The Future of Business with AI

0:00How do you open source development and monetize at the same time? you create an ability for anybody to build apps. You call it GPTs. You then have an app store that's the GPT store and you take a 30 % rake on everything. How much more successful would you be if you had lunch once a week with insanely successful entrepreneurs who share their biggest secrets on how they think and achieve success? Grab your seat at the table because this is Business Lunch with Roland Frazier and Ryan Dice.

0:33Welcome to another episode of Business Lunch. And today's a snackable episode with Roland where he's going to get into some more tactical strategies that you can start using to live a rich and happy life. If this is the first snackable episode you're hearing, I'd encourage you to go back and listen to some of the other episodes that Roland has put out. And if you want to get notified every time we release a new episode, go to the new businesslunchpodcast.com website and we'll send you detailed notes along with every episode. That's businesslunchpodcast.com, www.businesslunchpodcast.com, and you can sign up for the free email newsletter where you'll be able to get all the highlights and resources from the episodes.

1:08Hey, everybody, Roland Frazier and Ryan Dice here with another episode of the Business Lunch Podcast where we basically give you a seat at the table to see all of the things that we would normally be chatting about if we met for lunch. Ryan, how are you doing today? I'm doing exceptionally well. I enjoyed it. We got to hang out a lot this last week while we were, for those of you who don't know, we had partnered up with Damon John from ABC Shark Tank to run the Rise Nation Mastermind. It's a great, great mastermind of really incredible entrepreneurs. And we were in Miami, which admittedly, neither of us really wanted to go to.

1:46Nothing particularly against Miami. It's just that for you, it's very far away. Yeah. And for me, it's like, I'm not a Miami guy. you know, I'm like a dad joke guy. I feel like I don't belong in Miami. Like I get there and everybody's like, you don't belong here. I revert back to, you know, junior high and I'm the kid who's like, can I sit here? And everybody's like, no nerd leave. But we had a great time. My own kind of personal, you know, issues aside. I think part of it though, is like, like for me, I mean, I'm ripped. I like to kind of take, you know, put on the speedo and walk down the beach and you know flex and you know it's for those of us who are exceptionally buff i think that it's a you know it's kind of the perfect place to be that's so if it wasn't so far i'd be there all the time as i think what i'm saying just to show off uh yeah you know yeah show off the guns yeah uh so right there with you uh for those of you who are what who are listening to this and not watching it on youtube uh just have a mental picture um that of that that is definitely not true.

2:49No, but it was great. But while we were gone, we're in the room, we're hanging out, sharing, learning from other entrepreneurs. While we were in the room, though, OpenAI had a pretty important developer day meeting. And I just wanted to get your take on what were some of your big takeaways from it? It seemed like it was kind of a big deal. It seemed like they're from a feature set perspective, a lot of new features added. So I want to know, there's kind of two ways that I want to approach this. Number one, what does it mean for entrepreneurs, for small business owners who are looking to be AI companies, you know, who want to build on the back of it?

3:24I think maybe there's some discussions there. Before we get to that, can you just kind of hit some of the high points of maybe what was announced at this OpenAI Developer Day? There were a lot of just kind of, and it runs together because they announced so many things so quick. So just kind of talking about some of the features that are new now that are game changers. One is the ability to upload effectively 300 pages of content. And only recently were you able to upload anything at all, it was kind of limited. But now you can upload significantly larger amounts of data, which is pretty cool in many forms from pictures to Excel files to, you know, to CSV files and spreadsheets and PDFs and things like that.

4:13So that's what are the implications of that? Like, what does that mean I can do now that I couldn't do before? Yeah. So it means that you now can have a larger reference set for the AI to look at, to summarize, to have as data that is its specialized focus right now. So like when we assign a role, which is a good prompting technique, and we say, we tell the AI, hey, from the billions of possible things that you could be doing and answering questions about, we want you to act in this role. That really helps narrow it down, helps you get better information because now it's like, okay, you are a copywriter.

