Streamlined Solopreneur: Tips to Help Solopreneurs Automate Their Business

It has been 2 years since ChatGPT became publicly known, I guess, and really blew up. And in that time, a lot has happened. And while I have always been AI hesitant, it has fundamentally changed how I work and that's what we are going to talk about on today's episode of the Streamlined Solopreneur.

Welcome to the Streamlined Solopreneur, a show for busy solopreneurs to help you improve your systems and processes so you can build a business while spending your time the way you want. I know you're busy, so let's get started.

Alright. So let's dive into this episode. All of the show notes will be in the description of your podcast player as well as over at [streamlined.fm/446]. This is also the penultimate episode of the year. I have one more after this. It's my year-end review, which I guess I should do if I'm gonna talk about it. But, after that, I'll be taking a short break. There may be bonus episodes. There may be some crossover with my podcast advent calendar, but this is the second to last official episode of the year and it's been a good one. We'll get into that next week though. Today, I wanna talk about, well, my big problem and how I'm solving it or really how I am continuing to solve it.

AI is not a thing that is going to go away. This is something that we're going to be using for the rest of our existence and probably more and more. So my big problem is that I am AI hesitant, but I feel like I'm underutilizing it.

And I think I'm AI hesitant with good reason. Nearly every AI tool I have used has gotten progressively worse the more I used it. You know, I was a big fan of some of these podcast repurposing tools in the beginning and they just kind of as they learned more from the collective we or the collective us and less from just me, I feel like it got worse. It got further away from the way that I write and talk. and that just doesn't work for me.

And I also, you know, I also think that as we try to figure out the best way to use this technology, it's a hammer or it looks like a hammer to everybody and so everything else is a nail. Oh, I used, Google l m to summarize this talk I gave and then release a podcast episode about it. I'm like, oh, okay. It's cool that it could do that, but what's the practical use? Right? Or I didn't know how to fill out this one chapter in my book, so I just had ChatGPT write it. Well, okay. So now you're not writing your book.

And so when people I think when there's like a collective push towards something, I am, naturally resistant to it or I at least wanna think through it more. Right? I make a lot of snap decisions, but when I see a lot of people making the same snap decision, I take a beat and I go, is this the right decision? We all think the same thing. Have any of us really thought it through?

And so this is how I felt about AI. Everyone was like, AI is amazing. It can do everything for us. And I'm like, should it? Does it? Can it? And so that's my first hesitancy with AI. My other hesitancy or the other reason I don't like or I did I should say I did not like AI very much was it the ones I've used wouldn't remember my main directives, like never use the word Delve. I don't say Delve. I never say Delve, and so I don't want AI to use Delve when it's writing on my behalf. Or when I tell it to always write in the first person.

One time, I told ChatGPT, like, it wrote a thing that was, like, join Joe Casabona as he delves into all the differences between these two microphones or whatever it was. And I was like, right. Can you rewrite this description in the first person? And all it did was write join me, comma, Joe Casa Bona, comma. I'm like, okay. That's not really how I would say it. Right?

And so and I know that there are people who will say, you just need to train it more, and you just need to give it more of your stuff, and you just need to write the right prompts or whatever. But at this point, I am hesitant and I don't want to be. So since I conceived of this or since I kinda realized there was a 2 year anniversary for ChatGPT coming up, which was late November if we're keeping track and this episodes coming out slightly after that. I wanted to think about all of the ways that I do and can use AI and kind of talk through those things.

When I was 12 years old, our family van got stuck in the mud, like deep stuck in the mud. So we were standing there in the rain wondering how long it would take a tow truck to come help us. When out of nowhere, this behemoth of a man came to our aid and by himself pushed our van out of the mud, and got us unstuck. I'll never forget that and my dad still tells that story to this day.

Maybe you're feeling stuck and you need somebody to come alongside you and push you out of the mud. As a solopreneur, I know how overwhelming it can be to juggle everything. Daily tasks, growing your podcast, building your authority, and that's why I created the unstuck sessions. This is a flexible personalized coaching program where I'll help you automate processes, delegate efficiently, and focus on what really matters. It's an on-demand tailored to your schedule designed to help you make real progress program. And for the month of December, I am offering a special 33% discount for listeners. So if you are ready to get unstuck and start your 2025 the right way, head on over to [streamlined.fm/unstuck] to learn more.

