If you’re still thinking your coding skills alone will carry your AI career for the next decade, I have some uncomfortable news. I recently interviewed an AI engineer who’s been building systems for over 40 years, working with companies like Airbus, Disney, and dozens of luxury brands. And his advice might surprise you because we barely talked about code.

The first thing he told me was shocking. His number one foundational skill had nothing to do with information technology. Instead, he focused on becoming a subject matter expert in specific domains. And this single strategy helped him survive multiple AI winters and thrive through every hype cycle.

The Factory Floor Beats the Algorithm

Here’s a real story that illustrates this perfectly. He was approached by a major garment company in the 1980s to optimize their fabric cutting process. The person doing the calculations manually was about to retire, and they were stuck. But here’s the thing: he knew absolutely nothing about the textile industry. He didn’t even understand the vocabulary they were using.

So what did he do? He didn’t start with the code. He went physically into the factory. Every day after work, around 4 or 5 PM, he would show up with big pieces of paper and just watch how people worked. He learned the names of machines, their vocabulary, their constraints. He even convinced workers to let him use the machines during night shifts so he could cut fabric and roll it himself.

Through this process, he discovered around 10 real-life parameters that textbook algorithms completely missed. Things like margins you have to leave, maximum lengths for certain fabrics, and how the dyeing process affects everything. These weren’t in any computer science book. They were in the minds and hands of the people doing the work.

The result? He delivered a working system in three months. But more importantly, he became a subject matter expert in textile manufacturing. He sold that solution to every single major luxury company in France. At one point, he joked that no one in France was wearing something he hadn’t planned or worked on.

Your Ticket to the Future

Think about what happened there. He didn’t just build one system. He became THE expert that companies called when they needed textile optimization. When the next company approached him, he could honestly say he’d worked with 50 luxury brands. Who could compete with that level of domain knowledge?

This is what he called becoming a “theme” or subject matter expert. It’s your ticket to the future because you’re building something AI can’t easily replicate: deep contextual understanding of a specific industry combined with the relationships and trust that come from actually solving problems.

If you’re wondering how to apply this as a junior or mid-level engineer who’s mostly coding right now, his advice was straightforward. Find a domain that requires a lot of automation. He mentioned maintenance as a perfect example because everyone needs it, from airplanes to AI systems themselves. Plus, you can’t fully replace maintenance workers because physical work is still involved.

The beauty of this approach is that you can build your AI engineering career around industries where you’re not hurting people, where you’re augmenting rather than replacing. And that builds the kind of reputation that lasts decades.

Building Your Knowledge Repository

Here’s where it gets really interesting for today’s AI engineers. He mentioned that if he were starting now, he’d take all that domain knowledge and put it into a vector store or RAG system before even thinking about the AI implementation. All the rules, all the instructions, everything the workers told him about how things actually work.

Think about that for a moment. You could work with 20 companies in a specific domain, building up this incredible dataset of how things actually work in the real world. That knowledge base alone is worth millions, even before you add generative AI on top of it.

This aligns perfectly with what I’ve seen in AI agent development, where understanding the domain and the process is far more valuable than knowing the latest framework. The technology changes constantly, but the fundamental problems in manufacturing, logistics, healthcare, or any other industry remain remarkably consistent.

The Path Forward

So where does this leave you as an AI engineer? If coding is becoming commoditized and AI can generate implementations, your differentiation comes from understanding domains deeply. Pick an industry. Spend time with the people doing the work. Learn their vocabulary, their constraints, their exceptions.

Don’t just build solutions from your desk based on what you think the problem is. Go to the factory floor, the warehouse, the clinic, wherever the actual work happens. Understanding AI career paths means recognizing that technical skills are table stakes, but domain expertise is the competitive advantage.

The engineers who thrive over the next 40 years won’t be the ones who know the most about the latest AI model. They’ll be the ones who deeply understand specific industries and can apply whatever technology makes sense to solve real problems.

To see exactly how this veteran AI engineer built his career and survived multiple AI winters, watch the full video tutorial on YouTube. I walk through the complete interview where he shares decades of real-world experience. If you’re interested in learning more about building a sustainable AI engineering career, join the AI Engineering community where we share insights, resources, and support for your learning journey.