4:52You are a trainer of NFL pros specializing in passing, right? Then it's going to say, okay, okay, contextually, I'm going to look at this part of the database that has all of that in it, all the things that it was trained on. But we can then also say that would be generalized kind of what's out there stuff. And it would have had to have already been trained on it because it's not currently 100 % up to date, even though it can now go out to probably the best search engine ever invented, Bing, and find data that is completely not helpful. So what you can do that changes that now is you could basically have your own 50, 100, 200, 300 page book, document, training data set that then enhances the data that's already being queried and adds additional context to whatever answers you're going to get.

5:46So for us, like what we did was we uploaded scripts for how you talk to a potential seller of a business, including specific objection handling techniques and things like that. And so it can draw on its general knowledge of how do you handle objections. And I've given it a role of a business acquisitions M &A expert. But then when I add the extra data and I have a lot of room to do that, I can get entire scripts so that now it can act as a coach to help advise us or somebody we're training on how would you handle one of these calls? How would you handle a customer service call? How would you handle a sales call?

6:27That would be just one application of it. So it's just really, I think as much as anything, it's a data enhancement type thing. In terms of summarizing data, I've seen a couple of studies that came, I think, out of Stanford and a couple other places where they said that larger contextualized data availability like this typically has what they call a fall off in the middle so that it It pays attention to the first part of the data set and the last part of the data set, but kind of like skips over the messy middle. And so it may actually give you incorrect data. You know, it's still they're working on this.

7:06Like the data accuracy dropped dramatically, like dramatically in the middle and then came back at the end. So I'd be careful about it. Like I'd put some checks and balances on it. That was a big one. The other thing that I was probably most excited about, there were two things I was most excited about. One was feature-based, and that was it literally happened in between presenting in Miami on the thing I just talked about and trying to present it on our group coaching call for our AI program. They changed the interface, and they gave us access, and they did it to all of us that were on the call at the same time.

7:45they gave us an integrated access to chat GPT. So now we don't have to choose. Before you had to choose, do I want it to be able to access Dolly 3? Do I want it to be able to access a program? Do I want it to be able to access the amazing powers of Bing? Or do I want it to have ADA, which used to be, I think that's advanced data analysis, which used to be called the code interpreter, which is the ability to upload stuff. But now we can actually, they integrated all of those things. And so if we want to have it access the internet, upload data and refer to it and also draw pictures, it will do it all without having to switch.

8:25You used to not be able to do that without starting a new chat that was enabled with that particular plugin. So I thought that was interesting. And then the thing that, and by the way, that ability to upload that amount of data, I was just looking at a YouTube video. It says open AI just killed 99 % of AI startups and AI automation agencies. And then there was another one right before that that said, a open AI just completely destroyed my$50 million startup. It's like, so by doing all of these features that everybody wanted, that they were kind of, you know, creating workarounds for while open AI was, was getting some of this stuff dialed in there.

9:08Um, they're, they're obliterating those startups. Uh, and that's one of the things I thought would be fun for us to talk about today. The other thing that they announced to me is why I would absolutely buy AI, uh, open AI stock, even if it was a crazy overpriced, which it would be. And that is that, um, they announced that they're going to roll out the ability to do, well, actually it is now there the ability to do what they call GPTs. And even though that's confusing, GPTs are basically apps. So they're little things that they've got a GPT builder that allows you to build your own functional apps within the AI on OpenAI.

9:49And it's pretty good. I've gone through a couple of them. It's pretty good to work with. And effectively, it's democratizing software development for people that can say in plain English what they want to have happen. And then OpenAI will go out and it can access separate schema, which would be more programming oriented, or just kind of do what you need it to do pretty quick or create code to do what you need to do. And then those apps that people are building, OpenAI is launching, apparently later this month, they're launching an app store or a GPT store, which is modeled after Apple. I believe they're going to take 30 % of sales for themselves.

10:31And then it'll be a place where all kinds of new development is going to happen. So how do you open source development and monetize at the same time? you create an ability for anybody to build apps. You call it GPTs. You then have an app store that's the GPT store and you take a 30 % rake on everything. It seems pretty smart. And I, that's going to probably crush anybody else in terms of they're the first to do it. And they're, you know, still leaps and bounds ahead of everybody else. So to me, like it's a, it's a very, very big deal, the things that they announced and, and, and just was a tsunami of obliteration for all of these, you know, wannabe startups or, or startups that weren't wannabes that, that had these little workarounds and stuff like that.