Okay. So let's get into ways I can or do integrate AI into my current workflows. I will say that ChatGPT is the only AI app I use right now. I, don't really use it. Like, I don't use perplexity anymore because it had, like, a ridiculous hallucination and I was like, that's it. I'm done. I don't use Claude, which I've heard really good things about. I use ChatGPT and custom GPTs. And so I've been really impressed with ChatGPT's ability to punch up text I've already written. So I will write something, I will say, hey, make this clearer or improve the language. And by and large, that's been really good for me

Going in the other direction, right, if I give it like a recording I've done and I'm like write a blog post, it gives me some stuff I can use, but I largely rewrite most of that. So it's not great on that end. But, like, the punch-up stuff, improve the grammar, reword things, are there perspectives I'm missing? That stuff has been really good.

The other thing that I'm, you know, the images for ChatGPT, they're very and they are very obviously ChatGPT like they have that look. They're fine. I have 0 graphical ability. So if I want a custom graphic, you
know, like I've been doing for podcast Advent where it's like I want Christmas and I want advent or I want podcasting whatever I'm talking about. I can't generate something like that. Whatever I put together in Canva will look worse than what ChatGPT comes up with. And so I'm fine with that. Once I get an image I like, it does a really good job of making it a thumbnail too. So, like, that's also so I can make it like a box on the advent calendar. But something I do struggle with with the image generator is and this is probably my approach. Right? I've noticed that it gets really heavy handed with text when I give it a lot of text.

I mean, like, maybe it doesn't know what to do with some of the text, so I just put, tries to put it in the image. But, usually, I will give it a blog post and I'll say, take this blog post and create a featured image for it. Maybe what I should do is say take this blog post, summarize it into one sentence that could also serve as an image description, and then generate that image. Maybe I'll try that, but at this point, it generates an image. It has text on it. I say, can you remove the text? And then it adds more text. It's like, here's the image with no text, and I'm like, this is more text than what you gave me. So that's fine. I generally and eventually do get suitable images for my purposes. I will never be a graphic designer. I will never be like a digital artist. I don't purport myself to be, and I also don't say, hey. This could totally replace digital artists. For me, it's just I need something fast and this is good enough. If I ever get feedback from people reading the blog post, like these images look like garbage, I’ll probably do something different, but I'm unconvinced of that so far. So that is so those are 2 ways.

A third way for me is I've been using it more for searching and research because it searches the web now and site sources, and it's been very impressive. I heard the guys on Connected and Upgrade talk about this, mostly Mike Curley. But they've been using it instead of Google. So they'll ask ChatGPT something, It'll get its sources. It'll give a summary, and that feels more natural. Right? Instead of searching like, Al Borland actor age Home Improvement season 3, which is something I googled because I was wondering if he was older than me at that point and, he wasn't. I'm older than Al Borland in season 3. Richard Karn is his name. I'm older than Al Borland in season 3 of Home Improvement, which makes me sad. Anyway, I can just ask ChatGPT. Right? How old was the guy who played Al Borland in season 3 of Home Improvement and like it gives a very good description? Something I did recently that I was extremely impressed with was the overhead or the…Yeah.

I guess the overhead lights on our stovetop went out and, you know, they're not like normal light bulbs. And so I took a picture of 1 and I uploaded it to ChatGPT and I said, what kind of light bulb is this? And then it sent me the light bulb or what kind it is. And I said, where can I, like, can I buy this on Amazon? And it gave me a link to buy them on Amazon. Are there smart versions of these lights? Yes. And it gave me an Amazon link to smart versions of the lights. That was really impressive. I was very, very impressed by that. And so I can definitely see myself using that a lot. ChatGPT has already been a research buddy for me, or like, in I I don't know if this is just in programming, but we call it rubber ducking. But, like, what I'll generally do is just kinda ask it some free-form thinking questions about stuff, and it'll give me some sources and some perspectives and then I'll kinda take that and run with it. Yes. So rubber ducking, thank you, ChatGPT, is a problem-solving technique commonly used in programming and other fields requiring complex decision-making. The term comes from the idea of explaining your problem to a rubber duck or another inanimate object.

So thank you ChatGPT for confirming. This is another thing I do. I will confirm especially if I use a word in like, if I think I might be using a word incorrectly, I'll usually just say, like, hey, define this. There are 2 more things I wanna touch on here before moving on, like, the future and feedback.

