Can Anyone Build a Business With AI? This Guy Did It in 8 Months

So, a large prompt is in my mind something that's going to be two or three pages long and often accompanied with some other source document.

Did you just rely on, you know, the LLMs to teach you and to guide you or did you take on some studies? How did you learn to be a soloreneur?

I I teach a lot of business owners how to think about prompt engineering and and what I think prompt engineering it comes down I have it on one of my slides but it's really just knowing what the LLM needs to do the job well

what do you think's happening in the world like when you work with these companies is everyone going to lose their jobs are we going to is everyone going to transform to be solopreneurs where do you think things are going

well I think like the development of any technology

is AI hype or Good day and welcome to the show. I'm your host Pete Cooper and our topic is AI hype or help. Today's guest is Patrick Neland from Neland Solutions.

Patrick is a soloreneur built a business over the last few years helping companies and people transform with AI. We go in deep on some of the great ways that he's helping companies transform, but he also has a lot of insights, unique insights around how AI can really help you. So stay tuned. Let's get into it. Welcome Patrick. Welcome to the show. Um, how are you today?

I'm doing well. How are you Pete?

Yeah, I'm doing great. Look, Patrick, for the audience, Patrick Neland uh is here with his company, which is uh Neland's Consulting. Was it

Neland Solutions?

Nan Solutions. There we go. First mistake straight off straight out of the gate. There you go.

It's okay.

So, thanks for joining us, Patrick. Tell us a little bit about your business. Uh and then we'll get into your background and so forth.

Sure. So I'm the founder of Neland Solutions and we help small businesses figure out how to use AI and then we we ultimately help them use it.

So um that can take the form of uh training or consulting on an AI project, but a lot of times it it involves building an automation, building setting up an agent um or uh really just, you know, we offer a free AI assessment for all companies because I noticed that that's one of the biggest issues for small business owners today is I I get 50 new posts on my LinkedIn every day that tell me I'm an idiot for not using AI in this way. I'm I'm falling behind.

And so we created a free AI assessment where you actually just have a conversation with our AI agent about your business and she asks you all these questions and then I get the transcript of that and then I sit down over the next day or two and I I look at your business and think about hey how could AI actually help your business and then I write a report and record a little video to to walk you through that report and just give you a highlevel overview to to kind of quiet all the noise around it and and just help you have a better understanding of how AI can help your Patrick, um, as preparation for this podcast, I did do that assessment and I was really impressed.

Um, as you said, it was a female voice, which is what I'm used to with Chachi BT2.

Sure.

Um, does AI really have a gender though? I mean, that's another question, philosophical question we can get into. But I like the way you structured it. I thought it was really good. You're demonstrating AI, but you also put a human side to it. Your video was very genuine. So, really good job. And later on we're going to get into maybe put you on the spot and see if there's any insights that you can share with that assessment. But let's step back a little bit. How did you get into this field? What's your background?

So my background I did my undergrad in accounting and then I was introduced to the idea of doing a PhD in in accounting which most people don't think that you could get a PhD in accounting. Like what does that even mean? That sounds like the most boring degree you could possibly get. Um, but I was told that I could make, you know, $200 or $300, 000 a year and have the summers off. And so I thought, okay, this and I don't have to go into public accounting. This sounds fantastic. And so, you know, my second year of my undergrad degree, I started, you know, prepping for for that route. And I wasn't a good student in high school. So, I I I needed to catch up a little bit academically.

Um, so I I started working hard, started studying a lot more, and I started actually prepping for a master's program that I knew would be able to prepare me for the PhD program. And what people might not understand about an accounting PhD program is that it's not just advanced accounting. It's a research degree where you're working a lot with a data analytics and with statistics and with math and you're trying to understand answer complex questions about the world, about the accounting profession, about how markets work. And so I needed to get better at math and data analytics. So that's what I decided to structure my my master's degree around.

I I set out to get a master's degree in financial economics where I spent the majority of my time I you know learning a little bit about financial economics but mostly diving into data analytics and all of the other tools associated with data analytics like machine learning and deep learning and I was so fortunate that uh I I started working on a research team in my master's program where we were developing a Python micro simulation to simulate policy uh proposals uh like tax you know let's say a state wants to propose a new state tax policy well they would use our micro simulation model to kind of look at what are the effects of that model and and a couple months after starting to work on

that ChatGPT came out and so all of my data analytics classes all of my deep learning machine learning and my work on this Python development team it all you know culminated and got started at the birth of you know LL limbs like what we now see as AI. And so I I very quickly became an extrovert because I was learning about this stuff as it was being developed as it was coming out. Um so I went ahead and uh you know started my PhD program at Arizona State University thinking hey I have this background in data analytics and AI. How can that inform my research?

