How your real-world experience is becoming the most valuable skill in tech
I started a web design shop in 1996, while I was in the Navy.
The idea was simple. Small businesses in my community needed websites, and I could build them. Back then, that meant writing code by hand in a basic text program, one line at a time. I’d sit across from a florist, a dentist, a plumber. People who knew their businesses inside and out. My job was to take what they knew and turn it into something that worked on a screen. And my biggest challenge was never the technology. It was understanding their business well enough to build something that actually helped them.
Then Netscape released a tool called Composer.
Suddenly, the skill that made me valuable (writing website code by hand) became something anyone could do. Point, click, drag, drop, publish. No coding required. The playing field shifted overnight. A lot of people asked the obvious question: if anyone can build a website now, what happens to the people who were building them before?
Here’s what actually happened. The people who thrived weren’t the best coders. They were the people who understood their clients’ businesses deeply enough to build something worth building. The technology got easier for everyone. But the understanding of what to build and why? That stayed rare.
I’m watching the exact same thing happen right now. And this time, it’s much bigger.
From “Vibe Coding” to Something More Serious
In February 2025, a well-known AI scientist named Andrej Karpathy posted about a new way of building software. He called it “vibe coding.” The idea: you talk to an AI tool in plain English, tell it what you want, and it writes all the code for you. You don’t even need to read what it produces. Just describe what you need, let the AI build it, and keep giving it feedback until it works. Collins Dictionary made “vibe coding” their Word of the Year for 2025.
New tools popped up to make this real. The most beginner-friendly one is called Replit. You open it in your web browser (no software to install), and you describe what you want. Something like, “I need a reservation system where customers can pick a time and get a confirmation text.” The AI builds it. Not a picture of an app. A real, working app that stores information, looks good, and can go live on the internet.
But vibe coding was always a starting point. Karpathy himself recently said the term is already outdated. He now prefers “agentic engineering,” which basically means: AI still does the building, but with more care, better quality, and real human oversight. The fun experimental phase is growing into something you can actually run a business on.
This shift matters because it changes who gets to build software. And it changes what kind of knowledge matters most when you’re building it.
What I Call “Domain-Informed Design”
Here’s the idea I want to put into the world: domain-informed design.
For a long time, the problem with software wasn’t a lack of good ideas. It was that building those ideas required specialized coding skills. You could have twenty years of experience running restaurants, managing hotels, coordinating wellness retreats, or operating small businesses. But if you couldn’t write code (or couldn’t afford to hire someone who could), your knowledge stayed stuck in your head. The people who built software were engineers. The people who understood the actual problems were everyone else. And the gap between those two groups gave us a whole generation of cookie-cutter software that sort of worked for everybody but really worked for nobody.
Domain-informed design turns this on its head. It starts with a simple belief: the most valuable person in the room is not the one who knows how to build the app. It’s the one who knows what the app should do, who it’s for, and why everything else on the market falls short.
Think about it. A restaurant manager who spent fifteen years watching staff juggle paper waitlists and clunky software knows something that no engineer can learn from a meeting. A luxury travel advisor who understands that a wealthy client’s assistant needs to coordinate a private chef, a helicopter, and a spa appointment across three time zones knows something that no big booking website will ever handle with a dropdown menu. A wellness retreat operator who tracks each guest’s meals, treatments, and follow-up programs knows something that off-the-shelf wellness software was never designed for.
That kind of knowledge has always been valuable. What’s new is that you can now turn it directly into a working product. The person with the insight can be the person with the app.
What This Means If You’re Thinking About a Career Change
Let me be straightforward about the opportunity here.
If you have real experience in any field (hospitality, healthcare, real estate, education, fitness, farming, law, delivery and logistics, retail, or anything else) and you’ve ever thought, “Why doesn’t somebody build a tool that actually works the way we work?” the answer today is: you can build it yourself.
The tools are ready. Replit lets you create websites and iPhone apps right from your browser just by describing what you want. It sets up the information storage, the look and feel, and puts your app live on the internet. The AI picks the right technology and writes the code. Your job is simply to know what needs to be built and to keep guiding the AI until it gets there.
The money works. A small business that would never spend $50,000 on custom software will happily pay $2,000 to $5,000 for an app that solves a real problem for them, especially when the person who built it actually understands their industry. Monthly fees of $200 to $500 for keeping the app running and updated are easy for a business that’s saving hours of headaches every week.
And you don’t have to do it all alone. New marketplaces are popping up where you can hire experienced developers by the hour to help with the parts that are beyond what the AI can do right now. Fiverr and Upwork both have categories specifically for people who fix and improve AI-built apps. A site called Match.dev will pair you with a vetted developer within 48 hours. Another called Vibe Codors lets people buy and sell entire AI-built projects. Think of hiring a developer for a few hours the same way you’d hire an accountant to look over your books. You don’t need one full time. You just need one when it counts.