One is early thoughts on Apple Intelligence. Apple's Intelligence is not good. It's, like, pretty bad. I've turned off the summaries, like, the notification summaries which I think could be the most useful thing. Right? But they're not. They're bad. So bad that they're pointless. And so I mean, I just really haven't been, you know, and like the AI writing, it's like fine. Right? It's like every other AI writing tool. Maybe if I can integrate ChatGPT, I don't know if I can do that in the beta yet. Maybe that'll be better, but ChatGPT is definitely better than Apple Intelligence. Apple's really I know a lot of people have said this, but Apple's pretty far behind on this. And I know they don't just jump on the hype train. Well, it kinda feels like they did this time. Like Image Playground is weird and I hate it. Generating emoji, like that gen emoji is what it's called. That's fun because it's like benign most of the time. But that's really interesting. Now I will say one of the things that prompted this idea was Apple's definition of a photo, which, Nilay Patel did a review of the iPhone 16 Pro for the Verge when it came out back in September. I'm reading this from Mac stories. I'll link in the show notes. And here's the question that or here's the answer that he got from Apple's VP of Camera Software Engineering, John McCormack when asked about the company's definition of a photograph. Here's our view of what a photograph is. The way we like to think of it is it's a personal celebration of something that really actually happened.

What is striking about this definition is something that really actually happened, like Federico Vatici points out in his write-up of this is a really good baseline because Samsung has a definition that is basically nothing is real and Google says everyone has their own memories which is a very sad definition of what a photograph is. Like, I have a photo from my oldest smashing her smash cake on her first birthday and it's a really good capture. And, like, my someone else took it. So my wife and I are in it and we're laughing and to look at Samsung's definition, nothing is real. Like, what? Just however you wanna remember that is how you remember it, and this is basically Google's Google's version 2. But what that picture captures is real joy for 2 new parents in a challenging year. The 1st year of your kid's life is challenging. And so, Apple saying it's a personal celebration of something that really actually happened is something I love and it's why I'm willing to give them a pass on being kinda crappy at AI. Because while they're more focused on the real stuff that's happening, Google is focused on generating fake stuff like a fake conversation based on 50 documents you uploaded to Google l’m. And who knows what Samsung is up to?

But so that's my thoughts on Apple Intelligence and why I'm, like, cool with it. Now I will say that felt like a really strong way to end this segment, but I do want to mention one more thing. There's there's one more thing that I've been using ChatGPT for that's been really good and I wanna give my friend Alastair McDermott a shout-out here because he built a custom GPT that's like a sales-call analyzer, and I've been having a ton of sales calls lately. Like, yay me. Business is good. Huzzah. And I've just been, like, feeding the transcript into ChatGPT and saying, like, summarize this and tell me what I promised. And Alastair's custom GPT is a lot better at that. So it's like, you know, he didn't take this. He tells you to anonymize the stuff, And there's a whole big, long, very complicated prompt, and then it spits out an executive summary and the deliverables and some things to consider. And it's really well done, and that has been the starting point for creating my proposals. And so I'll take that and then I'll feed it to my ChatGPT. It's not a custom GPT. It's just like the one that I use for this particular conversation or this particular topic. I'll say, here's the executive summary of a call I had. Please take this and then use my proposal format to come up with a proposal.

And I'm totally fine with this because proposals are, like, kind of bland anyway. And so, like, I'll add my intro and my outro, and I always record a video to go along with my proposals. And so that's a personal touch. But, like, as far as, like, just listing deliverables and investment and stuff like that, I'm fine with an AI doing that, especially because I send the video which is the personal touch. So that's the fix. It's not small. I've been talking for a long time. But I, so I do wanna what I'll do is go over the feedback, and then I'll give a quick I'll close on the future of AI as I see it.

I'll be writing a lot more about this for podcast Advent because I always do I always last year and I'm gonna continue this year. The last three articles for Podcast Advent are The Ghost of Podcast Past, the Ghost of Podcast Present, and the Ghost of Podcast Future. So as far as feedback goes, Eric Melcore. Eric, I'm sorry if I'm saying your last name incorrectly. He responded to my question which was basically how has AI changed your your work or affected your work. It has definitely streamlined my writing and preparation process, nothing fancy, but now starting to use perplexity and taking the answer and plugging it into ChatGPT to get content ideas. Love this research and ideation are one of my primary uses for for AI as well especially ChatGPT. I'm always impressed by the stuff it thinks it quote unquote thinks of that I didn't think of.

