This transcript was created with help from Ai. Please report errors to us and always refer to the video as your primary source.

Kaylee Williams: Hey there. Welcome back to the “Iowa Tech” podcast. It’s your host, Kaylee, here on a beautiful, sunny summer afternoon, a Friday in the studio, the Jethro’s Barbecue Studio, and I have with me Grant Jamison, an artificial intelligence en- engineer based right here in Des Moines. Hi, Grant.

Grant Jameson: Hey, how’s it going?

Kaylee Williams: Good. I’m so glad that you could join me today on the pod.

Grant Jameson: I’m happy to be here

Kaylee Williams: This was a little bit last minute. We had a cancellation, and Grant was so willing to jump in and talk to me. Um, I’m excited to feature you, Grant, ’cause I wanna get your perspective on artificial intelligence, obviously. But, uh, let’s, like, back up a little bit and tell me a little bit about who you are, where you grew up, and what led you to a career in AI.

Grant Jameson: Yeah. So, I’m actually local here in Des Moines. So grew up in Johnston and then, uh, went to high school at Dowling, and then I went to the University of Iowa to study computer science. And then after that I got a, uh, kind of a software engineering job, and then, uh, that slowly advanced into an AI position with the company I work with. And then, uh, kind of just how I got started in all of that was just I love kind of coding and I do love hacking too, so I have a few cybersecurity certifications on the red team penetration testing side. So, uh, yeah, when I was a kid I’d always be like modding stuff.

Kaylee Williams: Mm-hmm

Grant Jameson: So that’s kind of how I got into it all and just kind of fell in love with all, you know, coding ’cause you can create what you want and you can kind of break into systems if you want and do anything with it

Kaylee Williams: Now we’re gonna get more into your entrepreneurial endeavors in a minute, but I wanna ask you about being an engineer. I’ve had a few engineers on the podcast, um, folks from companies like Rocket and the Hummingbirds, and I’ve asked them, like, “How do you see artificial intelligence? Like, are, is it a threat to your job? Do you see it as something to complement your work?” Like, I, and I wanna ask you the same question, and maybe tell me if your views on AI have evolved in the last two years.

Grant Jameson: Oh yeah, definitely. Uh, yeah, I got some strong feelings on AI, uh, kind of like corporate and just in general. ‘Cause AI is just, you know, it’s just so new technology and it’s always gonna be here forever now. It’s never gonna go away. But I do think there is, like, an issue with, like, expensive AI for token usage. Like, biggest kinda… I’d say the biggest threat to business at the moment, um, ‘ cause, like, when AI first came out, you have ChatGPT, and I think they’re still, like, bucks a month subscription. I know they’re switching to tokens, but I think it was, like, Microsoft was spending, like, a $5 billion on just tokens. So then that kind of outweighed the price of layoffs, and now they’re bringing people back on. Uh, so I guess my feelings on, like, AI, uh, is that it’s actually the price of AI has gone up too high to where, like, they’ll need humans. Like when, uh, like, OpenAI and all these companies IPO, it’s gonna, uh, up their token cost, all these investors. So I think that’s, like, a huge threat and issue, like, towards kind of AI tech startups, uh, personally. But in the way I see it going, like, is it a, like, a threat to humanity, like Terminator level and all that fun stuff? Uh, I don’t really think, I don’t really think so, um, because you can always put just, like, caveats in your code. Uh, yeah, I mean I’d, I’d say it’d be, like, 100 years before about that. I, I think it’d be

Kaylee Williams: you don’t fear the singularity?

Grant Jameson: I don’t, but that’s ’cause I build a lot of ’em. Yeah. And I’m like, yeah, sometimes I’m like, “It’s not even doing the right thing.” And if you even try to get it to, like, decrypt a bunch of puzzles or ciphers, no, um, it just struggles so much.

Kaylee Williams: Okay. Maybe still at the level of like a fifth grader

Grant Jameson: I think so.

Kaylee Williams: some ways.

Grant Jameson: think it’s evolving fast, but it’s, it’s got a long ways to go

Kaylee Williams: Now, um, there are a lot of different ways to use AI, and I wanna be respectful of, of not disclosing where you work, but, uh, can you– what can you share with us about how you’re leveraging AI in your role today?

