we recorded and like oh man we could talk about this a long time but that is kind of old school AM radios do you mind kind of sharing your passion with that because I absolutely love it so you know look I grew up in upstate New York my father was an executive with GE Audio so audio electronics at the time was radios tape recorders clock radios all that stuff so this is the 60s the 70s into the 80s and um and GE was one of the largest brands that made long distance AM radios and so before the internet if you wanted to get news from LA overnight you could get with a sensitive radio a signal from KFI in LA or WLS in Chicago or WABC in New York or wherever you are in the country because those are clear channel stations um they still are today it still works today all my long distance AM radios here.

2:22I have about 30 of them all still work and can pick up those stations. And it was quite the hobby because you'd get up early at four or five, 6 AM to get that bounce off the ionosphere and get some station from thousands of miles away. And you'd log it. You could even write to them and say, this is what I heard at this moment in time. And they would send you a confirmation that clearly you heard our station, right? Yeah. I would call in. So I was in West Michigan, I would call into WGN, WLS for sure. I remember when I lived in Utah for a while, I could get KFI and I called into the Phil Hendry show quite a few times.

3:02But you know, when back then when you had limited entertainment options, you know, you just had what three channels on TV and then just your local broadcast, like you just didn't have as many options. The big talent was in these major markets like your drive time host your morning host i mean that was i mean they were serious like wls at the time i mean there was some serious talent on those stations that you know for some kid growing up i'm like wow i want to be like these guys and today they're podcasters it's mr beast that's you know you know because there's more money in that than on radio But at the time, if you were one of the drive time anchors on WLS, a WABC, a KFI, that was as big as it got.

3:46That was a huge job. So look, fun old hobby. Probably hardly anyone understands it, but there was great excitement when you'd work the dial and you'd angle the radio and you were hoping. And there it was at midnight, you happened to pick up X from, you know, X thousands of miles away. And you'd write it down in your log and say, look what I got. So, you know, it was an entertaining hobby when there was nothing on TV at that time. Right. So really. Yeah, I love it. Well, Kevin, so we've talked about the past. Let's talk about the future. And I want to start off by sharing something that I saw today.

4:22Ray Kurzweil, a noted futurist. I'm sure you hear his name from time to time. I know, Ray. Yeah. Yeah. Yeah, so he's standing firm on AGI, and this is going to cross a couple of areas that you're, you know, well, you're adept at, but he's still standing firm at AGI. We should be seeing that in 2029. What's your take? Yeah, yeah. So my take is, sure, for sure, but do we care? I mean, I think the world has been, oh, what happens when we cross this boundary? Okay, let's just go back. Let me really tantalize people here. In around 1988 or 1989, Excel came out. We solved math, and it was, you could argue, AGI for math, meaning that Excel obviously has done math better than any of us can do math for 30 to 40 years now, right?

5:15Period. That's not arguable. Like, I can't do long division or calculus or anything like that the way Excel can. Period. Right? And so at that point, humans kind of gave up doing math. We do very little math. And nobody, no one would ask you to do math by hand and pay you to do so. It would be a ridiculous thing. So we did that. And now we've got language. And of course, language is a much broader problem. It took us 40 more years to solve the language problem. And then once we did that, people, oh, AGI, what happens when it's always smarter than us and it's smarter in every field? and I'll hold it.

5:53You can go to GBD5 today and ask almost any question. Josh, I don't know you that well. You're a very smart guy. But no matter what question you ask it, it's probably already smarter than you. So the fact that it'll be smarter in a few more fields, and it will be smarter in taking the bar, and it'll be smarter in whatever, passing SATs, blah, blah, blah. It's an interesting point. But it's already smarter than almost any human almost all the time. And the average human's an IQ of 100, and I can tell you, you know, GPT-5 in most subjects has an IQ of, you know, 180 or some crazy number, right? It's ridiculous.

6:32So I have just taken the point that that's not going to change people's lives. We have this fantasy point where we say it is all-encompassing and all knowledge. It already knows more than a trillion phrases that we've put on the web in the English language, right? Trillion phrases. Does it need to know$2 trillion? I mean, really? Do we really care? So other people are really all over it. I could care less. It's already smarter than us, and almost everything we're ever going to ask it to do, get used to it. And do you look at how AI has been evolving, let's say, over the past couple of years? I think that we could point at a few things and say, wow, we have been really evolving quickly.

