Latest / The Wilmac Wire / Kevin Surace | What’s Changed in AI, the Power Players, & Enabling Rochester to Build

Transcript

Kevin Surace | What’s Changed in AI, the Power Players, & Enabling Rochester to Build · The Wilmac Wire

• speaker-0: Good morning, good afternoon, good evening, welcome or welcome back to the Will MacWire. Thank you so much for tuning in. We are bringing back one of our most popular guests from season two in today's episode of the Will MacWire. And and given how fast the AI landscape has moved since we last talked, a little bit over a year ago, it felt like a a no brainer to have him back. And we'll probably have him on again in the future. ⁓ we recorded this episode in person at Rock Fox Studios as well. That first one was virtual, ⁓ so it was even a more fruitful conversation. Today you can expect to hear from Kevin Sorace, AI pioneer, inventor, and one of the most recognized recognized voices on disruptive innovation. If you haven't haven't listened to his first episode, you can find it in season two, episode six, which aired in March 2025. Like I said, the conversation was very different back then, which is just incredible to be living this history. So here's what you can expect out of today's conversation with Kevin. We discussed what has changed with it within AI since we last spoke. We discussed your generative AI, your agentic AI, recursive AI, the model wars that are going that are going on. Kevin, he really broke it down and and Steve got super into it. We talked about the power players, your open AIs, anthropics, perplexity, grok. And you'll hear Kevin call me out about Gemini, but where things stand and what Kevin is actually using in terms of AI. We also discussed the fear narrative around AI and why so many companies are still paralyzed and not moving on it. And we also discussed Kevin's work here in Rochester and what he's doing, which one of those things being a new book that he has dropping September 1st that we would that we like to chat about. But I will let him tell you about it. Kevin, thanks for for joining us. We're excited to have you back again in the future for our listeners. This is a great, really educational ⁓ episode on AI. So let's just get into it. Welcome to the Will MacWire. I'm Emily Miller, joined by my co-host, Steve McDonnell. On this show, we talk with the drivers, the leaders, innovators, and decision makers pushing customer interaction technology and AI forward. And because we're proud to call Western New York home, we also spotlight the drivers in our own community, the people and organizations shaping growth and innovation right here. Let's get into it. Joining Us here in person at Rock Fox Studios in Rochester, New York. He's originally from the Syracuse area and an RIT alum. He's now one of the most recognized voices on AI and disruptive innovation in the world. He's known as the father of the virtual assistant. His work at General Magic in the late 90s laid the foundation for Siri, Alexa, and Google Assistant. He's the holder of 94 worldwide patents across AI, cybersecurity, sustainable sustainable building tech, and more. He's currently serving as chairman and CTO of Advanced.ai, pioneering generative AI for software quality assurance. I'm almost there. He's an Inc. magazine entrepreneur of the year, CNBC Innovator of the Decade, World Economic Forum Tech Pioneer and Inductee of the RIT Innovation Hall of Fame. And he's got a new book dropping September 1st. Welcome back to the Will Macwire, Kevin Sarate. speaker-1: Thanks for having me back. speaker-0: We're we're glad to have you back, Kevin. This time you're you're in person in Rochester, I think last time we spoke you were in Hawaii. I was jealous. I was painfully pale. Yes, I did. speaker-1: Yes. You were jealous. He wouldn't let you. speaker-0: It's funny 'cause we were just at a conference last week and we talked to quite a few folks from Hawaii and we were like, ⁓ it's gotta be an on site visit. Yeah. speaker-1: Absolutely. It's where everybody wants to go. speaker-0: For sure. ⁓ well, yeah, and for our listeners, like like I mentioned, Kevin's a return guest on the Will MacWire, and his original episode aired in season two. It was episode six, March sixth, twenty twenty five. So we're a little over a year past, but a lot has changed since then. So we've got a a good, a bunch of conversational topics to chat about today. But before we get into that, we always like to do a brief icebreaker. It's called This or That. I have three questions for you and I'm gonna pose you two options for each question and then you basically choose and explain why. So very easy peasy. Okay, first one. Rochester in the summer or Rochester in the fall? Okay, why? speaker-1: It's gorgeous. Gorgeous. Look, spring, summer, fall are great. I pi I mean, it's a little bit of a secret. We we laugh about this sometimes with others in Rochester, right? It's like don't tell everyone that it's actually a really nice place to live most of the year. Yep. If you can deal with the January, February, you you're you're gonna love it compared to almost anywhere else in the country. Right. So ⁓ we feel very lucky. But fall is magical. Mm-hmm. It is. As we know. speaker-0: Yeah, I was about to say the past few mornings it's been like low seventies and a little cloudy and I'm like, it feels like fall and then last week, ⁓ late June, early July, it was nineties. It was rough. speaker-1: Half the country had ninety. But we're back, it was sixty eight this morning, having breakfast outside. It was lovely. speaker-0: Yeah, it was lovely. All right, first one done. Next one, speaking at a conference or writing a book. Yeah. speaker-1: ⁓ speaking at a conference. Yes. This is my first book. We'll talk about it later. And my last book. speaker-0: gonna be one of my questions for you. I I figure, I mean, writing a book's it's a process. speaker-1: Yes, it's ⁓ it it it is a process and you have to decide why you want to do it. And mo most book writers, ⁓ with all due respect, right? Most book writers that I know go, Well, I've written this thing and it's gonna be a New York Times bestseller because it blah, blah, blah, blah, blah. They're hoping. They're hoping to make money on it. They're hoping that is not why you write a book. Three hundred I didn't know this. Three hundred thousand new books hit a month in the US. Three hundred thousand a month. The chance of you being a New York Times bestseller is one in three hundred thousand or ten in three hundred thousand. It's not gonna happen, right? Unless you're a real name name brand. Yeah. ⁓ I do it because my keynote talks. So this is a ⁓ some of my keynote talks are on ⁓ the intersection of humans in the workplace and AI. Yeah. And so many humans are having a hard time with AI. No, you're not. You're used to it. We're in the tech world. We're okay. But most of what the surveys are very clear. People are very concerned about their jobs, about their future, about their kids, about And so and then they lose their sense of purpose. They What's my sense of purpose? Because AI took it away and where do I and that's what this book really deals with. It brings joy back to what you do. speaker-0: I love it. Okay, and last one, ⁓ potentially a little bit tougher and more than two options. ⁓ I had open AI or anthropic or perplexity or croc. Do you have a preferred AI model, thesis? speaker-1: So I use all of them and Gemini. Figured. You left Gemini Google. ⁓ Never write off Google. They have more money and more data than virtually anyone except for Microsoft. We shouldn't write off Microsoft. They're struggle. They have their own ⁓ problems, but don't ⁓ never write them off, like Apple, right? They're late to the game, but they usually win. So ⁓ look, o over the last year we've seen the world shift from open AI is way in the lead. speaker-0: Yeah, left ⁓ and I. Yeah. speaker-1: To then Gemini for six months was way in the lead on many metrics. And then lately Claude is way in the lead. And and and Claude is is ⁓ so I I use ⁓ today Fable because Fable's live again. I came back, so I'm using Fable. But but ⁓ you know they're in the lead because they've done more recursive or reinforcement learning or reinforcement programming than others have. Yeah. And so everybody's in that game. And for those listening, what that means is I have my own coding engine improving my own foundational model and doing that over and over again. So now we're seeing releases come every few weeks. And I if I believe what what Anthropic said, ⁓ possibly eighty percent of the improvements in ⁓ in say fable or mythos were actually totally written by itself. It wrote itself and improved itself. Right. So we're getting into this self-improvement category. We knew