Macpreneur

This Solopreneur Automated His Lead Gen on a Mac using AI with Lucas Erb

• Damien Schreurs • Season 7 • Episode 155

Send me a Text Message

🆓 Discover your AI Stage: https://macpreneur.com/aistage

In this episode, you'll learn how keynote speaker, published author, and technology consultant Lucas Erb uses his Mac to streamline his business. 

Discover how AI tools like Granola transform meeting productivity, how "vibe coding" is replacing traditional programming, and how workflow automation tools enable sophisticated decision-making processes.

Video and show notes at: https://macpreneur.com/episode155

Connect with Lucas Erb:

Highlights:

  • [00:00] Welcome  
  • [01:12] Introduction to Lucas Erb  
  • [08:19] Top Apps for Running a Business  
  • [13:28] AI in Software Development  
  • [20:55] The Future of Personalized AI Applications  
  • [24:31] Exploring N8N for Workflow Automation  
  • [29:51] Understanding AI Agents vs. Assistants  
  • [32:45] Advanced AI Automation Tools  
  • [35:36] Deep Research and AI Efficiency  
  • [42:31] Practical AI Tips for Solopreneurs  
  • [46:53] Lucas' Mac Wish  
  • [49:59] Final Thoughts about AI  
  • [51:45] Connect with Lucas Erb  
  • [52:18] Applying to be a guest too  

🎤 Want to be a guest on the show? Fill the application form available at https://macpreneur.com/apply or visit the show profile on Podmatch.

👥 Join the Macpreneur Community!
Simplify your digital life and streamline your business with fellow solopreneurs.
✨ Join the Waiting List

âś… Want to be more efficient on your Mac?
Answer a few questions about how you're currently dealing with unnecessary clicks, repetitive typing and file clutter. It's FREE and takes less than 2 minutes!
🆓 Get personalized time-saving tips today!

Follow me:

MP155 - This Solopreneur Automated His Lead Gen on a Mac using AI with Lucas Erb


Introduction to Lucas Erb

Damien Schreurs: Hello, today I have the pleasure to introduce Lucas Erb. Lucas is the founder of aiexperts.com, a one-stop shop for automating small and mid-sized businesses. He's a published author, keynote speaker, and a member of LinkedIn's Distinguished Invite Only Top Voice Program with a background in computer science and big data.

He has spent the last decade in tech as both a software engineer and consultant. For the past five years, he has advised Fortune 500 executives on emerging technologies like AI and automation. Today, he helps business leaders make confident strategic bets in an increasingly fast-moving landscape. Lucas, welcome to the show.

Lucas Erb: Thank you, Damien. Thank you for having me. I'm just thrilled to be here.

Damien Schreurs: Yeah, it's a pleasure to have you, a fellow technologist. And so let's dive right into the main topic.


Lucas's MacBook Pro Setup

Damien Schreurs: Which Mac are you using to run your business?

Lucas Erb: Yeah, it's a great question. So, the MacBook Pro is where I'm at, and right now, you can't see it on screen, obviously, but to the side of my desk, I have my Mac sitting in clamshell mode and plugged into an external monitor. It's set up so that I actually have a custom-made little wooden stand for the Mac; it stands there and slots in perfectly.

Then, when I want to meet some friends for coffee or go out to a meeting or the office, I can just unplug the Mac really quickly, throw it in my backpack, and run. That's why I sometimes operate it that way. I'll actually use it open, and things like that. 

But the MacBook Pro is where it's at. It's the most powerful, and it brings together all the things I need as an on-the-go New Yorker starting a business.


Damien Schreurs: Is it a one-cable situation or do you have multiple cables—one for the monitor, one for the hardware?

Lucas Erb: Yeah. This might be a superstition—tell me, Damien, with all your experience—but I have two cables. My thought is maybe I shouldn't put all of my eggs in one basket and run everything through one Thunderbolt slot (one USB-C). So I have, you know, the microphone, the camera, and the charger plugged into one dongle that's connected and does that. Then I have the monitor itself plugged into a separate cable as well.

Damien Schreurs: Yeah, there are monitors that act as hubs, and I guess maybe your situation is that the monitor is actually powering the MacBook at the same time. But, yeah, those monitors—like the Studio Display from Apple—you basically plug in one cable; it has power and then you can plug other things into the monitor and they automatically appear on the MacBook.