7:13And I got to tell you that there are other areas of AI where I'm like, man, I don't know. Like, for example, you know, OpenAI's GPT-5, I think, was a step backwards in my personal experience. And I know I'm not alone in that. How do you think we're doing right now? Yeah. So, look, GPT-5 made great advances in certain areas, including coding and a few other areas. I have, and actually in sort of standardized testing versus GPT-4 or 4.0, it is doing better generally most of the time. That said, right, don't forget that all of these LLMs, or multimodals, right, post-learning, have millions of rules placed on them by humans that are trying to make up for the fact that they were trained in an unsupervised fashion, meaning nobody told them what was fact from fiction.

8:06So if it reads a book and it says, you know, someone killed someone, that might as well be fact for the darn thing until a human says, I'm sorry, that was actually fiction, right? You read it and it appeared in the New York Times, but it was a review of a book. So you can't take it the same as news. So a lot of that's going on as well. And the other thing that's going on is these models are starting to asymptote a little bit, meaning that if you have a trillion phrases, having two trillion phrases isn't twice as good. it's maybe not even as good right I mean you can overload the thing in several ways and now we're starting to take in much more information that was generated by AI so it's kind of feeding on itself and there's a feedback loop there people are making real advances in reinforcement learning in a number of areas including doing some more interesting things with these models but sort of the net result is again whether I'm using that or Gemini or Claude or Reblexity or Grok, all the top models for almost anything I do.

9:06And some are better at certain things than others. Like I can upload certain documents to some and I can't upload them to others. I can do like Gemini is great for video evaluation. I can give it a video and say, what does this mean and what happened in it? And it's fabulous. I mean, almost 100 % accurate. And I can't even upload video to some other models, right? So everything's different. But if you know out of the top five models, what it is you want to accomplish, you're going to go to that model, get that accomplished, and it's world-class work. World-class. Like a friend of mine used to work at Amazon, and he says, you know, it wasn't that many.

9:39It was just three years ago we would have paid$150 ,000 for this report on the, for example, what's the overlap of taxis versus Uber versus Lyft over five years in Brazil? oh man that would have taken you know real money and real effort and then and then plot the forecast you know blah blah blah I can have that in a minute from the right model with sources for nothing for zero dollars this is unbelievable right so I I am amazed every day and if you ask correctly with enough information you almost never get a hallucination you you know you can force hallucinations are quite easy, right? But if you don't force a hallucination, you're really not getting it, especially if you ask for sources, and you say it must be backed up.

10:31And again, you know, my real quickly, my, my sort of prompts are often two to three paragraphs, not three words. And I see people write three words and go, well, I don't like what I got back. Well, good, you gave it junk. And so it gave you garbage in, garbage out. That's right. You gave it garbage, and it gave you garbage. Now write two paragraphs about how you really feel about this topic and then say, rephrase this into ways that a reporter would like, rephrase this into ways that the public would appreciate, right? Because I'm not an English major. So the pros of any of these LLMs beats my pros.

11:07It's sentence structures far better than mine and should be, right? And should be. But it's opinion. I don't want its opinion. I want my opinion, right? So here's my opinion. I may be right or wrong, but here's my opinion. So form your sentences and replies and structures aligned with sort of that opinion to answer this quest, right? Yeah. And that gets to something else that I've seen you talk about on LinkedIn, which by the way, Kevin, you're a great follow on LinkedIn to our friends that are listening. You can go to Kevin's website, kevinsterace.com. We have that directly linked. And then click to follow you on LinkedIn because you do share some really good thoughts around AI.

11:45One of them in particular is I think other news that we've heard recently about studies that are showing that, you know, those that are using a lot of AI, particularly students, are getting dumber. And you had a really great take on that. I did. So I'll tell you the take is this. You will get dumber if you use it as a cheat. Just like, Josh, if instead of me thinking about anything, I just texted you or emailed you and said, can you answer this question for me? Can you answer this question? And I might do that if it's about podcasting, right? Because you've got this podcast, you've done 2400 episodes or so I might act.