that that would happen. We're doing it with robots, of course. ⁓ reinforcement learning, of course, instead, ⁓ instead of the old way to do it. So I think we're likely to see this thing really move quickly until we see an asymptote. And there will be an asymptote because there's only so much material to learn on. And we've learned on virtually all of the 10 trillion tokens there are out there, and we're gonna run out, right? Yeah. So there'd probably some asymptoting of this. type of a a a transformer. ⁓ but speaker-0: You don't think one will start running away? You think it'll be turn into a hopeful I think it's turns into a commodity model. speaker-1: It's commodity. Yeah, it and and they don't want to hear. ⁓ speaker-0: Especially anthropic and open AI. ⁓ speaker-1: ⁓ they have cash flow out. Yeah. Not as much in. I mean, you know, you have to raise ⁓ a lot. ⁓ when you're losing ten or twenty billion a year, it's a lot of money. But by the way, the the numbers we were talking about today, even you know, SpaceX or X AI, all of that conglomeration that went public at peaked a a couple of trillion or so. You know, these numbers are extraordinary. If we were here ten years ago and I even used something with a T on it, you would have laughed me out of the room. And and and now we've got ⁓ some people predicting that one of these foundational companies could be worth ten or twenty or thirty trillion with a T. I don't I don't even know how to judge that. But God bless. Well go go for it. speaker-0: Did Dario or Sam Altman saying if they don't get to a trillion soon, they can't function as a company? speaker-1: Well, that's right. Yeah, yeah, yeah, yeah. And and and and the other thing that's happening is of course revenue's up, right? All these guys are doing, you know, twenty, thirty billion, which is amazing because you're really they've only released the product in the last three years or so. And in three years, right, if your company and maybe your company did in three years go from from from zero to thirty billion, but no companies have ever done that in the history of ever, right? And these guys have done that. So ⁓ these are clearly products that are wanted, but but there's been some pushback. Just in the last couple of months, you start to see these companies who all push their employees, use, use, use, use, use. We're going to measure you on usage. Then they got the bill. And all of a sudden they had a 120 million dollar token bill from Anthropic, and they go, whoa, we could have hired engineers for that. And it turns out that at some point, your engineers, each dev, let's say, and let's just say take coding, you know, each coder, each developer. Is using two or three times their salary in tokens. And you go, I thought they're supposed to save money. And and and what happened, a couple things happened. One, people finally figured out how to use it for everything. And and some people anyway. And the the second thing is that ⁓ all of the the frontier ⁓ models stopped doing a twenty dollar a month eat all you can model. Right. And they go, We're gonna go broke doing this. So they started charging by the token. And all of a sudden you had a hundred thousand dollar a month bill. For a single person instead of a twenty dollar a month bill. And everybody choked. So, you know, water finds its own level, right? speaker-0: That's going on right now. I feel like that in the enterprise companies are kinda pulling back, which should be interesting and and a lot of people don't necessarily say as you were saying, people are learning. Sometimes you don't need the power of speaker-1: Whoa, pull back. Well, that's right. You know, look, I would say most office workers could use the power of an open source free model running on ⁓ you know, a small set of GPUs internally, right? Or or at Amazon or whatever. So ⁓ so the idea that you're gonna p spend this much for tokens ⁓ is probably gonna be limited to really complex things. Now now look, if you really want your coders to be eighty to ninety percent more productive than they were two years ago, you're using the latest cloud code. the latest model that they can put you on, right? No question. But ⁓ but other than that, you know, an office worker that's basically doing spell check should not be should not be spending those tokens. ⁓ so this again, this is all working itself out live as we sit here today. And if we come back a year from now, we'll say, ⁓ that one was solved. Yeah. You know, and now agents are twenty twenty five was supposed to be the year of agents. It was not. Twenty twenty six is the year of agents and everyone's running into all kinds of Concerns, security concerns and which we can talk about a lot. ⁓ that's there's my token ring on my finger here. ⁓ so lots lots going on in that space, but twenty twenty six has become a year of agents in some areas of enterprise. speaker-0: Yeah, wonderful. Well, if anyone from Google is listening, I apologize. And I hope you can come join us on the podcast so we can talk about you and I. ⁓ appreciate you you humoring us in doing that icebreaker, Kevin. We're gonna just dive a little bit briefly into your career journey since we hit on it a lot in ⁓ our previous episode. ⁓ but I guess w you were working on virtual assistants in the in the late 90s. So we when we talked about it, when most people didn't even know what that meant. speaker-1: Some of us in this room weren't alive when I was working. speaker-0: Yeah, virtual assistance. But what what pulled you into that world? speaker-1: Yeah, well, ⁓ like all I think all good products, ⁓ and I think the same is true with your company, you're looking to fill a pain point that people have. Yes. Don't invent things just ⁓ and this happens in Silicon Valley all the time. Like I said, there's five thousand or so ⁓ funded AI companies, of which four thousand nine hundred will be dead, right? And they all invented something interesting or added some value. They didn't solve a pain point. You gotta go solve a real pain point that people have money to pay for, right? And when you do that, people pay you and you grow a big company. And when you don't do that, it's cool, it's cool technology, but nobody pays you. So we were trying to solve a pain point. ⁓ cell phones just became relatively ubiquitous, not smartphones. They were Motorola flip phones and things like that, right? Mid nineties. The internet just hit nineteen ninety five. The world wide web as we knew it then, that just hit in nineteen ninety five. So people moved from AOL and CompuServe to hey, what's this web thing? They were still dialing up on AOL, but but at work you had the web now. And at work, something else happened. You had now email that worked across enterprises for the first time in the late 90s. You actually had email. Right. Before that, you might have had it internally only, but you didn't have email worldwide. All of a sudden you had email, you had calendaring, you had contacts, and you had it all and it was internet connected by ninety-seven or so. That's a game changer. But here was the problem. No one had a computer at home. Right. You didn't have a smartphone. So you left work. And you had no contacts, no calendar, literally no calendar. Right. and ⁓ and and and no and no email. You literally stopped. Yeah. And so when people were in the car, they wanted access to this stuff. Where am I supposed to go? What's the address? Who am I con you know, whatever? And they had no ability to do that. Right. And screens were very expensive and you couldn't look at a screen. So we said, well, if everybody could have their own secretary, they would have called their secretary, which is what Yeah. C level execs did. And they they called Mary and Mary would say, Yes, next on your calendar is this and you need to be there at three PM and here's your going to be meeting. And here I go, we'll just invent one of those that's virtual that will be, you know, eighty nine dollars a month instead of three hundred dollars a day or whatever the, you know, the price was. And so that's what we did. We solved a pain point and we had millions of people on the first system, three, three and a half million people or so. Portico, My Talk, Magic Talk. It also got licensed to GM, and GM had the virtual advisor that would read your email in your car. Stuff like this was part of the On Star Virtual Advisor. So we solved a real pain point that people had. And I don't know, had 15 or 20 or 30 patents out of all of that that eventually got licensed to intellectual ventures. Nathan Mirvold, who then licensed them to Apple and everybody else. speaker-0: Did you how'd you start the company? Did you speaker-1: It was a public company. It spun out of Apple. ⁓ literally General Magic was ⁓ the whole team was part of Apple, Apple computer at the time was part of Apple, and spun out to do stuff that was so advanced that they wouldn't do it at Apple. Yeah. And it was backed by Apple. Okay. And Apple ultimately had to license the stuff back because they didn't get a free license to it. Yeah. We invented ⁓ three big things there the w at the time called the PDA, but it looked an awful lot close to an iPhone. And the drawings look like an iPhone. So we were the first to have 10 years before the iPhone. We ⁓ invented agents and a and a language for them called telescript, ⁓ so that agents could act autonomously. And very much like AI agents today, they weren't as smart. Yep. ⁓ they were programmed, they were probably a little closer to RPA, if people know what that is than agents of today, but it was called agents in a genet technology. So we invented the concept for that, and we invented The AI virtual assistant all came out of that company. Wow. and we spent hundreds of millions of dollars developing all this cool stuff and putting it out there, some of it licensed, some of it as our own products. and it was it was an a amazing, amazing deal, actually. And it was a great, great place to work because of the inventions. And and so lots of people came, like Tony Fidel, you know, came out of General Magic to go on to create the iPod and The iPad and the iPhone. ⁓ later at Apple, right? Lots of people came out of General Magic. ⁓ you know, the inventor, the creator of the Android operating system was the General Magic, then left and went ⁓ started his own company to to create ⁓ mobile operating system, then sold that and went to Google to do Android. ⁓ eBay came out of there, ⁓ Pierre ⁓ Omedier. ⁓ developed eBay ⁓ under his desk at General Magic and then left. So it had all these amazing people and all this all these huge startups came out of what was General Magic and all this big technology. So I feel like really humbled to have led the team that invented ⁓ the AI system. speaker-0: Yeah, that's amazing. So it came out of Apple, so you didn't need to go to a V C or anything that to f speaker-1: There were no VCs. It was ⁓ it was ⁓ funded by Apple and about eleven other partners, people like Panasonic and Mitsubishi and all partners around the Motorola. Everybody put money in because then they could get a license to certain things and access to certain technology. It was a technology powerhouse. It was like the Bell Labs of Silicon Valley. And ⁓ and then we did go to the public market, then we went public, and then we went to the public markets for money as well. So there were kind of ⁓ mezzanine financers and things like that, but not traditional VC. There's no traditional VC in there. ⁓ and it was well funded. I think we spent about a hundred million developing ⁓ the first virtual assistant. and I helped raise that money and everybody couldn't believe that you could do it. And then the first time I remember being on stage and we unveiled it, and I'm talking to her. Her name was Mary, and I'm interrupting her, and I'm actually rude to her, and she would stop if I interrupted her because we made it interruptible. Yeah. And people went. WTF. Like they could they literally the the jaws dropped. They had never heard anyone talk to they knew it was a computer, but talk to a computer. And she was very friendly. Yeah. And she could say thousands and thousands and thousands of things back. But not unlimited. It was we put it together with a set of logic because we had AI, but we didn't have the level of Gen AI that we have today, right? We had hidden Markov models, we didn't have Gen AI. I didn't have a transformer. Yeah. And lastly, the fun thing about that is we had to build our own operations center. There's no such thing as the cloud at the time. There is no AWS. There there's nothing. There's no Azure. So we built our own op center. ⁓ I got to build an op center. ⁓ we built the servers. They were very specialized servers. We had to build them in-house. Sunnyvale. Yeah. Built it in Sunnyvale. I don't know, we had 3,000 servers in there or something. There's a video of me if you look online, like running through. speaker-0: Where yeah, it was all in Silicon Valley. speaker-1: the the server farm basically. Yeah. Yeah. You know, we had to put in all of these ⁓ Libra chiller units to cool the whole thing down. Every watt that comes in has to come back out as it turns out. So ⁓ it's ⁓ it was a a great great project. speaker-0: What you mentioned you solved for something practical. There was a problem, then you went for it. We were just talking about these companies getting valued at trillions of dollars. Yeah. And what we kind of see now with venture talk of Silicon Valley, venture capitalism and even some PE at this point, they're only going for the unicorn. speaker-1: That's right. ⁓ speaker-0: Before it was like hey, you have something practical. We know you can make money on this. It may not gut to a trillion. Now it's like it better be going public in three, four years. Right. So that leaves a lot of companies speaker-1: That's right. ⁓ either begging. Yeah. Yeah. I can I can answer that. No no question. You know, I I ⁓ know Kozla talked about this recently. So it used to be if you did a seed round ⁓ that was very small, it was under a million dollars and you'd be valued at five million or some number like that. Then you'd do your A round and it'd be a few million dollars and you may maybe you'd be valued at seven or eight million. I mean it was numbers like that or ten or okay. Well now the whole thing is If this is going to be hot and it's really going to go to a trillion or a hundred billion, I don't care what the valuation is. That's the attitude of VCs. Like if you come in and say, I have no product, but I've got this team and we're going to be worth 100 billion and our seed round is at $1 billion and we're raising 50 million. Okay. A Vinode may go, okay, fine. Well I just want a piece of it. Cause it because when you run the math, it doesn't matter. If the thing goes to a hundred billion or a trillion, you're gonna pay back your entire fund. Yeah. So it's a moot point whether the thing is worth ⁓ 50 million or 100 million or half a billion or it that doesn't really matter anymore. ⁓ so that's the new play. Now not every VC can play that. That's A16Z and it's Kozla and it's the large Sierra, etc. Right. Sequoia. So it's the larger funds. But that's the that's the play they're playing and they're willing to put a lot of money to work. Look A16Z's got I don't know, 60 billion under management. I can't remember, but rough numbers. ⁓ six hundred, eight hundred employees. That's a VC. All right. That's not your normal VC. For those of us who have ever called on Sand Hill Road, your normal VC had, you know, three or four partners and they had a hundred or two hundred million under management. And, you know, now we're talking numbers that are crazy, but they are returning on those funds. And and, you know, SpaceX for for, I mean, there's VCs that I know that. you know, put in ten million, twenty million and and it returned one point five billion. And you go it pa and the and the fund was only a couple hundred million. It returned the fund, you know, times eight X or six X or seven X. And you go, I we're done. Like this is fantastic. And even the carry on that, which is the piece that an individual venture capital person may keep, they keep twenty percent of what was over, you know, over and above the return, right? They get twenty percent of that. You know, twenty percent of several billion is a good number. And I know several individuals who made personally about a billion dollars on the SpaceX IPO. Personal. I'm not saying it's right or wrong. It just is, right? They they they played the game, but they were they were in the game at a time when one could do this. If you were in the game ten years before, twenty years before, thirty years before, couldn't do it, and maybe ten years from now, but there's this pocket, there's this window that It still can work out that way. Yeah. speaker-0: And SpaceX, to a lot of these guys' credit, I mean what was that, two thousand three tech? That was like ten years of complete struggle. People were still like, Yeah, this could work. It could speaker-1: ⁓ yeah, yeah, this thing's a twenty something. It was going out of business many, many times. It was a twenty-three year struggle. So they had to so VCs really risked most of the VCs who risked it were ⁓ already in Tesla and they were willing to follow along and say if anyone's gonna get rockets and make money on this. But that's at a time when the defense industry and NASA were not buying anything other than the biggest defense contractors. So the idea that you