So the first independent research project that I set out was how does the adoption of generative AI among tax enforcement agencies affect um affect uh like corporate tax behavior. And so as a part of that research project I spoke with the director of the department uh the Arizona department of revenue and I said so how are you guys using AI? And his response was, "Oh, we we aren't using AI. We are no like we have like five initiatives that we're focused on. There's no way we could add another one. We just don't have the bandwidth. We don't have the expertise. " You know, like there was no way we could do it.

And so I thought I got feedback from a professor and he said, you know, you're either using the wrong methodology or you're 10 years too early on this question. And I thought, "Oh, that's lame. " Like this is what I this is like one of the only research topics I'm really excited about. And so it was at that point that I thought okay um maybe I should pivot take the expertise that I have and start helping small businesses uh implement AI into their operations. So that's why I started Neland Solutions and that's what I'm doing now is realizing hey there's an opportunity and that's what we do. What's that?

How many years have you been doing that now?

I've only been doing this for about six eight months. I started back in October so still very fresh to it. Um but yeah about eight months.

Well that's all it takes. I was talking to someone yesterday on another podcast launched in February has 20, 000 subscribers or clients depending on how you slice and dice it but

sure who've come in through the front door uh and are working or potentially working with her. Uh and so this whole idea of a soloreneur must resonate with you, right? You don't have a big team behind you, I assume.

No, I have my my brother who started this with me and he works very minimally. He's he's part-time and and he helps out with a couple of tasks. Yeah.

Yes, sir.

Well, this is fantastic. So, did you have I mean, you had to learn these uh web tools and building websites and, you know, all that sort of stuff. Um did you just rely on, you know, the LLMs to teach you and to guide you or did you take on some studies? How did you learn to be a soloreneur? Oh, without question, I relied on the LLMs, right? So, I I I've joined communities and I've joined like I've I've emailed half the people I know on LinkedIn just saying, "Hey, how did you start your business? I want to learn more about what you've done. " But I joined a community that was created by actually an influencer called uh Chris Kerner. You might have heard of him.

Um Chris Kerner, he has uh his podcast is called the Kerner Office and he has a bunch of content about just encouraging people to start businesses. Like no matter what that looks like, that can be a really large business or really small business. And so a lot of his content had resonated with me and I think that was a big part of me ultimately deciding to take that leap. But it was my experience with AI, I think, that made so many things feel possible.

Like uh hopping on to because I had created a website with like a drag and drop interface like using Google sites and that was my personal website but then I got into uh talking to claude code and you know I don't have a lot of web background I have some development background in Python but but in reality I can just say hey

how should I structure a website if these are the things that I want to accomplish with my website and it tells me how to structure it and I said okay like I also want to do this? How should I, you know, consider making this change? And and then after having that conversation, all you have to do is say, "All right, now do it. " And then it goes off and does it.

And and you learn really quickly there there are AI isms that exist and all of the code that it writes that the the shapes and the colors all feel a little bit like AI. But as you start working with it, you learn, hey, you know, I I'm going to go learn a little bit about design and know, hey, I I can tell if I can tell the LLM how to design my website, it it will obey. It will do it exactly how I'd like to do it. So, it's just been amazing building little apps, building games, hopping on like a a call with one of my brothers. Hey, do you remember this video game from when we were kids? Let's just make it. It's going to take us 15 minutes.

I remember I was on a call with my older brother and he lives in India.

Yeah. Yeah,

he works for the state department. He works at the embassy there in India. And he was trying to get us to play this uh this old Nintendo game. And he was using this emulator and we were trying to play and it just there was some sort of disconnect with his internet and my internet and and for some reason we weren't able to get this emulator running. And so while he and his wife and me and my wife, we were all chatting getting ready to play this this video game, uh I just really quickly vibecoded uh another game that we had played when we were kids.

And that's what we started playing for the rest of the night. And it took 15 minutes to create to recreate a video game from my memory. And uh

that's unbelievable.

It's just so cool.

So this idea of vibe coding um is really transforming what's happening and we we've seen I guess the software industry take a real hit. I was talking to someone earlier this week and he says it's just you know there's just no workaround for software engineers. And the way I kind of think about it is is you've got like a software projects like this and what used to be a big chunk of the value creation and the cost was sitting in the middle here. You'd have your requirements and you'd have your roll out.