12:22But if I start asking you about everything that I'm never thinking about anything, right? So, so on the other hand, people are getting smarter, if they use it to produce five or 10 times more content than they could have by first opining on that content, right? So So I'd like to create, for example, a blog post. I need it to be 600 words. I want you to learn from this website. Here's my opinion on what I want to see in this blog post. Here's some data that I've looked up that I think is interesting. And here's the sort of the net result that I'm trying to get, right? So I've thought about that.

13:00I put in two or three paragraphs, a lot of thought. It writes the blog post. Then I have to read and edit the whole thing, okay? Now, I was able to create that content easily 20 times faster than I could have by myself. Yeah, I did all the critical thinking part anyway. What used to bog me down was not the critical thinking. I knew what I thought about the subject. I still do. In fact, better today than three years ago because of AI. But what I couldn't do is put it in the right sentence structure and prose in a minute. I could do that in an hour or two, but not in a minute. So I knew what to think.

13:34I knew what I was thinking. I know my opinion. Here is my opinion. And in fact, the two paragraphs aren't formatted well, but it doesn't matter. I get my opinion across and I said, now format the output like this. Last example, I get requests for press quotes about five times a day and each of them have four or five questions for me. That's 25 questions a day. When I used to answer them one by one, it was over an hour per press response, because I had to think about the three or four sentences I would give them back that are quotable. Again, I'm not an English major. So now what I say is, here's my opinion on this topic.

14:16Yes, cybersecurity, this is going to happen. This is what's happening with deep fakes and vishing. And this is the response to it. And this is what's going to stop it. Boom, boom, boom, boom, I write my two, three paragraphs. And now I say, given these questions from this reporter, take my opinion and format the prose, format the sentences in ways that are quotable for the reportable. Boom. I have it in a minute or 30 seconds. Write my opinion, format it in a way that's appropriate for a report. I'm done in five minutes. It's still my opinion, but I couldn't have written those sentences as well.

14:49I want to make sure we only have a few minutes left. Kevin, share a little bit about your work, who you work with, if folks engage with you, or what would be the next step for a variety of folks that might be listening to our conversation right now? Sure. There's a form on my website, at least for keynote speaking and consulting and stuff like that. Look, I've been fortunate. I've built a lot of companies. I probably crushed a few companies and certainly did and have 95 patents, including in AI assistance, which kind of invented some years ago. But what I do is I, number one, keynote speak on three topics, the joy success cycle, my upcoming book, and Silicon Valley disruptive innovation, and today, AI and what it means to your work and your life, right?

15:38And with 50 concrete examples in 50 minutes, right? I just fly. so um so that's the biggest request and then obviously i work with companies on rolling out ai in an appropriate way setting up training having interactions uh uh with groups of people right because number one problem in companies today is not that they have haven't heard of ai but if you're not a tech company believe it or not because i do this around the country in the surveys about um 80 of people aren't doing anything 20 or 18 actually are kind of playing 2 % are doing real work. And if you're the CEO and you say, I have said for the last year, I want everyone to be in AI first, but they just don't know how to approach it.

16:20We have to make it more approachable. We have to show them what gains they're going to have. We have to make people clear that they don't need to be concerned about this or scared. Like if I'm not using AI two or three times an hour, I failed. This is just me, right? And I'm smarter because of it, because I'm not using it as a cheat. I'm having it do the stuff that I'm not really good at anyway and, and leveraging my opinion. Kevin, your website, Kevincerace.com. We have that linked up to our friend that's listening to our conversation. Kevin, it's been a great conversation. I wish we had three.

16:54I could talk on this for a couple of hours pretty easily. I appreciate all your insights. And I don't know if the AM radio still work today or what they're broadcasting. They do. They do. They still work. And the clear channels still broadcast overnight. So it all still works. Thank you so much for joining us. Thanks, Josh.

17:20If you've enjoyed this conversation, I'd love to invite you to share your message with the world at podverified.com, where thousands of podcast hosts and thoughtful guests are connecting the right way. Podverified helps match you with the right stages for your level of authority. No more wasting time on brand new shows that just aren't a fit. Your guest authority score is free and so are your first two podcast matches. Come hang out with me there. I'd love to learn more about your business, support your mission, and help promote you far and wide. Just head to www.podverified.com to get started.

18:03And of course, if you'd like to be a guest right here on The Thoughtful Entrepreneur, just go to upmyinfluence.com and click on podcast. Thanks for being a part of this movement. We rise by lifting others.