Lucas Erb: You know, if I had a do-over, I would buy one of those monitors right away. Because the idea of having one cable that slides right in—even some friends of mine have one of those docks where you can slide the Mac perfectly into that one slot so there is no cable. You just get home, put it right into place, and then you're off to the races; the power is on and everything works as expected. I love that idea.

Damien Schreurs: Yeah, I really like the idea. The only thing I would be a bit worried about, for me, would be related to having a hard drive connected to the hub or connected to the monitor.

Damien Schreurs: First of all, I would be paranoid—I would have a hard drive connected 24/7 to the monitor for Time Machine backups. The issue is this: when you have external storage attached to the Mac, you need to properly eject it before disconnecting the cables, right? If you just have the camera and microphone, it doesn't matter; you plug it in and out with no big deal. But with a hard drive, that might be an important tip for the listeners.

But whenever you tell macOS with Finder to copy a file from your Mac to an external hard drive—even though it looks like the copy is done—you might think it's copied, but actually...

Lucas Erb: It's still processing.

Damien Schreurs: ...it may still be processing, depending on what else the Mac is doing. Yeah, that's the thing.

Lucas Erb: If, by chance, you're running auto-saving and auto-timekeeping and it transfers a file during ejection, you're just going to lose the file. It will be permanently corrupted, right?

Damien Schreurs: Yeah, that's the issue.

Lucas Erb: That's a big issue. I'm curious—does this issue go away a little bit with some of these solid-state drives that people are using? There’s no spinning disc inside, so perhaps the file transfers more quickly and the data is a little more persistent, meaning you’re not risking anything by unplugging too quickly.

Damien Schreurs: Yeah. So transfer speed is one thing. It’s true that SSDs transfer faster, but it's really a caching issue. The computer tells Finder that it has copied the file, but behind the scenes there’s still part of the file either in memory or not yet fully registered onto the external drive. And that’s true whether it’s an SSD or a spinning hard drive. The only advantage of an SSD is that if you unplug it by mistake, it may not affect the copy as much.

Lucas Erb: Okay, really thoughtful.

Damien Schreurs: Yeah.

Lucas Erb: I remember there was a time, years ago, when one of my hard drives broke down. I unscrewed the little box, pulled off the top, and got to see those magnetic little sliding arms that they have inside, along with the little silver metallic plates that store the data. It’s entirely fascinating how we were able to do this.

Damien Schreurs: Yeah, and the density that they're able to cram into those hard drives nowadays is amazing.

Lucas Erb: Absolutely.

Damien Schreurs: And so, if we talk about density and having space on our Mac, what is taking up space in your applications folder?


Top Apps for Running a Business

Damien Schreurs: What are the top apps that you rely upon to run your business?

Lucas Erb: There are so many—too many to count. But, in the interest of reducing application fatigue, I think we all probably have about a hundred apps that we switch between, and maybe there’s a hundred more on our "to try" list. I’m more interested in naming the few that have added the most value to me recently.

For myself, as an automation and AI guy, I'm looking for applications that, at their core, are still just very beneficial and intuitive with a user interface. They add value to my day-to-day, helping with a specific task or workflow, and then they add a little bit of AI just as the sugar on top, the cherry on top, the icing on the cake that helps things run smoother. It makes my job as the operator or user of an application less cumbersome.


Granola: Simplifying Meeting Transcriptions

Lucas Erb: I’ll share a couple of examples, maybe starting with the simplest one called Granola. I love this app. I don't know if you're familiar.

Damien Schreurs: I've heard one guest mention using Granola, yes, but I've never used it myself.

Lucas Erb: Okay, so this is one of many trying to solve the problem of meeting transcriptions. It transcribes meetings for me. The cool thing about Granola is that it has a very intuitive interface—kind of like Apple Notes. Typically, you open a new note and jot down your ideas. Granola, however, transcribes the meeting live. Then, as you take your own notes during the meeting, highlighting the key points and ideas (or even adding a screenshot if you like), after the fact, the app processes the transcript using AI and summarizes everything that happened.

You can send that off to friends, or even add automations on top that might send an automated email after your meeting. The real special thing is it takes the basic functionality I needed and makes it as simple as possible.

There are other good applications in the market, like Otter.ai or Firefly, but sometimes simplicity is the ultimate form of abundance—it’s just better when it’s simple.