11:20And they were charging people for, but ultimately that's a pretty big risk, right? Yeah. So let's, I want to tie, there's a couple different directions. I want to go with this because I'm, I'm really fascinated in this as well. I want to come back to, if you're okay with the business model of open AI and what that means for other companies that want to plug into it. And also the lessons that people can learn in terms of building on top of other people's platforms. I think there's a lot of really good lessons to be learned in that. But before we go down that, because I know that's going to be a pretty long, a pretty deep rabbit hole.

11:54I'm just like, as you're saying all these things, I'm thinking I'm a reasonably technical guy, you know, and I'm, I'm, I can dive in and I can figure out a lot of stuff. I think one of the things that you and I kind of bonded over early on and, you know, our relationship was we kind of would both geek out on SEO type stuff, which definitely has its own technical kind of thing. So I'm not afraid of any of this stuff. And I'm also not old dog, unwilling to learn new tricks, but it seems like one, there's some complexities here because AI at the end of the day is software, right? The best AI, the people who are the best at AI are the best at prompt engineering, which really is coding, if we're being honest.

12:32So that could be intimidating and not everybody's brain even just thinks that way. But let's say I'm willing to dip my toe in that water. It's changing so freaking fast that keeping up is nearly impossible. So I'm a business. I'm Joe's plumbing or I've got an IT consulting company, something like that. Do I just need to, is it time just to hire an AI specialist in the company? Somebody who like their entire job is just to kind of run AI? Like it almost reminds me, you remember back in the day, early days of the internet, there was this role called a webmaster. Yep. Do you remember the webmaster role?

13:12I do. Which is a stupid name, but it's like this idea of like, okay, Your job is to be a master of all things web, worldwide web. Are we looking for like an AI master? Is that a role that companies need to be hiring for? It's a great question. I would say it would depend on the size and type of your company. I think if you're Apple, you're left behind if you don't do it. And they're buying AI chips and building their AI team because they said, I think Cook said even that they got caught flat footed on it a little bit. I think Google got caught flat-footed on it. They were like, yeah, we got this A-style stuff and we're developing it.

13:53And someday we might come out with it and then OpenAI is like, yeah, here it is. And then Google's like, whoa. You don't need to wait a minute. People aren't even searching on Google anymore? Holy crap, right? Yeah, who would have thought that? Yeah, what happened to AI and Google – I'm sorry, what happened to Google and Apple in terms of missing AI is similar to what happened with Microsoft in missing the internet. you know and google and missing social like if you're google you're like god dang we got one of these things right and have missed every other major thing but like but let's say for those big companies it's table stakes they gotta do it yeah like and and even you know any company that's technical let's assume that they're gonna make a play in it what about because you mentioned it so like you said it and plumbing and i would say it you gotta have you you gotta get up to speed if you're it you gotta get up to speed if you're plumbing then i think you can wait i think that it would be smart to, but, you know, but I've seen a lot of, of plumbing businesses and you have to, and we have a whole company called scalable that like looks at those companies as they're like completely lost as I don't have a customer list.

14:59You should get your customer list into a CRM. That would be what I would say. So like, if you've got, if you've got those things done, then then yeah, you could benefit from this, but I believe that it's kind of, to me, like the internet. And so when the internet came out, there were early adopters and there were people that kind of waited to see, and I'm not sure how damaged non-tech companies were by waiting to see. And so it was kind of like you had to have a website, but there's still plenty of people that don't have websites and do just fine. So I don't believe it would be a panic. We should all go out and hire prompt engineers at$300 ,000 a year or AI specialists, because most of them are not specialists anyway.

15:48And really, as we've delved deeper and deeper into prompting, it is very code-like. You're going to get to that if you're going to be building your own stuff, but there's so much stuff that other people will be building. And as these apps come out, it's kind of like, should I learn how to program and build an app to have an app for my business that does a function that I want? I would say no. And we've hired independent app companies. We've programmed and built apps ourselves. And I would not spend any time learning how to program an app if I wasn't just curious about it. I'd say the same thing about GPTs.