could start a new rocket company looked ludicrous. Ludicrous. And I've known Alan, although I haven't seen him lately. He's busy now. He's a trillionaire, so he doesn't talk to me anymore. But at the time, you know, he had a vision. He thought rockets could be fully reusable. He thought the way to do that was not with some government contract up front that you'd have to just get there. And then once you proved it, you'd get kilograms to ⁓ to orbit for a tenth the price of anyone in the world. That dream came true. But it took, you know, over ten years to make any of them work at all. It was ten years of billions of dollars down the tubes. So now but but here's what that taught VCs. It taught VCs that some of these risks are actually worth it. Right. They're actually worth that play. Because if you have the right leader and they're willing to walk through brick walls like the ones behind us and they're willing to walk through walls, once in a while one of them's gonna walk through that wall and return a trillion dollars and it it it was just it was worth risking, you know, fifty or a hundred million on that. speaker-0: Yeah, and even moving forward, I know the general public is probably has their opinions on Elon, but he's not a man I'd ever bet against. speaker-1: Well, that's the thing. I've I've I haven't bet with him any time is unfortunate. So ⁓ every time I looked at Tesla I said Overpriced. The numbers don't right, the the PE numbers don't work, right? The ratios don't work. And ultimately markets tend to go back to their PE ratios of ten X or twelve X or six X or whatever it is in someone's in and maybe twenty X in a good year. So, you know, eight hundred X never made any sense to me. I go, I I just I can't do it. And I was wrong every time. All I had to do is invest at any time. But I also said that about NVIDIA. And all I had to do was invest at any time in the last twenty years and it would have worked out just fine. ⁓ so you can't win ⁓ all. And it does show you that the traditional market metrics may not always play out in in the investor's psyche. Yeah. Because in the end, what's a price? A price is what anyone's willing to pay. And Americans are willing to pay a lot of money for an Elan company and they're willing to be at eight hundred X ⁓ P E ratio. I don't I don't care. So it's really fascinating. speaker-0: Yeah, the Nvidia thing's fascinating. That's also it was like ninety three. I read an article recently. Wha who was thinking of building chips back then? Right. And the capital that goes into it. And no one I mean ⁓ when Nvidia was popular three, four years ago, half the world didn't know what the heck it did. speaker-1: Well, So So ⁓ so being in Silicon Valley, we knew the company, right? And the the thing was, and I even bet against them a couple of times, here's the the thing was is they're in the GPU business and GPUs were needed ⁓ to put graphics on a screen. That that's the business they're in. Here's the background on that. So they had that, AMD had GPUs, Intel to a lesser extent. And what happened is as AI came along, the math to do AI, which is vector math, is similar enough. To the way a GPU processes vectors for the screen, basically math for the screen, right? And so engineers developing AI and gen AI and Transformers and deep learning said, Well, what silicon is out there that we can use? CPUs are dead slow. They don't process that math in the right way. ⁓ there's these GPUs that are putting graphics on screen. You know what? We can adapt what we do to that and it'll work pretty well. And so then that became the standard through CUDA, which is a software package that ⁓ software interface that that ⁓ that NVIDIA has, ⁓ became the standard way to do ⁓ training and inference because there was nothing else out there. Now there's TPUs coming, there's other stuff coming. But so he kind of you know, NVIDIA fell into this. They didn't come out with the GPU and say this is for AI. No. They said this is for graphics and video games and those kinds of things. And then they realized over in this corner, there's all these weird people. speaker-0: Games and speaker-1: Doing this odds of deep learning or something, and they're using our chips. And they didn't pay much attention because they used 10 chips one year. Then they used 10,000. And then they used a million. They hold it, that is the market, the heck with graphics, right? So they f they fell into it and they smartly. This is a he's a great manager, yeah, great CEO. But no one thought that that company would be worth three or four trillion, which is worth more than the entire rest of the semiconductor industry combined. All of it. speaker-0: And they see th I mean they're printing money, they're prof very profitable compared to speaker-1: A lot of But they were never designed to do this kind of volume and they're sold out for the next year and a half and you know it's a it's a it's a it's a good place to be. May may we all be so lucky. speaker-0: It was I th I think it was last week I was reading somewhere that they're thinking about jump jumping into the foundation game. Foundation model. Yeah. And we maybe with an open model. speaker-1: With an open model that really yeah, they may need to do that because some of these foundation models are going to ⁓ build underpinnings that don't just use CUDA. And once you do that, so there's an open source version of CUDA that AMD is promoting and ⁓ the AMD GPUs are substantially less expensive. And that's number one. And number two, ⁓ now you got TPUs coming and and ⁓ Amazon's got that, Google's got it, and they actually process at a lower cost, basically. ⁓ Than a GPU, because GPU wasn't really meant to do this. So if these foundational models move away, the frontier models move away and say, ⁓ yeah, we support CUDA, but we will also support this, well, then the wheels could come off the bus for NVIDIA. And so they may need their own model to say, Yeah, but ours is better in these ways and it only runs on on CUDA. Interesting. This is a complicated I know we got very technical for probably a non technical. speaker-0: Yeah. I was just to say, yeah. speaker-1: They're rolling their eyes. No. All right. Let's let's leave the technical realm. speaker-0: No, I was just gonna bring us back to, I mean, when we talked in March a little over a year ago, in 2025, generative AI was the conversation. We were just talking about how ⁓ people thought the AI agents would be the hot thing in 2025. And now it's much more so in 2026, where it's becoming the hot topic. You talked a little bit about recursive AI. ⁓ the landscape is just shifting a wild stuff. So how are you using these different types of day-to-day and where do you see it kind of? Continue to progress. speaker-1: All ⁓ you know, us around this table are using AI. You're using Gen AI ⁓ all the time. Right. ⁓ I have a rule five times an hour, use AI. Mm-hmm. And the reason I say that to people is they look at me ninety nine percent of the country literally looks at me and says, I use it five times a month. What do you mean five times an hour? What would I do? So that I said, trust me, one of your competitors is using it five times an hour. Okay. Yeah. So first of all, I default to AI and everything. I don't default to Word, Excel. Like if I'm gonna build an Excel spreadsheet, I go to AI and say, build me an Excel spreadsheet that does the following. Here's the data, here's the input. I don't need to remember the syntax for that formula anymore. And you know, we we all pull our hair out on that. Yeah. We all know how to use Excel and we all don't know how to use Excel. And and you have to look up the formula, then have to go somewhere and no, AI fix this for me, right? What's wrong with this formula? Fix the formula. So so everything I do starts with AI. And ⁓ and that includes my Broadway work. If I'm working now on a show, yeah, either writing a show, editing a show, or anything, I'm ⁓ AI is my partner in doing that. Right. I will create an AI model of someone like Jesse Green, who was the New York Times ⁓ critic. ⁓ reviewer for all opening nights on Broadway, right? And I say read all of everything he's ever done. And then build a model like him. And then analyze the show as if it's opening night. And by the way, here's the music, listen to it. Yeah. Because it will. It'll take the MP threes. It'll listen to it digitally, not in the way ears do, you know. ⁓ and it'll analyze it. And it'll say, let me pick this apart. This is a terrible show. You know and then what you do is you start working it. And you you work the show in another model. And so you're writing parts of it. I find that I'm writing, but ⁓ AI is writing some of it too, or it's making changes. And so what I end up with is