Damien Schreurs: So how does it work? You're in a meeting with a note open, and you’re taking some notes yourself manually. Is Granola rewriting what you’re writing on the fly or how does it work?

Lucas Erb: There's essentially a note section for you to use during the meeting; it simply captures your notes without interjecting while you’re typing, which could be confusing. After the meeting ends, it creates a separate automated summary note.

Damien Schreurs: Okay.

Lucas Erb: The cool thing about that tab is that you can format it based on the type of meeting. Say you’re an entrepreneur pitching an investor—you might want to take elaborate notes on the investor’s feedback, questions you couldn’t answer, etc. But if you're on a call with a client, you might want the AI to focus on the customer’s pain points. The AI lets you customize that per meeting.

Damien Schreurs: Wow, so it's like a meeting template or note-taking template where you tell Granola, "Okay, this is this kind of meeting; summarize or focus your attention on these points," and it only surfaces those items from the transcript. Do I understand correctly?

Lucas Erb: That's it. Plus, you have the full transcript available if you need to review specific sentences later. They even have a live web feature so you can click in and have a conversation with those notes if needed in the future. It’s really helpful.

Damien Schreurs: I understand that it’s still outside of Apple Notes—it's a separate application—and afterwards you can copy-paste from Granola to Apple Notes if you want. 

Lucas Erb: Yep, exactly. You could even send emails with those notes, but if you want an Apple Note or a Notion note, Granola only helps you with its own note-taking; you would need to port some over if necessary.

Damien Schreurs: Okay.

Lucas Erb: Can I get a little more geeky? Can I go for another one? Okay, so now we're getting a bit technical.


AI in Software Development

Lucas Erb: A little more technical on the AI stuff—if anyone here is thinking about the latest in automation or AI for code, this is what gets me most excited. My background is in computer science, and I’ve spent a lot of time developing various applications over the years. Two quick thoughts: first, the need to know software engineering and computer science as a skill set has diminished for most businesses. It's a revolutionary concept.

You might be thinking, "This guy is crazy; there's no way we're there yet." There’s a popular graph floating around that shows software engineering job openings on Indeed.com. To all my friends in software, I’m sorry to say this is a bit miserable: there was a bell curve where interest in new software engineering positions peaked around 2021—right when ChatGPT came out. Now, open job listings have nearly vanished, and there’s much less interest in hiring more developers because AI allows developers to accomplish more with less.

One tool I’m particularly excited about is called Cursor. It’s essentially an AI pair programmer: you have the code on the left, and the AI on the right. The really special thing is that you don’t really have to write everything yourself. Some people are doing what they call "vibe coding," fully embracing that the AI will write the code to solve the problem. You sit back and become the director of a team of AI agents that write code, understand dependencies, and even install things for you. There are legitimate security concerns, and you still need to be smart about what you’re doing. 

But what this means is that people who have never built an app before might start developing applications. There’s a viral video of an eight-year-old learning to code what took me my entire first year of computer science to grasp. With enough curiosity and creativity, this child experiments with new tools and builds something truly incredible. I find that opportunity tremendous.

Damien Schreurs: I still see value in learning computer science principles, but you’re right—maybe not going as deep as I did. When I started university, I learned programming with Pascal, then Fortran, C++, and eventually settled on Python. 

Interestingly, I recently discovered another AI automation platform that relies heavily on JavaScript. I have no idea how JavaScript works, but that tool spit out some code. I gave it to ChatGPT and then to Gemini to help me understand it. I know Python and programming, so it wasn’t too difficult to reverse engineer what the code was doing.

Lucas Erb: Exactly.

Damien Schreurs: I think as long as children develop logic, critical thinking, and maybe a little collaboration, that will be plenty. And regarding job listings for programmers, there may be less need for hiring full-time programmers, but those who might not find a traditional job can sustain themselves as independent developers more easily thanks to AI.

Lucas Erb: I agree. Existing software engineers can now implement a new feature or front end 10 to 100 times faster. Solving complex underlying problems remains difficult—like Uber, for example. Uber had to figure out which of the thousand drivers around you to route to your location, optimizing route efficiency, fuel consumption, and ensuring subsequent ride requests. All these calculations are deeply mathematical. AI does not remove critical thinking or creativity; it just makes it easier to achieve once you know what needs to be built. So, if you’re someone without extensive development experience, you might not become a 100x developer, but you can certainly pick things up and build a front-end demo that gets you to a conversation and helps your idea evolve into something real.