And that's really valuable for me because I want to I don't want all of my writing to, like, contribute to some monoculture. Right? I want my writing to bring in various perspectives. This is one of the reasons I continue to have people on my podcast. I don't I don't want my show and my content to be group think or just like one point of view. I want a lot of points of view and ChatGPT and AI do help a lot with that. Susan Bowles, a friend of the show, I'm gonna say. Susan, thanks for writing in. For me, she says, for me, I've really started using it to help get boring work done.

I don't use it for content creation, etcetera, but I use it to help me do work that I always avoid like writing SOPs or summarizing calls and action steps. Yeah. This is great. I was on a call, like, a discovery call this morning. Well, it was really like I'm there's an organization that I volunteered to do their website for to get it up and running for Giving Tuesday. And so I was walking through how to do some stuff. He had some questions about how to manage the site. So we recorded it and I said, here's what you should do.

Take this call, transcribe it, put it into ChatGPT, and ask it to write out, you know, say, like, there were several points in this video where Joe walked me through a process. Can you find those and write out the steps? Kat Corbett, again, Kat, thanks for writing in. Sorry if I'm saying your last name incorrectly. She says I love this. She says it has literally never been useful to me. Every time I tried to use it for pod copy or email drafts, it gave me a generic slop. I would be embarrassed to send it out, but it pushes me to write better copy myself, so I guess there's that. This is great.

She's right, And Jason Snell put this perfectly on upgrade. He said that AI writing is mediocre writing. If you are bad at writing or making content, what ChatGPT puts out is fine. But if you have any talent or, like, any kind of perspective, it's it's going to be worse. And so I agree that this kind of stuff that Kat has tried to use it for is bad, but it's done the same thing for me. I for me, I kind of use it if I'm stuck on something. I'll give it a prompt like, hey. Here's a video where I talked about a thing. Can you write a blog post based on this video or this podcast? Or I'm trying to write a podcast on this. Here's what I think. Can you write this? And then I read it and I'm like, this is garbage and then I start writing something better. So, like, again, it's kind of like the rubber duck in that situation. I love Kat's perspective here, and it that does bring me to the future.

So, The Coming Wave is a book that I read recently. So here's what I think. Okay. So it's, Michael Bhaskar and Mustafa Suleiman. Once again, I hope I'm saying those names right. And it doesn't just cover AI. It covers Biotech as well. And especially as the convergence of AI potentially autonomously thinking computers and biotech, and synthetic, biology, and robotics. Combining those things could be problematic, and I absolutely agree. The problem is the if I'm not gonna give, like, a deep I'm not, like, one of those Goodreads reviewers who's, like, high on his own supply. My main problem with the book is that the authors have a really hard time straddling the line between technology will save us and technology will destroy us all. But my main takeaway is that we in the creator space are thinking very small about AI. We're not we're not thinking of how the actual physical world will be impacted, And the biggest that made this point hit home for me was probably a throwaway line. With robotics and AI, we'll have, like, robot busboys to take away or stuff and maybe even robot waiters and waitresses to take and bring our orders. And that is a deeply dehumanizing world. It's taking away a human touch point.

And in the creator space, this is the thing that will always set us apart from AI. I posted a video on TikTok about this, and LinkedIn, I guess, where I wrote an article, and the same day I wrote the article, I listened to a story about Tetris. And that story informed how I wrote the article and how I use that analogy. And it doesn't matter how much I could have trained ChatGPT on everything I've written from kindergarten up until yesterday. Right? It wouldn't have known I listened to that podcast. I wouldn't have known how I internalized that. It wouldn't have known that I was working on an article where I could take what I internalized and use it as an analogy in the article. It would have never written that.

And so this is the thing that AI can't replace, the human element, the stories, which is why, like Cat said, most of what it puts out is crap because it takes that away from us and we need to remember that. As we move forward, first of all, I would love to hear how you're using AI and how you plan to use AI more in 2025. You can go to streamlinedfeedback.com, and then there's even a way to upload a voice note now.

So, I can include that in the show if you leave good feedback. But as we move into the new year and these tools get better and smarter and faster, we need to remember what Susan said that we should be using it for the boring tasks, the ones that we don't really want to do, and then we'll be freed up to work on more of the creative work.

Alright. That's it for this episode. Thanks so much for listening. Again, I'd love to hear your feedback. You can head over to [streamlinefeedback.com] for that. That link, all of the show notes will be over at [streamlined.fm/446]. Be sure to tune in next week for the last episode of the year. It'll be my year in review, another solo episode. And if you write in with feedback on this episode, I will read it or play it if you send a voice note on that episode.

So thank you so much for listening, and until next time, I'll see you out there.