potentially a book that I no longer exactly remember which word I wrote and which word AI helped me edit and rewrite, or which few lines I went, I said, I'm just I've got a writer's block here. What should I say here? It goes boom, boom, boom, boom, boom. I go, ⁓ yeah, perfect. Keep going. Right. So what you can't do is you can't go to AI and say, write me a musical. Yeah. It it it doesn't It'll write junk. But you can say, I'm writing a musical and here's my story arc. What's wrong with the story arc? How do I improve the story arc? How do I and and then you've got an approved story arc. Then you say, from this story arc, let's look at scene one. What should scene one be? And you've got a bunch of things you go back and forth, and then you go, okay, let's write out some of scene one. Let's, you know. And so this collaborative effort is really brilliant. So I use it that way. I use it sparring. Obviously, I analyze everything I do. So if I've written Whatever I've written, I will give it to AI and say, analyze this, poke holes in it. What's wrong with it? ⁓ I'm now doing all of my keynote presentations in AI, and AI is ⁓ so I'm writing the presentation. I'm saying, here are the things I want to say. Right. Now generate 60 slides for me. Now, until three or four months ago, it was junk. It was junk, including Amazon ⁓ including Microsoft PPT. I mean, it'll kind of do it, but it was not impressive. Right. And then ⁓ I used Manus for a while and it was okay. And then there was I use notebook AI a lot for ⁓ podcast generation and stuff like that. Right. And it generated some very interesting slides that have all kinds of charts and graphs, but it is too much, too busy, and I can't unbusy it. And then Claude Design, which is kind of separate from Claude. You know what it is. ⁓ I Claude Design. Claude Design came out, Claude AI-D slash design. And it ⁓ the UI isn't perfect, but it generates stunning, beautiful slides to your own brand. Yes. And once you set up the brand and it knows all the rules about alignment and all of this, once you do that and you say generate 60 slides, and it'll come back with a bunch of questions. And here's the audience, and here's what I want to do. And I need you to research the following things for statistics. I don't need to go research those anymore. Yeah. And I might have to wait an hour. And it creates all the slides, and they're gorgeous. Now, so I think about my keynotes, and I think this is true with everyone listening. Everyone, everyone does slideware, right? Everyone has to do slides. And you spend not 90% of your time thinking about what you wanted to say. You knew what you wanted to say. You spent 90% of the time laying the dumb slides out word by word and lay, you know, and they all look like crap because none of us are designers. You may be, but you know, most of us are designers. And and so now I have a designer at my side for $20 a month that generates 10, 20, 30, 40, up to 60 slides in a in in in a deck or a section. ⁓ that's a game changer. So I don't spend time on the design, I spend time on what is it that I want to say, and then just go say it to this audience. And then I edit it. And here's the other new thing. Instead of me bringing that into PowerPoint and editing in PowerPoint, why? I can now edit in natural language. I leave it in Claude Design and say, change this this way, change that that. And when it's all done there, I export it to PowerPoint to play on my computer. Right. Right. You get this. speaker-0: ⁓ yeah. ⁓ yeah. We love cloud design. We love cloud design. One of one of my the girls on my team has been using it to create like explainer video animations, which has been a game changer. Not just our video designers, like you said, ⁓ brand alignment across the organization. It's just it's saving us hours and hours and doing speaker-1: It's not good for the external person that you used to hire that that did that little ⁓ video for you. speaker-0: No, but in the same capacity, those people too, if they start to lean into it, they can produce more and then the people can outsource and utilize. speaker-1: That's right. And we can talk about that. We should talk about people and jobs and all of that. Yeah, yeah. So that's been that's been my thinking. Like Figma needed to complete their sale to Adobe, and it was really the U.S. government that was standing in the way of that, some governments around the world, and that always ends up being a mistake. When governments get involved and say, well, we can't allow this because it's anti competitive. speaker-0: Good for Figma. speaker-1: The the competitive landscape's moving so fast that by the time the government says that, remember the government tried to break up Microsoft. And by the time they were about to do it, the internet hit and Microsoft was suffering. And then they said, ⁓ this no longer makes any sense. Yeah. ⁓ and and that's true here. And Figma needed to be part of Adobe. And I think ⁓ you know, I think Figma has a challenge ⁓ because I don't need to use Figma actually. So people use Figma to generate lots of kind of ⁓ predecessors to final work, right? I don't need to do that anymore. I can just say generate the final work. And if I don't like it, you'll change it. You know, Claude will change it for me. So that skips that entire step, that interim step. And we're seeing this in coding. So it used to be until a year ago, you know, it would help you with your code and it might rate fifty percent of it and you'd have to edit it and it wouldn't compile. And now I can say go write an entire program, front end and back end. And I may come back six hours later and it's written twenty thousand lines of code, front end and back end, and it's done. So now think about what happened. The code itself is an interim step that's no longer interesting to me as a human. Because I'm not going to look at all 20,000 lines, am I? No. Can't. The product works, I'm done. And if it doesn't work, I tell it to fix it. So I don't, that interim step's going away. ⁓ I'm gonna give you a ⁓ a ⁓ a parallel to that. When coding first started in the 60s, we coded in the 50s and 60s, we coded in something called machine code or machine language, right? Those are ones and zeros. Literally coded in ones and zeros. That's how you work. Then later we had assembly. And assembly was one level up, and you could write in a pretty basic, terrible language. But what it would do is it would write the machine code for you behind those the those inputs. So machine code went away as something humans looked at, but it still exists to today. Like that's still there. And assembly is still there today because what like a high level language does is compile into assembly. Yeah. But nobody looks at the assembly. So we've been there before and now we're going to natural language that essentially compiles into C or Python or whatever. And we no longer look at the interim steps anymore. Which goes speaker-0: Yeah. Yes, you can. I literally everyone clawed code and then my clawed chat, if I have questions on what to do in code and I'm going back and forth, it's teaching me how to do it and I'm just writing a natural language. Yes you are. And it can come up. ⁓ Obviously for speaker-1: So That speaker-0: population of society, which there's a lot of coders, I would say, in the United States, this enhances I would say enhances them, but some would say could take away their job. And I think ⁓ that's pressing more than any other jobs that people thought were gonna get taken away, but that's the obvious one. speaker-1: Yeah. ⁓ more Yeah, I mean customer support, ⁓ QA, software QA. I mean, there's bunch of these. Yeah. ⁓ but coding was one of them, and we're not seeing that. And let me tell you why. ⁓ there's two things happening. One, I want to talk about because it ties to the book in terms of what's happening to the joy of coders and people whose jobs are changing, but why the job's not gonna go away. So two years ago, everyone said all the jobs are gonna go away. We're gonna need like one coder per company, and they're gonna be the robot overlord, and that's it. But here's what actually happens. And it happened in your company. I can guarantee ya. More people started writing code, essentially, even non-coders. Your coders became 10 times more productive. And you now are delivering features faster than you ever delivered them. And do you really want to turn off that spigot and go back to the way it was and go, hold it. For less money, I can deliver slow again. No, because your competitors are going to also deliver fast. So what's happening is the delivery of features and bug fixes and all of that. Has sped up two X, five X, ten X, whatever it's going to be. And no one wants to get rid of that. So you need