Now that piece in the middle has just shrunk down

because so many of us you and I because I'm not a coder either. I I'll fiddle around with Python and other things and I was trained in as an engineer but I don't do it regularly. So I wouldn't call myself a coder, but I can do things like that, too. I I stood up an app in 5 hours, but this just back to my point, this whole thing has shrunk in the middle, but the piece at the middle, the front and the back are still where I think where the the value creation is. But that's good in in a way because people can now say, well, I have this idea. Let me see if I can build it and then I can explore it.

And it feeds just straight into what you have with your done with your soloreneur. So there enormous opportunities for anybody that wants to take the jump and become an entrepreneur, right?

Absolutely. I get on calls with people and I like there's still a learning curve. There's still a learning of what's capable, but like once you learn that new mindset of I have this employee who's like incredibly fast and capable, like if I don't know how to do something, I just need to ask it. Like if there's a problem, I just need to say, "Hey, there's a problem. " And if I get to a point where it's not all working out, the good news is I've only been vibing this app for two hours.

I can just start over. Now with this massive wealth of knowledge over that I've learned over the last two hours. Now I know what I want it to look like. Now I know I want like what I want it to do. I want it to do this and this. My requirements became so much clearer. Instead of me paying a development team over the course of like two months to build an MVP, I can build the MVP in two hours uh and then learn a lot, do it again, and and it's just anyone can do it. Like genuinely, anyone can do it. I was teaching my mom recently how she could use Claude to to manipulate an app so that she could just automatically add stuff to her grocery list just by talking to it. It was anyone can do it.

Yeah, it is like that, isn't it? And I think you just underlined something where you said, "Hey, there's a problem. " And you talk to the chat uh the chat bot and say hey there's a problem you know that goes right back to a management philosophy that I was taught 20 or 30 years ago when I from an engineer I started working in management with people and I was told that it's not what the answers are the questions that you ask that's the important part of being a good manager it's asking good questions constantly asking good questions because often the solution is in the person's head anyway right? He just doesn't know how to find it.

And I think this is such a good analogy between that.

You you are effectively managing your

your employee, which happens to now be your LLM, and you just have to ask it the right questions,

right? I I teach a lot of business owners how to think about prompt engineering. And what I think prompt engineering it comes down I have it on one of my slides but it's really just knowing what the LLM needs to do the job well like that's that's what prompt engineering is and and it's a skill of developing

I have all of this wealth of knowledge and vision in my head how do I articulate it to an LLM because the LLM can can do just about anything but it doesn't have what's in my head. How do I how do I efficiently get it out so that it can perform?

Yeah, that's a skill for sure.

In your day, how much time do you spend uh prompt engineering?

Well, there's kind of like two ways that I think about it and I don't know why there's not a ton of in between. Um but the I don't know. I just I talk to Claude a lot like just throughout the day. Hey, I need you to do this. But um often it's executing prompts that I have engineered um like beforehand. So let's say I wanted to prepare some sort of document.

So the the assessment that I'm going to give you, I'm going to look at the transcript today and I'm going to uh look through it and think, you know, from a real human perspective is experiencing this because the LLM is not experiencing 2026 the way that you and I are and running a business in 2026. So it doesn't have that context. I need to look at the transcript myself and determine what what's there. But I have a really big prompt that goes out and does research on tools and it goes out and actually creates the the actual report that you're going to get after I tell it, hey, this is what I want to include in the report, you know, and this is why I want to include these things.

And so

every time I have some sort of project or or some sort of repeatable task, that's where I would say I am prompt engineering. I'm thinking about what is the most efficient way for this LLM to achieve this task to the level that if I sat down for 10 or 15 hours and completed this task, it would almost replicate what I can do. Like that's that's a prompt engineering task. That's what I'm I'm working with a client who works in the energy sector right now and and he's trying to automate the creation of these reports on on federal federal regulations

and that's all we're talking about is how do you get consistent output on this really big and complex task.

Um so you're asking how much time do I spend doing it? I I don't know probably at least an hour hour or two a day. I'm I'm carefully thinking about how can I prompt LLMs better, but a lot of it's just me pressing I my my hotkey for whisper flow is control space and hey, can you do this for me? Like I need this ASAP, you know, that sort of thing.

Yeah, I was in uh the Middle East last year and uh I read the paper over there. They said they were going to invest in this is in Dubai. They were going to invest in prompt engineering and they planned over the next two years to have 1 million prompt engineers.