The Future of Personalized AI Applications

Damien Schreurs: The way I see AI right now is that it fosters personalization. In the education space, for instance, we’re about to see much more personalized learning. When it comes to applications, AI will also create personalized solutions. Someone might have a very specific need and, instead of paying someone, they could simply subscribe to an AI service that builds their own custom app. It’s a market of one—a perfect solution, because nothing will be as good as the thing you create to scratch your own itch.

On the creativity side, very few people know this, but I released a chronic music album about 10 years ago on some platform—I don’t recall the name—where I challenged myself to create songs using just an iPad; it was 100% human creativity. Then, about four weeks or a month ago, I was on the couch thinking, "Hmm, could I make a song about my cat?" And I did that in 20 minutes using an app—I think it was UD or Suno. 

Lucas Erb: Suno is fantastic. I'm not completely sure of the details.

Damien Schreurs: Yes, so I provided a topic and style for the lyrics, and the AI created three iterations. One of them was particularly good.

Lucas Erb: Sometimes it makes songs that are actually good enough that I’d listen to them on Spotify—I’d hit the like button repeatedly. They’re entirely generated by AI. It begs the question: Is the ability to create uniquely beautiful or interesting music a human skill, or can it be abstracted away by algorithms? Real art will remain, much like painting remains coveted despite the advent of photography.

I do think, however, that AI will make it easier for folks like you and me—non-musicians—to create songs, which might make it a bit harder for true artists to stand out.

Damien Schreurs: Yeah, and as we’re already in the AI segment, let’s continue.


Exploring N8N for Workflow Automation

Damien Schreurs: I wanted to come back to N8N. Could you explain why you like it so much, how you use it, and how it differs from Zapier or Integromat (meg.com)?

Lucas Erb: That’s the perfect question. When you think about all these automation and workflow automation tools—Zapier is probably the best known. Many of us have been using Zapier for 10 years as small business entrepreneurs needing to, say, send an automated email to a client or automatically save a tweet to a Google note.

For a long time, automation was deterministic; that is, you had a very clear expected outcome given a specific input—if this happens, then do that. With AI, however, we can address more ambiguous problems. Instead of having code that says, “When you see a Twitter post with heading A, add it to note A; if it’s heading B, add it to note B,” you can now process a hundred random Twitter posts through an AI that organizes them into a well-structured table for later analysis.

A tactical example of N8N’s magic is that it allows for no-code, click-and-drag interfaces, letting you connect virtually any open application online. Imagine one of the automations I use for my business:


Automating Lead Generation with AI

Lucas Erb: Imagine that one of your key lead-generation methods involves researching prospective customers in a niche. First, you add a list of those customers and their business URLs into a spreadsheet. Then, for each URL, you research the appropriate contact—perhaps the operations lead, CEO, or CIO. You use an email finder application to get their contact details, then extract their LinkedIn URL, and finally reach out with a personalized message, explaining how your app can help their business.

What hasn’t been done before at the scale we’re seeing now with automation tools and AI is fully automating that entire outbound process. For my business, I can plug in a list of URLs—say, targeting auto dealerships in the US—and have an N8N agent process each URL. It can automatically extract the names and roles (CIO, CEO, CTO, Chief Operating Officer), verify emails, gather LinkedIn information, and craft a personalized message for each contact saying, “I would love to help you—this isn’t some sleazy automation, but a genuine offer to work together.” It used to take hundreds of hours to implement this; now it’s automated with AI.


Damien Schreurs: I am offering AI training here in Luxembourg—that’s my main business. I’m now offering artificial intelligence training, among other things.


Understanding AI Agents vs. Assistants

Damien Schreurs: My pet peeve right now is when people misuse the word “agent” when they really mean “assistant.”

Lucas Erb: Right.