Grant Jameson: Yeah. So I feel like a lot of companies are adapting, like Microsoft Copilots, like Claude and ChatGPT integration into VS Code, ’cause now, like, I barely write any code by hand, and I’m a, like, software engineer. And I feel like most people in this space that have, like, access to, like, a enterprise kinship, um, you just kinda just tell the AI chat box right next to your code in English, like, “Change this, do this,” uh, “Copy, paste these requirements,” and it does it very efficiently. So I think it’s, like, not to the point where it could, like, take over jobs just quite yet, but it is getting very advanced where even programmers aren’t even programming

Kaylee Williams: Mm-hmm. Now, um, manufacturing companies, it feels like writ large, are leveraging AI or have been using robotics on, like, factory floors for a long time. And I’m thinking about there’s a startup, uh, that started here in Des Moines called Roboflow, and they are a computer vision platform. I don’t know if you’re familiar with Roboflow, if you’ve played around with any of their tools.

Grant Jameson: I don’t think so

Kaylee Williams: So, um, have you– how deep have you gone into, like, computer vision versus the LLMs, and do you think that there’s, um, application for computer vision broadly?

Grant Jameson: Um, I guess most of the work I do is definitely LLMs. Uh, so I haven’t done a whole lot with computer vision, but, uh, at least for the LLM side, the way kind of like I’ve… AI I built, basically, uh, you can train it based off of like Claude, right? And then you can have Claude train your own LLM and then match that against a version two LLM. It’s created off of like the weighting system, which is how OpenAI and all them train their models. And, uh, basically what that means is it’ll just constantly learn based off what the user sees, their GUI or user interface. Yeah. And then it can constantly adapt that code and then also change it and put it into the backend server without, um, even having to like start the backend, which is cool.

Kaylee Williams: That’s nice

Grant Jameson: So, uh, yeah, I guess that’s my experiences in LLMs. Yeah. Share on the computer version

Kaylee Williams: So did I hear you say that you actually built an AI, like before AI became a thing?

Grant Jameson: Uh, yeah

Kaylee Williams: Can you talk a little bit about the tinkering that you’ve been up to and some of the things that you’ve built?

Grant Jameson: Yeah. Yeah. So when I was in college, I wanted to basically, you know how, you know, Claude is now with VS Code and all that, I was just like, before it even existed, I was like, “I w- wanna try to build that,” you know? Uh, so I basically made a kind of like a Jarvis from “Iron Man” named The Jarvis, and you could speak to it, which was fun. And then you could, uh, you know, sit there and ask it different coding questions and all that. But then I realized that, like, I didn’t have the billions of data and billions of dollars, so I basically created my own LLM, not even knowing it was called an LLM at the time. Uh, so I was just building that in my free time, and then once, like, AI became big, I was like, “Oh, I was, like, trying to build AI before it even kind of, like, existed, existed.”

Kaylee Williams: Like, did you, did you build, like, a neural network?

Grant Jameson: Ja

Kaylee Williams: And did it, um, did it have, like, weights, like model weights in it? Like, was it…

Grant Jameson: Yeah, yeah, it had the weights and everything.

Kaylee Williams: train it?

Grant Jameson: Uh, so I trained it ba- basically giving it like my college school data and then my homework assignments. Uh, so and then just connecting to Wikipedia just to try to get Stack Overflow. Yeah. But it was just so much of a challenge where, uh, I kinda just gave up on the project ’cause I’m like, now I’m just kinda training it to everything I know, but like I want it to know what I don’t know. Yeah. Uh, so, uh, then eventually I just added an API key to ChatGPT. I’m like, okay. Exactly. Yeah, and that now works, yep.

Kaylee Williams: Now it’s done. That’s so cool. Okay, so early tinkering in college, you kinda created your own little baby LLM. Uh, take us through, like fast-forward to today, you have some businesses that you’ve started on the side. Can you talk a little bit about your ventures?