Um

uh trained uh I think that whole question about what is a prompt engine anybody that's just come into the AI world you know they might have used ChatGPT and they're putting a question in that's a very basic level

right

but then you talked about um a very large prompt tell me why

tell me what that is tell me what a large prompt is in your mind

so a large prompt is in my find something that's going to be two or three pages long

and often accompanied with some other source document like some other I want this information

uh transformed manipulated and like analyzed and summarized in this really particular format.

So think about uh the the client that I was just talking about. He has a 100 pages of regulatory like filings

and he wants to basically create uh for the industry some sort of report that gives an insight and a summary in a way that you know certain groups of decision makers can make decisions. So that's where that two or three page prompt comes in where it's saying hey I need a headline. I need a sub headline. I need this information and this information. Like that is in my mind true prompt engineering. Like, hey, we're we're kind of getting mixed results here. Let me let me break it up into two separate prompts.

I'm going to prompt, you know, for I want the first section extracted this way and then the second section extracted this way. Uh, one big uh parallel use case that I think of is attorneys. They have massive amounts of documentation that needs to be sorted through. But if I just tell Claude, "Hey, I need you to summarize this," it's not going to be pulling out those really specific pieces of information that serve the attorney.

And so in reality, they're going to get a summary and they're going to read it, but they're not going to have a lot of confidence that it pulled out that important information that the lawyer like after 15, 20 years of experience knows I really only need two or three sentences in there. I need the LLM to extract those two or three sentences. So then the challenge, the task becomes prompt engineering so that we can effectively pull out those two or three sentences in a structured reliable way.

Um, yeah.

So I had a colleague of mine who was a previous podcaster as well that went, you know, going back nearly six months, but he said to me because we did one of our level up sessions where we actually take people into, you know, our one-hour session and we say, "Where are you at? We'll see if we can add to that, right? " And we did. We we gave him a couple of extra tools and he was very happy about that. But he said, "Oh," he said, "The trouble is I write, I give it a prompt and then it comes back and I say, yeah, but you haven't done enough detail or and then it does it with more detail. " And he said, "Why didn't you just do it like that in the first place? "

Yeah, exactly.

So, that's that's a little a basic example of where prompt engineering really helps. The more scriptive you are, the more information you give it,

the more likely you are to get the exact answer.

I'd say on average that's 100% true, but I've noticed there's this um this use case where it's it's that's not what I want to do.

So, let's say there's uh some sort of domain that I don't have a lot of expertise in.

So, I can't legitimately be very descriptive. I can't be very prescriptive to this model to say this is exactly what I want and this is how I want it done.

So the first one that comes to mind is like image generation. I'm not a photographer. I am not an artist.

And so if I try to be really prescriptive about the image that I want created,

it looks like a third grader came up with the idea for the image because that's the the art skill level that I have. So in those situations, it actually serves me to be a lot less prescriptive and just say and really just produce like 10 different options for me to pick from because I could see which one is good, but I'll say, "Hey, you know, create an image. " And then I'll copy and paste that prompt six or seven times and then one of them will come close to what my vision was, but the composition of them is going to be so much better. So there are situations where you want to be more general and interesting.

That's an example of where I want to be general. I think I have the same approach to art as you do. I I I mean I music would be one area, but when it comes to visual art, no idea. But I'd say to someone when I see it, I know like if I walk into a room, I know the room's set out nicely or if I see a piece of art, I like it or I don't like it. Like I I can form an opinion very quickly, but I have no understanding what goes into the process of building it.

Right. Right.

Like all you and I can say is I want this room to look nice. Like I want this picture to look good.

I can't say I need the sun to be in the top right corner and and I need it to shine on these people doing this thing and and I want to express this feeling. When I have tried to do that or when I have had ChatGPT write that really descriptive prompt, it's

it it doesn't land like I can't do it. Uh so being more general there is helpful.

Yeah, I mean I image generation and I think in video generation u for AI maybe to me it was a bit of a take your opinion to me it was a bit of a a fad like it's it was exciting it's good to be able to do it everyone plays around and we burn a lot of you know

uh electricity from the data centers creating you know the

but you know I think it's it's almost come and gone maybe not yet for everyone but it's like ah what's the point I mean yeah I can make a funny image and a meme or whatever whatever. But what's the point? Um because I see a backlash, you know, in almost like no, I don't want to see the AI generated meme anymore. That's that's par. I want to get back to genuine people.