And here’s the best part: these AI tools get noticeably better every few months. Something that requires a professional developer today might be something you can handle on your own by the end of the year. The more you learn now, the further ahead you’ll be when the tools catch up.
My 1998 Lesson, Applied to 2026
When Netscape Composer showed up and anyone could drag and drop a website together, the first reaction was fear. People thought it would kill the web design business. It didn’t. What it killed was the business of only being able to write code. The business of truly understanding clients and building things that worked for real businesses? That thrived.
I see the same pattern playing out now, but on a much bigger scale. These new AI tools will make the building part of software available to everyone. Knowing how to write code will stop being the thing that keeps most people out. But knowing a business or industry deeply enough to design software that truly fits? That becomes more valuable, not less.
This is what excites me most. For the first time in the history of software, the advantage goes to the person who doesn’t know how to code but deeply knows the problem they’re solving. Not knowing engineering isn’t a weakness anymore. It might actually be a strength. You think about the problem, not the technical details. You design from the real world, not from a programmer’s habits. And the AI handles the translation between the two.
A More Humane Future
There’s a bigger story here, and it’s one I care about a lot.
For years, automation has been talked about as a threat. Machines replacing people. AI taking jobs. Technology as something that makes humans less important. And there are real concerns worth paying attention to.
But there’s another way to look at this moment. What if this is technology finally catching up to what people have always been capable of?
Think about all the people with great ideas who never built them because they couldn’t code. Think about the small-business owners running their operations on spreadsheets and sticky notes because custom software was too expensive. Think about all the people who assumed tech was a closed door because they didn’t have the “right” background or the “right” degree.
Domain-informed design opens that door. It says: your experience matters. Your understanding of real human problems matters. Your ability to sit across from a business owner and truly get what their days are like, what frustrates them, what they wish worked better? That’s not a “soft” skill. That’s the skill. That’s what makes technology feel human instead of cold.
The best software has always been built by people who understand the humans it serves. We just spent a few decades in a world where the barrier to building was so high that we forgot this. Now the barrier is coming down, and we get to remember something important: technology works best when it’s shaped by the people closest to the problem.
That’s the future I’m working toward. Not a world where AI replaces human judgment, but one where it gives human understanding more reach. Where someone who spent twenty years in hospitality or healthcare or education gets to turn that experience directly into tools that make those industries better. Where the real advantage goes to empathy, to deep knowledge, and to the willingness to listen closely to what people actually need.
Where to Start
If any of this speaks to you, if you’ve been thinking about a career change, or if you have industry knowledge that you’ve never thought of as a “tech skill,” here’s a simple path forward.
Start with Replit. Go to replit.com. It’s free to try. It’s the friendliest tool for building real apps without any coding background. Make something small for yourself first. A tip calculator. A personal schedule tracker. Just get comfortable with how it feels to tell an AI what to build.
Learn a little bit about how apps work. You don’t need a degree. Just spend a few weeks watching free videos on YouTube or sites like freeCodeCamp. Learn what a database is (it’s just where your app stores information). Learn what makes a website talk to a server. Even a basic understanding of these things will make you much more effective at guiding the AI.
Build something for your industry. Use what you know to create an app that solves a real problem you’ve actually lived through. This becomes your calling card. The thing you show potential clients to prove you understand their world.
Get help when you need it. When you run into something the AI can’t handle (and you will), hire a developer for a few hours through Fiverr, Upwork, or Match.dev. This costs a fraction of what full-time help would, and it keeps you moving.
Call yourself what you are: a specialist. Don’t say “I build apps.” Say “I build custom tools for restaurants” or “I create technology for wellness retreats.” Being specific builds trust, lets you charge more, and leans into the one thing no AI can copy: your real-world experience.
The Window Is Open
I’ve seen a window like this open before, back in 1998, when the web was brand new and anyone with a little knowledge and a lot of energy could build something real. That window eventually got smaller as the industry grew up and got more competitive.
This window is wider. What you can build through it is more powerful. But it won’t stay this wide forever. The people who start now, who learn the tools and build their reputation while the field is still young, will have a huge head start over everyone who waits.
The technology is here. The support networks are forming. The businesses that need what you can build are all around you, still running on spreadsheets and generic software that wasn’t made for them.
The only piece that’s missing is you deciding to step through.
This post is part of an ongoing conversation at betterplaces.blog about building a more humane future alongside the technologies that are changing how we work and live. If you’re exploring a career change into building technology for the industries you know best, I’d love to hear from you.
Better Places 