Damien Schreurs: I often get asked, “What’s an agent? What’s the difference between an agent and an assistant?” For me, an "agent" implies agency—i.e., the ability to make decisions, some level of autonomy. If an AI system can make decisions on your behalf, it’s an agent; if not, it's merely an assistant. The tool I mentioned earlier is called Mind Studio, which advertises itself as a way to create agents with no-code or low-code. Although I still needed some JavaScript in the loop, the system is totally deterministic: I've set up all the if-this-then-that logic, and the only AI component is the input-output processing that proofs text—similar to copying and pasting text into ChatGPT for proofreading. I tried to automate the process for my YouTube subtitles. 

When I feed ChatGPT or Claude an SRT file with timestamps and text, I ask it only to proofread the text part so as not to mess up the formatting. I realized I couldn’t do that in one go—I had to split it into small parts. The moral of the story: I thought I had created an agent, but I ended up creating an assistant.


Advanced AI Automation Tools

Damien Schreurs: From what I have seen with N8N—for those unfamiliar with Zapier or Integromat—it essentially uses different modules where you plug one after another. Previously, with Zapier, you could only create simple actions like “Create a post” or “Save a row in a Google Sheet.” With ChatGPT, however, we can embed AI actions. We now give prompts plus additional context from previous steps. But as you said, it’s still deterministic.

What I’ve seen with N8N is that it includes a special AI node with multiple branches. In Zapier, each box has one input and one output, but with N8N, you can have multiple inputs and outputs along with dynamic decision-making, allowing the prompt to guide how decisions are made.

Lucas Erb: In short, prompting the AI is how you introduce decision making. For example, if you’re posting content on the internet, part of your process involves researching current news. Say you’re a sports poster who tweets daily about sports news—the decision-making isn’t as simple as “if tweet A then post note A.” Instead, you might have an agent that sifts through hundreds of articles, narrows them down from 100 to 50 to 25 to 10, and then selects the most newsworthy article for you to post.


Deep Research and AI Efficiency

Lucas Erb: This is possible because we now have deep reasoning agents. Today’s agents go beyond the early ChatGPT era of predicting the next character—they are more aligned with our goals and desires. They’ve proven to be highly persuasive, sometimes even more so than humans, mathematically competent, and capable of performing tasks like passing the bar exam—something I, as a human, would likely never achieve.

This opens up a world where we could allow an agent to make decisions for us—or at least reduce our workload, leaving the final decision to us (a "human in the loop"). For example, an AI agent could populate a Google Sheet with suggested email texts for outreach, and you could then tweak the generated text before sending it.

Damien Schreurs: Exactly. Every decision that you would have made manually can now be aided by AI. This goes beyond simple automation; many folks are working to build general-purpose intelligence for online use. We might soon have a world where you can simply speak into ChatGPT or Claude and say, “Hey, visit Airbnb and pull up the 10 most interesting A-frame cabins in New Jersey, as I want a break from the city.” While I like to think through decisions myself, the more we trust AI to narrow down options for us, the more we can save time and mental energy.

Damien Schreurs: I mentioned in my training sessions that the easiest way to get your feet wet with agents is to run a deep research query in Gemini or ChatGPT, and you’ll be amazed by the number of articles and websites it retrieves. I once ran a deep research query for 21 minutes about the US tax system.

Lucas Erb: Wow.

Damien Schreurs: It sifted through more than 60 sources. If I had tried to do it myself, it would have taken at least half a day, maybe a full day. I might have even gotten distracted—got a cup of tea, let my cat in, then checked back. It saved me a tremendous amount of time.

Lucas Erb: Incredible.


Practical AI Tips for Solopreneurs

Damien Schreurs: We’re almost at the end of the show. What tip or trick did you discover on your Mac recently that you wish you had known before?

Lucas Erb: Such a good question. One of my favorite tools that always gets a “wow” reaction from friends is a little Chrome extension. I love Chrome extensions because they can take your web browsing experience and make it just a little bit—say, 1%—better. Because many of us use our browsers for everything, a small improvement can be very valuable. One Chrome extension I frequently use is called One Tab, which I’m not sure if you use, Damien.

Damien Schreurs: I know about it; I don't use it much, but I’m aware of it.

Lucas Erb: One Tab gets lots of reactions because, having grown up with computers and computer science, I’ve seen engineers leave their browsers open all year long—hundreds of tiny tabs, each only an icon across the screen. For those with powerful machines (like M2 or M3 Macs), it’s manageable, but it gets confusing after a while.