16:31I wouldn't go out of my way to hire somebody on my team before I knew what the use cases that I wanted was, because it's likely if you wait 90 days, that there'll be so many apps that you don't know what to do with. And then you'll be able to look at the rating system and see which ones are good. And then that's the one you'll get that does the thing that you want to accomplish. So what about somebody in the company though? That's like, you know, we, and we talked about this, this, you know, there was a digital digital transformation movement, where it's we need to have people come in who are just going to tell us about all these amazing technologies.

17:09Like, for example, you know, hey, Bob, you no longer have to actually type out a memo, hit print, put in a yellow envelope, give it to Sandy, and then Sandy's going to walk it down the hall to Jim and Jim's going to open it. He's going to respond and put it in this like tube. And it's going to be like you can now send this thing called an email. Yeah, right. There were roles, entire consulting industries created around this concept of digital transformation. What about, you know, should companies, even if they are a plumber, because I'm sure there's productivity gains to be made by their teams and people doing these things.

17:41Is it at a point where we should, you know, different small businesses should be looking at hiring a consultant to AI-ify their business, or maybe bring in, again, the webmaster, you know, this person who is going to help to say like, oh, here's all the stuff we could be doing. or are you saying for most companies, unless you're tech, you're probably okay just to chill for a little bit longer? I feel like we're right at the inflection point. So I do think that there's no way that you could not benefit from having AI come in, but it would be a prioritization thing. I think, like I said, if you don't, if you've got all your receipts and shoe boxes that you take to your accountant every year, you know, maybe get your accounting in order, but AI will impact all of that.

18:23The question is how much time is going to go by before AI is just an integrated part of all the things that we do versus its own separate thing. It's kind of like it's its own separate thing right now, but it's also being integrated into major apps like Photoshop and Microsoft Office Suite and things like that. But at the same time, could we reduce our copywriting staff by maybe 80 % because we have hired, we've got somebody that knows how to work copywriting AI. Yeah. The CEO of Disney, I think just, I think it was Disney. Oh no, or maybe it was Pixar. Just said that the cost of producing animated films with AI is going to decline by 90%.

19:07So like from productivity gains, absolutely. Could you benefit from it? Yes. As long as you've got your other stuff, you know, a little bit together. So to me, it's a little bit harder just because you and I see a lot of businesses that really have some very basic things they should probably prioritize. But could they benefit, assuming that the AI consultant was actually good and knew what they were doing? I can't imagine that they couldn't benefit from it. It reminds me, it just, it tells me that this is different, that AI is a big deal. Because I remember when everybody felt like they needed to have a webmaster.

19:45like every company was like you need a webmaster and to your point i because back then like in 1999 i was that webmaster yeah you know i was the company being hired by literally the lactation consultant um that was the very my very first job the very first thing i ever had was building a website for a lactation consultant which when you're 19 years old is freaking awkward but money is money yeah um now four kids for the record respect all the lactation consultants out there in nursing moms. But like, you know, this person felt like I've got to have a website. In reality, she didn't. Her clients were all by referrals.

20:20Could have been as simple, but it was like, no, I need this like fancy website with all this stuff. And I need to have an e-zine, you know, which was the very first like email newsletter. Like, so I remember that phase. Which are hot again now, just called newsletters. It's really funny. I know. Everything old is new again. But like, I didn't see that again until the social media a manager. Yeah. And then that was when we saw the next big phase. Yeah. Right. And you sort of saw a little bit of a blip when every company felt like they needed to have an app. And then sort of the open web just got, you know, more.

20:51And blockchain. Now just websites just worked. Yeah. I mean, it just, it just worked. And so that went away. But there was a time when, and everybody, again, if you, if you knew the Facebooks, you know, you were a kid and you were on the Facebooks And Uncle Charlie, who owns a plumbing thing, wasn't. He was going to hire you, at least as an intern. I'm seeing it again for the first time since that with AI. So I think this is going to create a tremendous opportunity. For those of you who are early in your career, consultants, if you can be up on this, I think you're going to have an edge. But I will tell you, just like a lot of those, just like that lactation consultant didn't need to hire me as her webmaster, just like tons of companies did not need to hire a social media manager.