Are you seeing that

without question? Like, but one one sentiment that I've heard many times and that I agree with is that people want value. And so

people like in a lot of use cases don't care if something is AI generated as long as it produces the same level of value. Now there are some pieces of content that

the knowledge that it is real is a huge component of the value. So, I remember the first time that I was looking at a video where I was genuinely, excuse me, faked into believing that this was real. And it was a a compilation of videos like on on Instagram where a dog was saving the life of like a 2-year-old child. And it it brought me to tears.

I just was so amazed that so many times these dogs would would come and and and oh, the TV is going to fall over and it's going to save this kid. And in reality, I believe that's something that happens. Like it can happen. But the fact that they were all framed in just the same way with the same perfect timing, even though the the actual footage, if you saw any one of them in isolation,

looks perfectly real, but I saw them over and over again over the next few days. And all the comments were like AI, AI, AI. And it's like,

who would even have a camera sitting there at that time?

Like just the the volume of these videos that can exist is inherently limit limited by like probability and statistics like

either they are rehearsed in real life which would be wrong endangering your kid

and or they are AI generated and I was just blown away. So like obviously the joy that I felt from that was that I believed that they were real

and that was an issue. But my my wife and I on the other end, we've seen a commercial and we don't want to like if we're forced to watch commercials, we don't want to watch commercials where an image is hallucinating something like that's that's annoying.

Um but Google has this AI generated Google Fiber um commercial that has beavers building a dam and it's just these beavers that are talking about like Google Fiber and you can tell that they're AI generated just by the art style. It's it's very this is an AI generated video, but it's it's well done. It's well produced and it it you know there's the the the Google logo, there's the Google other animations that are overlaid on top of this AI generated video. And I am perfectly okay with watching that and I don't like I think that they probably spent several thousand dollars on this as opposed to several million dollars hiring people to

do the CGI Beaver.

M um so

you think we're entertained by it. Is that is that where the value was in the entertainment? The the fact that you got a feeling a positive some sort of feeling from it.

I think so. Like like the value like hey I just want to look at something that's visually pleasing like and and if AI is done poorly then it's not visually pleasing but if it's done well it's visually pleasing. I'm not under the impression that beavers can actually talk. I'm not so like as long as it's honest. I think one area of video generation that's really popular that I have an issue with is um like user generated content, giving reviews.

So, think about like influencers um the way that they'll say, "Hey, I just started using this supplement and it's changed my life. "

Um I don't like the idea of that being completely AI generated

unless it's disclosed. Like, but if it's disclosed, then what's the point of the testimonial? And so, but but people have been doing this for decades where you hire an actor to say, "I've been using this for a long time. " But I I think again it comes back to

like what is the average perception of like your audience? Like do they genuinely believe this is a real person? Like so many people on Instagram believe that this influencer is their friend?

And so if if they believe that this influencer is their friend and it's AI generated, I think that that can be more damaging to society. So, I think it it's important for people to think about it carefully, but I don't think that a lot of people are. I think they're just going to produce the content that converts and and it's it's good. Um,

I think where we're going with this is that, you know, AI generated content has its place where it adds value, it works. Um, as long as it's not trying to be something that it's not. I think if we know it's AI I mean we watch science fiction for you know since the

70s I don't watch Star Wars because I think it's real.

I watch Star Wars knowing full well that it's not real

and but you like to see the story unfold and have everything connected through um what's generated there. And I think yeah there's there's that and then at the other end you do want to also I think there's still this need for integrity. I think that's where it's cutting through here. It's the integrity. If it's AI generated, it's a sci-fi movie, you know, then, you know, we know that and we're prepared for that and we take it for what it can be, an exciting future or an exciting other world versus what your example was of somebody something that's AI generated saying, "I'm taking a supplement.

" When it's obviously

not possible for an AI bot to take,

right, a supplement.

Exactly. Exactly. My wife and I, we have a joke about this uh this the beaver commercial that I was telling about that Google produced.

She asked me the question while we were watching TV. She said, "Why is that beaver wearing a hat? " Like, cuz it's wearing a little hat made out of sticks.

I said, "You know what? I know exactly why. "

The first time they generated the picture of this beaver, it happened to have a hat on it. And there were prompt engineers all over Google saying, "All right, I need you to reproduce it, but without the hat. Like, get rid of the hat. The hat cannot be there.

" And just every time the hat kept coming back and kept coming back. So they ultimately just had to roll with it. Now the beaver wears the hat in the video because they they couldn't prompt it away.

They couldn't prompt it away. I've had that experience in my presentation and I can't no matter what I say to it, it will not change the one thing.