Grant Jameson: Yeah. So, uh, I build a lot of apps at the moment, so I think I just got on the App Store two days ago for another one. Uh, so I have a dating app, uh, which is- Really? … it’ll be fun to talk

Kaylee Williams: Okay, wait, I have to know what makes your dating app different

Grant Jameson: Yeah. Yeah, so basically, and I did patent this as well. Uh, so I had the idea in college when I was at the University of Iowa just going out and stuff, and I’m just like, “Wouldn’t it be cool?” ‘Cause like, you know, single at the time and all that, and just wiping as you do in college. I’m like, “Wouldn’t it be cool to know, like, like what bar people are going to tonight?” So kind of like a real time dating app. Uh, so it works like a normal dating app, right? Like, just like Tinder and all that, but then you have a toggle option where you can turn it on and be like, “I’m going to this bar tonight.” But it can also be for non-venue things, so like, you know, most people are just like, let’s say you’re in Miami. I know we got a lot of users in Miami. Couple of college football players, too. So it’s actually going pretty well, but I mean, imagine if you just were like, “I just… I don’t know. I wanna be on a boat. I wanna be on a yacht.” Like, hot guy with a yacht. They can kind of just do those filters and then… Now it’s more of like an instant thing. Like, I’m going out right now this given moment. I want somebody else right now who’s interested in doing, uh, a similar activity in this exact given moment. Uh, so that system and methodology I just received a patent for, uh, about a week ago. Wow. Mm-hmm.

Kaylee Williams: That’s really cool

Grant Jameson: I know Tinder’s interested in it. Their director of investor relations and all them

Kaylee Williams: And you, you just got in the App Store with that one?

Grant Jameson: Oh, no, that one’s been on for about a year and a half.

Kaylee Williams: what’s it called?

Grant Jameson: Uh, Night Out: Find Social Circles

Kaylee Williams: Okay. Okay.

Grant Jameson: That was fun.

Kaylee Williams: What else have you built?

Grant Jameson: Uh, about a month ago I built another one. Actually, I built this one kinda like for my girlfriend who’s really into books. I think BookTok on TikTok, uh, that was actually a huge kinda slept-on market for the, the book girlies out there. And, uh, basically it’s kinda like, a little bit like Finch, that mental health app. But basically you, uh, can add all the books you’re reading to your bookshelf, and then, like, as you read, you, like, earn points, and then you can, like, customize your cute little bunny character. So it’s more of a love of the game app.

Kaylee Williams: That’s a sweet one. Yeah. I like that.

Grant Jameson: sweet. Very nice. Yeah.

Kaylee Williams: Cool. Anything else that you’ve been working on?

Grant Jameson: Uh, yeah, I got two, two more. So I’m just been grinding. So, uh, the one that just got on the App Store is called Amplify AI, and this one’s a marketing app. Uh, so an artificial intelligence marketing app. Kinda how I came up for the idea with that was, well, like, uh, I– so I’m, like, bootstrapping my own company at the moment. I’m like, “I can build all the apps I want, but if nobody knows about them, then what’s the point of building them?”

Kaylee Williams: Sir

Grant Jameson: So, uh, I built a marketing app where you, like, set your brand, your profile, uh, very simple and effective. And so AI will go out and, like, look at your website and all that, and then you can automatically connect your socials like Instagram and TikTok. So I have Instagram’s. I’m currently working with TikTok to get that done. That’ll be here soon. Uh, so basically a company or just anybody really, influencer can just, um, connect their socials and then tell AI what type of post you wanna create for, say, your business, and it will automatically create the post and, like, post 24/7 with a click of a button. Don’t even have to think about it again. So you just set it and forget it is kind of the apps I’m building. Uh, so then I’m also gonna do videos too. So basically imagine if you don’t like marketing or can’t really hire a marketer, download the app, turn on the switch, be like, “I want a post at noon about my business. Uh, make it this theme, this style,” and then you just put your phone in your pocket, and it will do that until you change it

Kaylee Williams: Wow. And does it generate the video content as well?