21:32I think the general advice, like you said, is if you don't see an immediate application, now may not be the time to jump in headfirst. If you wait a few quarters. But I will also say it was important enough to me. I bought AIFImybusiness.com. I was like, to me, that is a thing. That will be just like digital transformation. And truly, it should be part of it, you know, the same, I think. But you need that. And you still need somebody that takes care of your website. Most people have terrible people. So webmaster as not maybe a master of the entire web, but webmaster as the master of your website and social media manager, we hire outside and internal people to do that because it's just too much and it's too specialized for us to do and keep on.

22:20And personally, we would rather shoot ourselves than do it. But also, I think that's gonna be the case with AI. I think you're going to have to have somebody that is going to, if you're not willing to be keeping up with it, you're going to want somebody that is able to help you with that. And at least that could do an assessment, like an AI assessment of your business and say, where are the opportunities to have productivity and revenue and profitability gains by implementing AI in the business? That is going to be a big value to any business that's actually up and running and has any people or processes at all.

23:05So last kind of thing, and then we can, you know, we can close out. I want to be thoughtful of time, but it seemed like when open AI first came out, there was a big question out there about, are they going to function kind of like the internet, where they're going to essentially be open, as the name would suggest, we're open AI, we're going to be open, and we're going to let a lot of people build on top of us. That's kind of what the World Wide Web and the internet is and businesses were built on top of it. Or are they going to be a bit more closed like what we see with Apple, right? And where you can have an app store, but it's within their kind of ecosystem, right?

23:42It seems like a lot of businesses who thought that they were building businesses on top of a new internet, an AI infrastructure are finding out very quickly that, oh, crap, that's not what this is at all. In reality, we tried to cozy up next to Apple and Apple's like Tootles. Or I remember the game developers on Facebook when Facebook was wide open and they could have Facebook games. And now Facebook's like, nope, you don't get to do this anymore. So what does this say about OpenAI's business model? And what does it say about entrepreneurs who are looking to build businesses on top of some of these platforms?

24:19I think that it is always dangerous to build a business based on a single feature, which I would say Snap, or that could quickly be integrated into other bigger businesses like Facebook did with Stories and Snap and things like that, that I wouldn't build my future and my business on something that was a feature. And I certainly wouldn't build it on something that was an enhancement to something that exists because the risk of being completely obliterated, like all these guys that were complaining about the, like they were building their business on, you can upload PDFs to chat GPT. And we're going to charge you for that.

25:08That's a logical to me, product development, product roadmap thing that they're going to do. But if you wanted to build an app, if you wanted to build, what's the big company that, Clash of Clans, right? Clash of Clans, you can build a multi-billion dollar business on the App Store. Just don't build the App Store, right? Build the thing that the app store, like, I think you have to think about what is the goal of OpenAI? The goal of OpenAI, if assuming that it is becoming a profit-oriented company, is let's create effectively an app store that allows people to develop useful things to businesses.

25:53Don't develop things for our business. Develop things that will be helpful to other businesses and other people. And I think if that's the big distinction that you take away, then you're likely to be safe and you can build the$50 million or$50 billion startup on that. Look at Roblox, right? So it's just you've got to build value more than just it's obvious that a feature that people want is this, and I'm going to build that feature for Apple. Apple's going to just eat you alive. That's my thought. What do you think? Yeah, I mean, I think the question you have to ask is, do your customers know that they're actually customers of this main platform, right?

26:40There's no question anytime somebody downloads an app from the Apple App Store, they understand that they're an Apple user first and foremost and that this is an add-on. if you go and download an app from the Shopify store or from one of these, like everybody there clearly understands that they first and foremost are a customer of the platform and then something else is additive to it. As long as you're additive, you'll likely be welcome for at least a while until they decide that they want to do that too. As a lot of the note-taking apps have found out in the Apple App Store, a lot of the different tools in there.

27:14I mean, I don't think you're immune from it, but I think you got a shot. If what you're trying to do is to build something that is your brand, where you are essentially white labeling the technology of open AI or some of these non-open source, because I mean, it is not open anymore. It is a, I get that it's still structured on a nonprofit. It is a for-profit. I don't know how they did that. I don't know the legal side behind it. It's going to be interesting to see how that evolves, especially if they're going to go public. Yeah, exactly. It's clearly a for-profit deal. And so that is their motive.