Presentations are the worst. I I don't know why like what we understand about how it views presentations on the back end. There's a disconnect. We Yeah. There's a disconnect. Yeah. So, this is interesting. So, um you've discovered all of this in your journey, which I guess is going back quite a few years now because when did ChatGPT launch?

Was it 202?

November 2022.

Yeah. So, you were right there at the forefront now or 2026. So, I guess it was four years.

Yeah.

Four years is uh a long time in the IO world. I mean, there's a lot of people with a lot less experience,

right,

doing um doing doing what you do. So, you're quite deep into this. Um are you still what do you think's happening in the world like when you work with these companies? Is everyone going to lose their jobs? Are we going to is everyone going to transform to be solarreneurs? Where do you think things are going?

I don't think well I think like the development of any technology right um the newspaper industry slowed down dramatically after the creation of the internet because we could access the same information in a more timely manner than ever before for a lot less money. Like that's just the truth. There's going to be a handful of industries or jobs that are affected in a similar way and that happens at the advancement of any technology.

What's exciting about AI is that I think it produces a unique opportunity for there to be soloreneurs that either the the business model that they are now rolling with wasn't possible before like before AI because of the leverage that it provides or it enables somebody who never would have had the courage to step out on their own because they were there were so many blockers preventing them from starting their business like oh I know that a website isn't required for a lot of businesses to get started started, but that's a mental blocker for a lot of people.

And so, if they can hop on and just make a website in 10 minutes and it have the cool buttons and features that they're looking for, all of a sudden, that person feels like they can start a business. And so, I think there's those are two big opportunities. There's a business idea that I have that I I just created a landing page for that that may turn into something, but it's it's called history pen pals. Um, and it's essentially, you know, I I'm married, we have four kids, and we're homeschooling our kids. And the idea is that kids don't love history on average and they spend too much time on screens.

And so history pen pals is an opportunity for them to have a a a physical like visceral experience with a fictional character or even a real historical character from the past where on the back end we are um writing letters that depict a story or an experience uh and address it to a child and it's handwritten on a piece of paper. Um the content's obviously generated by AI but the kid their experience gets to be hey I'm I'm learning how to write better. I'm learning about history and I'm I'm learning about all of these things. Um, and having this unique screen-free experience where I'm pumped to go to the mailbox and pull something out, you know, like a little kid is.

Wow.

And this business

I don't think could have existed before AI in any reasonable measure. Like what's what's somebody willing to pay for that? You know what, like a 200 bucks for for one session of of letters over the course of a few months.

Having somebody craft and write letters by hand at the quality that you can say this is highquality education and this is entertaining like the the level of craftsmanship you would need to endow a bunch of writers with to to actually write these letters for it to be the product you'd be paying so much money it just never would have been profitable never ever but it's this product that's good for society that that I think is fun to do it would be, you know, your first package comes with a fountain pen. It's this awesome business idea that just couldn't have existed before.

And there's hundreds and thousands of others that people are starting similarly that just no one could have justified the capital to do this experiment whether or not this business will be successful. It's too niche. It's too there's this isn't going to be a multi-billion dollar company. You know, this is a company that might bring in a few hundred, 000 a year if it's done well. And if one person can do that, then it's worth all of their time. But you know, so I think there's a huge opportunity there.

You're really solving uh some real problems as you said, kids with with screen time, not writing. So there's definitely a problem there. I think it resonates with parents.

And what was

I was waiting for you to tell me, but you did. And that is that you actually get a real letter. You actually get a physical letter.

You know, when I was a kid, there was this thing called um pen pals or whatever it was, and you get connected with somebody overseas. Uh, and I ended up I think it was a guy in Brazil. I was in Australia. He was in Brazil. He sent me a letter. I sent him a letter. And then we both fizzled out. It was really hard to write a letter.

But getting the letter was fantastic.

Writing the letter was no fun.

Exactly.

So would the experience be um that they would write a letter back or

Absolutely. Yeah.

So, they write a letter back and uh depending on how the parents want to do it, if they're okay with the the long feedback loop, they can mail it back to them. But I I more so envision the parent scanning a QR code and just taking a picture of the child's writing and that being sent digitally back to us where we're able to to form a response letter so that we don't have to deal with the extra postage. Remember, like it just reducing friction. And it's a lot easier to scan a QR code and take a picture than it is to like, "Mom, when are like we got to go to the post office to drop off this letter or, you know, whatever it is. " Uh, I think it's a lot easier.