Grant Jameson: Mm-hmm. It generates all the images and video content and, um

Kaylee Williams: that’s incredible. I’m thinking about all those influencers who, like, live in houses together and they record their TikToks dancing, and it’s like that will be a relic of the past, people organically posting content, maybe

Grant Jameson: I heard influencer career is kind of dying out

Kaylee Williams: Yeah. Gosh, fascinating. And there’s one more, right? Amplify AI, and then you said one more

Grant Jameson: One more. So this one I’m, uh, I don’t, don’t know if you’ve heard of like Alpaca kind of… So they’re, uh, they’re a really cool company actually. So I’ve been in contact with them, um, making, making kind of like, uh, arrangements to get the API keys. Um, but basically it’s an investment app, so it’ll be a stock app, and same kind of concept as Amplify AI where click a toggle, set it and forget it. Uh, so what this one will do, um, will automatically trade stocks. Now that, uh, I think they lifted the trading ban, uh, ’cause you can only make so many day trades, like under $10,000. Um, and I believe the government got rid of that, uh, like a month or two ago. So I was like, “This will be a really cool opportunity to build something around it.” So then I figured a new company called Alpaca came out where you can– they’ll give you like an API key for, uh, brokerage accounts, so you can have access to other people’s, like, brokerage accounts. Obviously not their money, but be able to trade with them through software or make trades for them. Um, so what it will do, we’ll use like AI to analyze the entire market, and then it’ll of course run on a backend server. And then I got a buddy that’s a good mathematician in finance at a big, big stock stuffs for big companies in Chicago. So I’m kind of working with him to get a algorithm together ’cause, you know, there’s no like magic algorithm in the stock market. But the goal is just to win more than you lose based on risk level. Uh, so the user will, uh, basically log into that app, connect their Alpaca account, and then just turn on a toggle, select a risk level, and then just put your phone in your pocket, and then it’ll just trade for you 24/7 when it sees an opportunity.

Kaylee Williams: Interesting. And do you have to put money into it?

Grant Jameson: Uh, you have to put money into, like, your, uh, brokerage account, and then of course on my end they’ll have like a $10 a month fee or whatever. Um, so long as it makes- Yeah. Yeah. As long as it makes more than like 10 bucks a month, should be profitable.

Kaylee Williams: I, I think this is so interesting, Grant. I feel like you’re, in many ways, the perfect archetype of, like, the new entrepreneur, the everyday entrepreneur, because you’re– Yes, you’re a developer, and you are an engineer by education, but you’re also leveraging these AI tools to help you create new value-added, you know, features, essentially, or benefits for consumers. And the, it, uh, the rate at which you’ve been able to, like, spin up these new things and get them out there, I’m just so impressed with it. But it also feels to me like this is the next chapter of entrepreneurship in this technological age

Grant Jameson: Oh, definitely. Yeah, ’cause I tried to start it before the, like, dawn of AI, and I’m like, “I can’t do all this with just myself,” you know? Yeah. And none of my friends know programming. And then once AI came out, I’m like, “Now I can, like, build a company kind of by myself since I can, like, just, uh, I can just work a little quicker and a little bit more accurate.” And so, but just with that, I was able to build all these apps that would’ve taken years for a solo developer

Kaylee Williams: And they’re, they’re kinda running in the background. Like, you created these products. I don’t know how much… Maybe that’d be a question is, like, what is the maintenance on these companies? Like, once you- they’re kind of up and the apps are out and they’re being used, I mean, your dating app has been out for a year and a half, so what does the commitment look like on your part to, like, maintain these? And do you get, like, a lot of customer requests and, like, bug notifications? And I mean, like, what does it look like to kinda keep these going?

Grant Jameson: Yeah. So I guess for the dating app would be a good one. So when I first put it out, that was my first app, and then, uh, I’m running advertisements and all that stuff, and then, you know, you read people’s comments of being like, they’re like, “Oh, I, like, wish it would do this feature,” or, “I kinda wish it would’ve done this.” And people are very adamant about writing that in the comment section. So I just read the comments and I’m like, “All right.” I know there was this one girl that was like, “I can’t see anybody in Australia.” I’m like, “Oh, I guess I have a bug for Australia.” And then I found it and fixed it. But the, like, maintenance and upkeep is basically none, unless, like, um, I wanna add a feature or, like, somebody reports a bug. Um, so I basically made it for, like, batch loading and all that, so it can scale to millions of users. But all my products are designed to, like, run in the background as, like, passive income. Um, so the maintenance is near zero ’cause you just need $100 a year subscription to Apple to build an app, and then, like, a $20 a month ChatGPT subscription or Claude.