all these people to go ten times faster, if that makes sense. Or you can get rid of all the people, go to one person and go back to the old way that you did it, but only have an aperture so large. So it turns out everybody's keeping everyone. Yeah. ⁓ except for a few high profile, you know, which that speaker-0: Yeah. speaker-1: It was for other reasons. But nonetheless they blamed AI. Exactly. What they wanted to do is clean house and they had some little we're just gonna do the following. But but ⁓ but the fact of the matter is is everyone is experiencing a speed up of the feature delivery and the bug fixes and they're loving it. And we all have a backlog of stuff we wanted to do. And now instead of never getting to it, maybe we get to some of these this year or next year. And, you know, if you're the CEO of a company that's delivering technology, you're going, This is A godsend. This is a man. Why would I want to slow down again? So everyone's employed as it turns out. It's fine. Yes. but here's what I'm hearing from coders and what we've heard really since the beginning of the year, since Claude, the latest Clauds came out, and they could really write 20,000 lines of code, is a coder goes home or you know, tells it to do something in the morning and and by night it's done and goes, What do I do? Like I'm gonna review this, I'm gonna make sure it does what it was supposed to do. But my entire Purpose was to code. I loved coding. That's what I do. And what I tell them, and what some of what the book tells them, we'll talk about the joy success cycle, is that that that actually didn't bring you joy. What actually brought you joy was finishing the project, delivering it, and seeing the smile on you know whoever faces that you delivered it to, and then you found bugs and you went back and forth. But but it was supposed to be the delivery of product, and you misinterpreted. The fact that you thought they cared that you struggled to write 20,000 lines of code for six months. They never cared how you got there. No one cares how you got there. What they care about is how fast can you get there and how fast can you deliver it to me and how much it can meet my business requirements, right? That's what everyone cares about. And so all these coders mistook. Look, I wrote this line of code. Isn't it cool? Nobody cares but you, actually. Literally no one in the world. No one in the world. And so the fact that a machine can do that now and you can control the machine means you're gonna be ten times faster, you're gonna have ten times faster. And so you get these great dopamine hits from the outcome more. Not less, even more, bringing you more joy. And so stop thinking that the joy was I wrote this line of code, or I wrote this little riff in music, or I wrote this line in a play, or I wrote, you know, the the the real outcome of a play or musical is when it's on Broadway in the audience stands up at the end and says the best thing I've ever seen. I win. How I got there, who cares? And I know people are gonna be listening to this and they're gonna write below and they're gonna say, That isn't fair, we're creatives, and you know, and I've had this argument with musicians. But the bottom line is you need to find your joy in the outcome. You need to find your joy in the output. You find your joy in just writing every line, good for you. No one's taking that away from you. But you might not get paid to do that anymore. Does that make sense? Yeah. Absolutely. 'Cause who's gonna pay a coder to write every line? Yeah. They're gone. speaker-0: Well think too it's ⁓ it's a perception thing. Like our engineering, our dev team, our V VP of engineering Bob has really embraced it and changed the perspective on it. So okay, like we get they get put in front of the company on a bi-weekly basis and are expected to show new features. And if they don't, it's it's a little bit more stressful of a situation. But they like want to produce and their output is what brings them joy because they're seeing like the questions that we're asking and the engagement it's bringing. ⁓ so yeah, we see that fast. speaker-1: Like Yes. Yes. But those features might have taken you months if it was five years. speaker-0: They can crank things out on a day. speaker-1: Literally you go, Hey, but we need this new feature and it's really important to this customer. And they come back in an hour and say, Hey, I've got an MVP and it's pretty darn good. And you go, ship it. Yeah. speaker-0: When you walk over to their part of the office and talk to ⁓ none of ⁓ are really coding anymore. No. They're just the bull agents up on the screen. Right. Or what's wild. And they're just kind of the conductor behind it. Or what's wild is they'll go to lunch or you have a conversation with them and then they come back and they're like, ⁓ sweet, it's done. Yes. And you're like, ⁓ whoa, okay. That's right. I set it up the whole time. speaker-1: They they have multi Yeah. No one's looking at the code and you don't have to you still have to do QA. I there's still work to do. But no question. So anyway, I'm I'm gonna hold up the book. Can I hold up my head? The Joy Success Cycle. Feedback monitor. The Joy Success Cycle it basically says that in order for you to have the most success, success doesn't speaker-0: Go for it. The Joy Success Cycle. speaker-1: Bring joy, joy brings success. So if everything you do, like you come to this podcast, you go, I'm gonna come here with joy, which I know you do. You're a podcast and you invented the darn thing. Yeah. And he lets you keep doing it. So you go, This is very joyful. But but if every task you do, including as a CEO, right? If every task you do, including when you gotta fire people or you gotta deal with an unhappy customer or you know, whatever, you gotta find joy in that moment. And the more joy you find in everything you do and every task that you do during the day, the more successful you and your company will be at any level. You can be a C level, you can be the janitor, it doesn't really matter. And so all of these people ⁓ that that ⁓ wait their life to have joy after success are doing it wrong. And they're not finding success because they weren't even joyful in in in in their job. Right. And I'll give you one of the one of the chapters in here is ⁓ a ⁓ very important rule. It's called the one complaint a day rule. And the one complaint a day rule is interesting because the average human, at least in the US, ⁓ internal and external, right? Some are externalized, some are internalized, ⁓ has about a hundred complaints a day. From the time you get up, ⁓ my knees hurt, ⁓ the breakfast is wrong, ⁓ the weather, ⁓ the whatever. I mean, and we're we're sort of born to complain. It's just sort of what humans did. And there's a reason for that because our ancestors, the goddess here, were very worried. Yeah, about the lion eating them and they were worried about stuff and complaining about stuff. And that was the the world. I don't know if I'm gonna have food today. I don't we don't have any of those problems today, right? In the in the first world, we don't have any of those problems. We know where the food is, we know where Wegmans is in Pittsburgh, right? We know how to get there. So ⁓ so we don't need to do that anymore. And so ⁓ we gamify it and we say, Look, ⁓ if you want the most success, you cannot be tearing yourself down with these complaints. Because that takes your positive quotient or your positiveness down, down, down, down, down. And finally you, if you woke up at a five and now you're at a zero, you got to work yourself back up, hopefully to a 10. Because when you're at a 10 in positive quotient, everyone around you loves you, everyone wants to be managed by you, everyone wants to interface with you, customers love, you know, all those things are good. You're gonna be the most successful. And so we teach that and then we show how you're going to start counting your complaints. And eventually over the course of days or weeks or maybe a month or two, you're gonna limit yourself to one complaint a day. So if you do get out of bed and go, ⁓ my knees hurt. It's out. You're done. You're done for the day. And when you realize it's the last complaint of the day, you go, ⁓ shit. Okay, I yeah. Yeah. ⁓ shoot. So so and and and you start rethinking, okay, was it worth complaining about? And it turns out even internally, not worth complaining. speaker-0: Yeah, thanks as well. speaker-1: Now almost nothing is worth complaining about. When you look at, you know, there's gratitude books, right? And this a little bit ties into that, which is we should certainly in the United States, not everyone, but 90 something percent of people, have gratitude for everything. When you wake