So, I might be able to even do it for myself like for uh maybe I'm interested in talking to Albert Einstein or something like that. So,

so I talked to a marketing specialist in the homeschool space and I'm obviously targeting homeschool at the beginning because that's that's where I'm living, but I think in reality there's a much bigger market available. I wouldn't know how to market both at the same time.

But I think about just an adult like getting a really high quality fountain pen in the mail for Christmas and then a subscription to to write with this fountain pen to a historical figure where I learn about their life and I learn about you know like what lessons did they learn that I can apply in my life like I think that would be an awesome experience for anyone. Uh so there's definitely a bigger opportunity there.

You know, there was this thing called mass customization which came out a little while back, especially when we started moving into 3D printers and the idea of

I mean, this was it's been around for 20 years where

well, just a little history, I was with Nokia um back in the 90s and they were all black phones, right? But Nokia realized that, you know what, you could personalize your phone by having a snap-on cover chunk. So, there was this whole market that just exploded where you'd have these little snap-on covers and everyone had different colors and styles. And that was the concept of mass customization.

Um, and kind of this this is a really interesting idea that with this technology, with this AI technology, it can take mass mass customization to a whole new level like in intellectually just like what you've done with this idea of pen pals. Yeah. Things that just wouldn't have been at all possible.

Yeah.

And I think there's so many people who are afraid of AI, and I get that there are still

risks here, right? There needs to be strong quality assurance on the quality of these letters to make sure that they're not hallucinating.

But like with a fine-tuned model that can stay in its lanes, if it does, like I just don't see very many negatives to this type of business where like this is just good for society use of AI. And I don't think that people are recognizing that there are a lot of things like that created by AI. Just like with the internet, a huge source of bad and a huge source of good. And I think AI is the same way. So, I hope if more people can get on board and get excited about the good that can come from it, more good will exist from it.

Mhm. Are you seeing any bad examples of of the use of AI?

Well, I mean, there's sex robots and there's the creation of romantic partners.

Uh, I was I don't even feel good telling you that my AI assistant is a girl. Like, because she's not. It's It's a female voice, so I don't know how else to describe it. I could just say it, but she is is informative because she's accessed or it's accessed through through natural language, through voice. And so there's obviously an an inherent like genderification of it.

But treating these things as humans, other than it being an opportunity for us to rehearse manners, like we should still say please. We should still say thank you

because if we don't, then we're developing the habit of not being polite.

Um, but other than that, these are machines that that function and operate and can be like super super helpful. But them filling some sort of emotional role in our life that that proxies for human connection, I think is a huge risk. I was listening to a podcast a few weeks ago where this is her entire business model, the guest on the the podcast show. It was uh the AI fix uh is the name of the podcast, but there was a woman on and her whole thing is, you know, you're going through a divorce and your sister's asleep. It's 3: 00 a. M. You need to talk to somebody like talk to this this AI person or even to the point of your your family member dies.

Let's take all of the texts that you have with them and and transfer that over to an LLM so that that person can continue to live in some sense. And and I think that that's I think those are bad uses of AI. I think that that's ultimately damaging to our our spirits and to our our mental health

in the long run. It can it can offer temporary relief. But

right, I was thinking there is this kind of temporary relief. I was analyzing myself about why using a chat is is is easier and better and more pleasant than the way I used like we had this example of trying to find someone to fix your roof, right?

You'd have to go to a portal, you'd have to find someone in your area, you'd have to get the ranking and you'd have to do a lot of searching and then you get to the guy and find he's not available for three weeks. You need him this afternoon or he doesn't do the shingles that you do. Whereas, if you come in from the AI side of things, it meets you exactly where you are. I've got a roof problem. I got this type of shingles. Oh, whereabouts are you located? Blah blah blah. Do you need something now? Like, it's it's meeting you exactly where you are. And that's

that's to me is much more pleasant than having to do all the searching myself.

Oh, absolutely.

So, but in that situation, it's not replacing human connection. It's replacing a bunch of clicks like for you and a bunch of staring at a screen. And so,

oh, it's just, yeah, like that that is the use case for AI, like replacing our usage of screens, not replacing our

good. Really good insight. It comes back to your your pen pals. I mean, I'd like to have less screen time and I'd like to see other people have screen time.

And then when you said replacing uh human connection, I think that's an important point, too.

Are people going to find themselves migrating to the chatbot rather than putting the effort into a real relationship because it's just more comforting like they're getting they're there to please you know they're they're there to to make you feel good and you do feel good so just keep going back instead of saying oh I got to deal with my brother or whatever you know

I think that it's a a source of uh I mean think about there's a everyone has problems and struggles in their lives and there's a small percentage of the population who has decided to turn to drugs or other substances or other sources of entertainment to uh alleviate that pain.