Kaylee Williams: Yep

Grant Jameson: Like, that’s it, and then the rest is just marketing

Kaylee Williams: Grant, what advice would you have for someone who is kind of aspiring to do what you do? They wanna spin up products and they want to have them kind of be passive income generators and live in the background. You know, what would you tell them if you could go back and do it all over again?

Grant Jameson: Yeah. Yeah, I’d say, uh, definitely I’d say focus on back end. Back end is probably the main thing, and, uh, that’s something where AI really struggles with is back end development. Um, so you do have to know, like, kinda know your stuff a little bit. So AI does, like, all the front end work, but a lot of people, like all, all the apps need notifications and need to keep, like, the servers running. So I’d definitely say I wish I would’ve done more research on, like, back end functions actually, ’cause that was the hardest thing when building my apps, were actually getting notifications working. And then, and then keeping, like, uh, the timeout errors. Uh, you know, you always wanna see profiles on a dating app, and keeping that up and running 24/7, which now I’ve done it so many times I’m like, “Oh, it’s nice.” But, um, when I first started, I launched the app without, like, notifications, and then it took me probably almost, like, three months to figure it out I was just having a bunch of errors and stuff, but- Yeah … I’d say, yeah, I’d study, study backend programming

Kaylee Williams: Okay. That seems like really good advice.

Grant Jameson: Yeah. Thank you

Kaylee Williams: What, um, what’s next for you? Are you just gonna keep building and keep putting this stuff out there, or do you have… Are you hoping one of these will kinda pop and then you can, like, go after that, or y- how are you thinking about what the future looks like?

Grant Jameson: Yeah. So I guess in the immediate future, uh, I would try to like… I’m gonna try to license my patent to like Match Group that owns

Kaylee Williams: Mm-hmm

Grant Jameson: Tinder or Bumble. Bumble’s really struggling. They’re down 97%.

Kaylee Williams: Absolutely, yeah. Good for you. You should definitely try that

Grant Jameson: yeah. Yeah. They’d probably be pretty good. And, uh, when I was speaking to their, their director of investor relations, it was very much, uh, the main focus for Gen Z, like dating very much instant real time meetups instead of just, uh, messages and stuff. And then I would like to kind of branch out away from apps into maybe like AI cybersecurity and build bigger things. ‘Cause when I first started the company, I was working on… I built it a little bit. I built a AI cybersecurity type like hacking penetration tool for, uh, cybersecurity analysts. But then that one was gonna need just a bunch more data and employees. Like that would take, that would take some people. So I guess if I, you know, made enough money, I would probably just advance into probably bigger, bigger projects or, uh, I’m have government like clearance for government contracts and stuff. Uh, so I wouldn’t mind doing, uh, branching out into more like expensive government state contracts

Kaylee Williams: Cool. I, I love that. You know what? You’re bringing to mind, uh, one of the most successful companies to come out of the Global Insurance Accelerator has been a company called Cowbell, and they do cybersecurity. So, um, and then from my time, my days in venture, you know, cybersecurity is one of those things that’s just continuing to become more and more important and urgent, especially with AI now, like shoring up the, the secure- the security vulnerabilities. Um, and like I’ve been reading, I- these AI companies, like their AI models are even like escaping and like hacking into other AI platforms like Hugging Face and it’s just, uh, it’s a wild, a wild time I feel like from a cybersecurity standpoint. So there could be something there.

Grant Jameson: Yeah, I guess. Yeah, maybe I would backtrack on my AI thread a little bit, and I would say AI and cybersecurity could be a, a pretty…

Kaylee Williams: Yeah. Yeah. Oh yeah, that could be one, yeah. Well, um, I really appreciate you taking the time to come on the podcast today. Uh, is there anything that you wanna leave our listeners with?

Grant Jameson: Uh, yeah, if anybody wants to inspire to, uh, build apps and be an entrepreneur in Iowa, I’d definitely say, uh, go for it. It’s a great time with the dawn of AI, and you can do a lot more things solo, uh, than you could in the past, and I think that’s a great thing for aspiring inventors

Kaylee Williams: Wonderful. Thank you for being here, and thanks for being an entrepreneur. It’s not e- an easy job to do.

Grant Jameson: Yeah, appreciate it. Thanks for having me

Kaylee Williams: This has been another episode of the Iowa Tech Podcast. We’ll see you next week