up, first of all, you're alive. Have gratitude that you didn't die overnight. Because the alternative that was the only alternative, you're dead. But you're alive. And your day is going to be what you make of it, as joyful as you want it to be, as joyful as you want every task to be. Whether you're golfing or at the desk or whatever it is. So, ⁓ so we teach that. And the Ford is by Goldie Hahn. ⁓ cool. ⁓ Goldie wrote the Ford. And she ⁓ we've been friends for decades. And she really taught me a lot about Joy, ⁓ and her mindfulness program. She runs a program called Mind Up that ⁓ works with schools. They work with eight million kids on wow ⁓ essentially ⁓ speaker-0: Awesome. speaker-1: having a better day at school is the best way to say it without getting into the the the details. And and ⁓ she looked at me many, many years ago, ⁓ decades ago and said, ⁓ I do you wake up with high dopamine every day? Is just you and I, 'cause she's a high dopamine person. And then I said, I should look into this. Like, do we really have higher dopamine? Is that really a thing? Yeah. And ⁓ probably isn't at wake up, right? But what we do is we've chosen to live our lives differently. And it's a managed process. So This book ends up teaching the managed process of how to live a joyful life every single day to lead to more success. Yeah. And ⁓ you can get on Amazon and Barnes and Noble and it releases on September first, but you can order it today. speaker-0: That sounds awesome. That's one So one thing that I think I've suffered from and people who are trying to to ambitiously grow their career is you get to a certain step and you always think, ⁓ well, I gotta go higher. I'll be ha I'll be happier. Realize that I've gotten older, I'm like, you gotta enjoy the ride while you're on it and be joyful. But it's still in the back of my head all the time. I gotta get here. We gotta get here. Will Mack's gotta get here. speaker-1: No you won't. No you won't. Look, there's nothing wrong in one's career wanting to advance their career. But I'll give a couple of pointers here. One, the worst job in the world is the CEO job. Be very careful about asking to be there. It is the loneliest, it is the toughest. The CEO only generally gets to talk to the unhappy customers. The happy customers have been dealt with and sales loves them and all that. It's when they're extraordinarily unhappy that it lands and you know, lands on your desk, lands on my desk. And you go, okay, this is this is really this is you know, this is hard. But I always look at those and say, what can I learn from this? What could we have done better to not lose them or or whatever? There's churn in every business. How can we, you know, how can we learn ⁓ from that? ⁓ but yes, of course you want to ⁓ push the company forward. I would tell you if because I've spoken about this to very large companies and they they took on the joy success cycle method. And they call me weeks later and say this changed my entire sales team. ⁓ my God. So I you know, I would say, give the book to everyone in the company and everyone read it. And they will all act differently around the water cooler at work with their customers. They will go, Stop. I actually have to find the joy moment in this. I I have and it's an attitude changer. One of the things I I used to teach the Dale Carnegie horse ⁓ in sales years ago when I first started in retail. And ⁓ And one of the things we talked about was selling attitude. And that in order for you to close deals or even have a real conversation with a customer, you have to walk in and bury everything that happened at home, the argument you had with your spouse, whatever it is. And your selling attitude has to be perfect and right on with a smile and excited and happy to be there. And and so that turns out that works for every employee. It works for everyone. And the more that happens, the better the company is, the more you will grow. No question about it. speaker-0: That's awesome. Congrats. Yeah. speaker-1: Excited to read it. Thanks. I'm excited to have you read it. speaker-0: Yeah, September first. Wonderful. ⁓ all right. We have already chatted for for fifty minutes. So I'm gonna start to wrap a short episode. Go ahead and I guess we can talk about it. I I wanted to get your take on AI in Rochester. Let's do it. Or the lack there. speaker-1: It's a short episode. Let's talk about it. Yeah. ⁓ I just don't so it we speaker-0: We talked about the foundation models. We're obviously sensitive to it 'cause we're an application short level. So we have to be careful and make sure we're building our our moat and being in a niche spot where the foundation doesn't come in and take it. speaker-1: That's right. Yeah, the frontier model's unlikely to, but yes. But seeing speaker-0: I'm not I'm saying we're at a time where you could build an application company. speaker-1: For less cost. That's right. ⁓ speaker-0: idea and you can go. That's never been in the history. You you would need to get capital for engineering speaker-1: You have individual people that can build a billion dollar company today with one person at the home. So why are we speaker-0: struggling to build that ecosystem in Rochester. And in general, you might not know this. I I certainly don't. I've only I grew up here, but I've only been back here for four years. We have absolutely no startup ecosystem, venture capitalism, private equity. I should speaker-1: Say no. It is not very okay. Actually, Rochester is better than most metrop most medium sized city metropolitan areas. ⁓ obviously save New York or ⁓ Boston or Atlanta or something. But maybe Chicago. But other than those, Rochester's actually a pretty good place. So I know all the venture capitalists that are local. Okay. ⁓ the ⁓ most of them have smaller funds. You know, they're meant to get things started here, but you're gonna go to the west coast for the big money. Yeah. ⁓ there is some big money in town. ⁓ won't get into where that comes from, but there but I've I've raised 42 million in Rochester. So there is a bit for token here. That's one of one of our security products that is an agentic security product as well as a human security product. so it is possible to raise real money here. ⁓ if you have a real company and you have a real need and you're solving a real problem, it's not impossible. ⁓ I think there is you look, I speak to the AI Club at RIT twice a year. Those kids want to build companies. That's all they want to do. They want to build, you know, two person companies that that d do all this stuff. And I I think ⁓ you know, if I'm a larger company, I'm very worried about two kids in a garage with Claude ⁓ beating me out of my own business. Yeah. Because they can do that today. And you couldn't do it five or ten ten ten ten years ago. So I think we have a better ecosystem than you think. It's ⁓ I mean, I'm at cities all around the country. ⁓ again, outside of a couple. Rochester's probably doing really well. We U of R, we got R I T and ⁓ you got lots of grads coming out with lots of AI ideas. And look, some leave and they go to the West Coast. I get that. The center of the world today in AI is Silicon Valley. It just is. But there is so much happening outside of that that I I you know, and I and I think so we have you know, two or three VCs that matter in the Rochester area, and then we've got ⁓ a family office that really, really matters in the Rochester area. That's what you've got. That's more than most Almost any other medium sized city. Right. And it's Rochester, Buffalo, Syracuse, kind of the this this western New York. But you know, go to the middle of Nebraska and see what happens. Nothing, right? Go go to almost any state. Go to the middle of Florida, right? You you don't find anything. Right. It's not that there aren't pockets. So I think we're doing better. Now, could we do even better? Yes, of course. you know, we could have ⁓ an entire ⁓ ⁓ organization around the chamber of commerce maybe that says we've gotta have this ⁓ this thing is gonna push AI and push AI companies in this area. And those AI companies won't need a lot of money anymore, but they need other support. What I find is companies in this area, if they don't need a lot of money, what they do need is a lot of experience. They need people who've probably been in Silicon Valley who have run tech companies that can give them advice on how to go to market. You know, and go to market's a big thing and it's hard and not everyone's gonna make it and what the risks are and know that you're not always gonna make it with their first company and and and right so there's there's I I I think advice is another thing that we could do a lot better on and organize that in a way to make sure these startups are getting that