I think that there's going to be a group of people who look to LLMs to be the solution to their pain. And I'll be honest, like there have been a couple of situations where um maybe a boss at work did something. I genuinely wasn't sure how to react to it or I wasn't sure how I should word this email to try to resolve this very human problem. And I went to ChatGPT to to sort out this problem. And I think that it can be helpful there. But I was no under no illusion that like this is my this is a proxy for my wife helping me with my problems. This is a proxy for my therapist helping me with problems. But I yeah, I think that there will be a large group of people who decide to do that.

But I think just like with drugs, like it'll be we used to think that cigarettes were great. We used to think that cocaine was great. And it won't be long before we fully realize uh a lot of the damaging effects that that come from not addressing our problems the proper way with with real human connection. That's the source of fulfillment. I don't think that any lasting fulfillment can come from from LLMs. There's no there's no soul or spirit in them.

Yeah. Yeah. I I understand and I agree. I mean the your point about technology coming and having two sides to it.

I think there's quite a few cases where kids have grown up and become gamers and becoming in obsessed with and they're locked in their room um and the parents don't even not even allowed in the room without permission whilst they're game uh because that's their thing and they're literally peeing themselves in the chair or whatever.

So obsessed by the gaming and I said

sure

you know look I used to play Space Invaders uh and I got my coin into the slot and once the dollar had gone that was it I couldn't do anymore I couldn't play Space Invaders so the technology has come and made I guess made uh turned into a monster for some people not just a piece of entertainment and I think that's the big takeaway here isn't it you know um I think use but not overuse not over reliance would be a couple of themes in here.

Absolutely.

Um but also I think you know drilling into what AI can really do um in terms of I really love your idea of you said it very clearly being able to create a business model that couldn't possibly exist before. That's really interesting, isn't it? And that's where I think we should kind of as entrepreneurs as we both are really think about um easy ways, quick ways, better ways to create value for people that just wasn't on the table before, you know,

right?

Fantastic. Patrick, we're running out of time. Do you have any questions?

I don't think so. I really love what you're doing here. This has been a lot of fun to have this conversation.

And there are things that I don't have enough uh time to just chat with people about the the philosophies and the the pros and cons of AI and where it's good and where it shouldn't be used. So I I love it.

Yeah, me too. This has been great. I I I find the podcast is very real and um interesting because we don't know what we're going to be talking about and we both have to be engaged and listening to one another in a real human interaction. No AI is being used in the creation of this podcast.

Exactly. So, one final question before you for you before we wrap up. Patrick, is AI hype or help?

I don't know what response you usually get, but this is the lamest one that you'll you'll get is that it's it's both, right? Like I said, AI is a source of good and it's a source of bad. It's a huge source of hype and it's a huge source of of help. If you're looking for your information on social media and Twitter, it's going to be a source of hype. If you are in the weeds, like using the tools, talking to to folks who use it regularly and saying, "Hey, I have this problem. " And I know that it can probably be automated or augmented or improved with AI. I'm going to systematically figure out how to do that.

It is without question a help.

It it can genuinely speed up the creation of anything. Um the the work of anything where the majority of it is is is clicking or reading or or you know pasting all of these different processes. Um yeah, it's it's a help for me for sure.

Yeah, Patrick, this has been great. We're going to have you back on because you're such a great guest. Such great insights. I love the things you're doing. Nean Solutions, the pen pals, I really wish you great success with that. So, let's have you back again soon. Thanks.

Thanks. I appreciate it. I'd love to. Thanks.

All right. Well, that was a great chat with Patrick. I hope you enjoyed it as much as I did.

Now, Patrick was a seasoned soloreneur and started around the time that ChatGPT was launched. So, he's really been doing this for a while. He successfully helps companies and people use AI much better. And we talked about prompt engineering um about how to create deep prompts that are useful in certain uh applications versus going very light on other prompts and when you should use them. Also we talked about mass customization. So this is where you know you can actually get new business models which are highly customized solutions for people with AI which wasn't something that you could do most recently. So there's new business models coming up for solo entrepreneurs and soloreneurs.

And then finally, we talked a little bit about AI and the fake AI versus um you know the real content. So when is it appropriate in each case? And it's not always black and white. Sometimes AI generated content can be very entertaining and other times we need the integrity of a real human. So this is an emerging fascinating field. If you want to know more about Skillion or Nean Solutions, we'll leave the links in the description.