With One Tab, you simply click a button to collapse all open tabs into one. It saves the links in a nested structure so you can later reopen them all with one click. This helps compartmentalize your work—for instance, if you’re working on a project that requires a Google Sheet, YouTube research, and ChatGPT conversations, you can open those tabs, do your work, and then collapse them into One Tab when you're done. It declutters both your workflow and your mind. I hadn’t known of another way to handle this until One Tab existed.

Damien Schreurs: That’s a great tip. I’m guilty of having many tabs open, and I sometimes manage the issue by using different profiles in Chrome, Edge, or Safari. I diffuse the problem by switching browsers.

Lucas Erb: Yep.

Damien Schreurs: One more tip is using profiles to manage tabs efficiently.

Lucas Erb: I also want to mention another tip—the high-quality video you saw on screen. Your tip about using an iPhone as your video camera is tremendous. We all have a 4K camera in our pockets now—a high-quality, DSLR-type camera. Being able to connect the iPhone for use as a webcam is possibly the most efficient and ingenious trick I’ve heard in a while.

Damien Schreurs: Mm-hmm. I agree. I wouldn’t use my iMac’s built-in webcam, even though it’s a 2020 iMac with a 1080p camera. The iPhone’s camera is far richer. That’s the beauty of the Apple ecosystem: the iPhone communicates so well with the Mac. In macOS, I have a little button that quickly lets me change the appearance, activate the bokeh effect or portrait mode, and even use Studio Light. I'm using Studio Light right now.

Lucas Erb: Tremendous.

Damien Schreurs: Last question: if you were in charge of the Mac division at Apple, what would be your first priority—whether on the hardware side or the software side?

Lucas Erb: It’s a big red flag right now for Apple. If anyone at Apple is listening, consider this: recent news is that Johnny, the man behind so many beautiful Apple products we know and love, has recently solidified a massive multi-billion dollar deal with OpenAI. They plan to build a new device at the intersection of beautiful, applesque design and AI-driven conversational agents that quickly access information.

I think the feature I’d request is this: can we envision what a fully integrated AI experience on a Mac would look like? Imagine applying Apple’s best practices in security, beautiful and powerful hardware design, and creating a truly AI-native macOS and iOS interface, re-engineered from the ground up. I believe we haven't yet grasped how revolutionary AI will be in the way we interact with computers—and Apple has the best chance of creating a compelling AI-first experience.

Damien Schreurs: Yeah, I agree. I don’t think they’re making that mistake. The mistake would be trying to compete head-to-head with OpenAI, Anthropic, and Google. But an AI-native macOS and iOS interface is missing at the moment, and if they can figure that out, they will be well-played.

Lucas Erb: Agreed.

Damien Schreurs: And that nicely wraps up the theme of today’s episode.


Final Thoughts and Contact Information

Lucas Erb: If I can give a quick shout-out—if anyone listening wants to dive deeper, Damien is a great contact. But please reach out to me as well. Our goal is to help people move beyond just using ChatGPT and Claude. These tools are extremely powerful, but there are ways to implement them into your workflow that save on efficiency, productivity, time, and ultimately, money.

Back in the early days of the internet, maybe your business didn’t have a website because it wasn’t necessary. But now, having a website, an Instagram presence, and a LinkedIn presence is table stakes. I think we’re about to see a similar revolution with AI. Every business will need to rethink not just using the latest tools, but building feature-rich AI solutions that solve their unique problems in a way that enriches—rather than replaces—human connection. There's tremendous opportunity out there, and I’d love to talk to anyone who has questions.

Damien Schreurs: Very good. So, Lucas, where can people find you online?

Lucas Erb: I’m frequently on LinkedIn—search for Lucas Erb—and you’ll find me on a profile with a robot dog. I’m also very available via email at lucas@aiexperts.com. And you can visit aiexperts.com for more information.

Damien Schreurs: Great, we’ll put a link in the show notes and in the description of the YouTube video.

Lucas Erb: Thank you, Damien. Thank you for having me.

Damien Schreurs: Thank you very much, Lucas.

So if, like Lucas, you would like to share how you are using your Mac to run your solopreneur business, it's simple: just visit Macpreneur.com/apply

If you are on PodMatch, just click the button and you will land on the show profile. Otherwise, fill out the application form and I will get back to you within a few days. Once again, that’s Macpreneur.com/apply.

Until next time, I'm Damien Schreurs, wishing you a great day.