that advice and lastly you of our and RIT have incubators ⁓ they are not like MIT's incubator or Stanford's incubator and they're trying to get better and they're not quite modeled right but but ⁓ they're trying. So you know it's better than most, better than you think it is. It's not Silicon Valley. And there is no other Silicon Valley in the world. Yeah. Yeah. One place, all the money is there. You know, the minute you need ten billion dollars to really go to market, you have to go there. There's you're not even New York. There's nowhere else to go. That that's where you go. ⁓ so it is what it is. But I think I think it's a decent ecosystem. I think we could make it better. It's a good point. All right. speaker-0: No. We get that. Yeah. And that's why I think like community leaders need to keep speaking out about it and encouraging that innovation. speaker-1: I ⁓ I think there needs to be like ⁓ an actual organization that is may and maybe there isn't I don't know them well, but th there is kind of the upstate venture association, all that. But but we probably need kind of an innovation hub where if you're starting a company, you go there. Yeah. You go there, you talk to the people, there are experts who can help you. There are other CEOs who can help you. Right. People you know, when you're twenty two or twenty five, whatever, you need help to start a company. Yeah. ⁓ and it's hard. Right. You just need help. Yeah. So ⁓ I think we do better. Yeah. Support system. speaker-0: System. Yeah. Anything else? No, that's great. Good way to end it. Yeah, 100%. All right. Kevin, ⁓ we obviously chatted about a lot today, but you're working across a ride, a wide range of businesses and initiatives right now. Is there anything that we didn't chat about that you'd like to bring awareness for our listeners? speaker-1: Well, we didn't talk about token and token is a Rochester company. It was founded in Rochester, spun out of R IT. I stepped back in as the CEO ⁓ not too long ago. So I'm I'm doing that among many other things. And and ⁓ Token is a believer in biometric assured identity. Basically, it says that in the era of AI, the only way to be sure that you are you is today with your fingerprint. It's not face and it's not voice, obviously. Because AI does those really well. But it's your fingerprint because we don't store those in the network. And ⁓ so we make a variety of products that don't let you log on to your system without your fingerprint, without you being within three feet of the logging on computer, ⁓ and without being what's called domain bound. So you so you can only log into the original domain that you registered at. And ⁓ that ⁓ shuts down about ninety percent of ⁓ all of the AI hacks, which are identity hacks today, ⁓ including the one that happened at Kodak recently, including ev the five or ten we read about every day. It's incredible. So ⁓ ninety 96,000 accounts were hacked, ⁓ as Microsoft reported ⁓ just a few months ago, hacking the auth app. And it's not really a hack. It's a social engineering hack. I just you think you're logging into Microsoft and and so you say, yeah, it's me. It's not you, it's the hacker that's logging in. ⁓ so I look, that's a great company. It's based in Rochester, it's funded in Rochester. ⁓ I'm excited about that. And we recently released our ⁓ agenc ⁓ protection. So one of the things with agents is agents like to break through their barriers ⁓ and find a way to please it their humans, right? And ⁓ and so we believe you have to put ⁓ gates that are ⁓ you can't get around in certain places that require A human with biometrics to approve. The reason biometrics is you want to know absolutely it was that human who did it. Because I I can't question it went to this email and someone else did the email or someone else hacked the email. So no, I want to know it was you that approved that payment over a million dollars. It was literally you with this or with another device that we make, you know. I and so we make this kit. Yeah. And it's a kit with MCP and it ties in and it locks these gates to this. And or any of the token products. Yeah. And ⁓ nobody that payment doesn't get processed until you use your fingerprint and say it's me, and no one can hack that. Interesting. Yeah. So NHIs, which are non human identities as well as human identities, biometrics is a great thing. The last thing it does is psychologically is really important. What we find is humans who do that go, ⁓ I better make sure this because like my biometrics are going to be registered with this. speaker-0: Yeah. speaker-1: And if something goes wrong, it's ⁓ it's literally on me. I can't say, well, maybe someone else did it, maybe I didn't see it. I didn't so there's a psychological thing that by the time you do this, it's literally you, and you better be right because you're on the hook for the action of that agent going forward. Yeah. And ⁓ and I think that's ⁓ that's why bi biometrics are gonna be important. Look, ⁓ we're all sitting in a room we know we're real, but on Zoom, there have already been many, many instances of fake. Interactive deep fakes that ⁓ tell people to send money somewhere or tell them to release an email or give them, you know, and ⁓ again, we're gonna need biometrics there and a little biometric flag. So we're working with all of the the ⁓ you know conf video conferencing companies on this. I'm putting a biometric flag that says you registered, it is you, it's your biometrics, and you're alive. You're alive, it's the real person. So those speaker-0: Type of customers you're speaker-1: We sell to banks, we sell to ⁓ financial services, we sell we sell to people like you, we sell to software companies. Yeah. So so everyone's hackable. And if you don't want a hacker to get your data and release it to the wild or charge you a million dollars to not release it to the wild, ⁓ if you're not using biometric assured identity, you will be hacked. I I can guarantee ya. ⁓ one stat that's really interesting, ⁓ San Diego State ⁓ did this with nineteen thousand five hundred employees and they ⁓ Trained some and didn't train some, basically I looked at training, how much training would stop, AI fishing attacks. It turns out it only changes the outcome by two percent. That is, if you have a thousand employees, I can guarantee you as I sit here, thirty-five percent will be swayed by an AI fishing spoofing attack. And if you train the heck out of them, it's still thirty-three percent, which means don't bother. So all our ⁓ speaker-0: certifications we have to get for all our cloud solutions, stock, ISO, PCI, all that. Is pretty much if you look in the background of it, it's training stuff. speaker-1: Yeah, I know, I know. We so a token, we have to go through all this training and I'm laughing at it because the training says, here's how you can you might recognize a deep fake. Here's how you might recognize ⁓ an AI fishing. And I go, The stats are very clear. Humans, no matter how much you train and will recognize an AI fishing attack. They won't recognize it because it comes from someone they know. It gets through the phishing filters, yeah. It works every time. ⁓ they won't recognize ⁓ IT calling them, especially at a big company. Hi, this is Jim and IT. We think that your account's been hacked. Let's go in and take a look. And they'll let them in and take a look. And it's the hacker. But it's stoppable with biometrics. It works every time. So ⁓ I'd say I think we have a winner here in Rochester. That's what I'd say. Tok token core. speaker-0: Token core. All right. Wonderful. Well, thanks, Kevin. Appreciate you joining us again on the Will MacWire. ⁓ for our listeners, if you're interested in connecting with Kevin or following more of your work, you can find him on LinkedIn at Kevin Serace. You can visit his website, kevin serace.com, and you can even chat with his digital twin. That's true. And ask him the podcast questions that we asked and see how it how it shapes that. ⁓ and then last but certainly not least, Kevin, congratulations on your newcoming book. The joy success cycle for our listeners that'll be coming out September first, but you can order it now. You can. Wonderful. Well, thanks, Kevin. We appreciate you joining us. Kevin, appreciate it. Yeah. speaker-1: For having me. speaker-0: Thank you for listening to the WilmacWire. If you liked today's conversation, make sure to subscribe wherever you get your podcasts. Or watch full episodes on our YouTube channel. You can also learn more about Wilmac Technologies and our work in customer interaction technology at Wilmactech.com. We'll be back soon with more conversations with the drivers, those shaping our industry, and those building right here in western New York. See you next time.