What Construction Taught Me About AI Tools
Why mastering your craft beats chasing the latest AI model
I’ve been thinking lately about a conversation that’s happening everywhere in knowledge work: how do we actually get good at using AI? The answer, surprisingly, came to me from a construction site I worked on years ago, watching a veteran carpenter choose his hammer.
Lessons from the job site
Right after high school and during part of college, I spent time working construction projects. I started as a helper with various trades, working alongside plumbers and sheet metal workers, essentially doing whatever needed to be done around the shop. Later, I moved into framing and learned house construction, working on both commercial and retail projects.
What struck me most about working with these experienced tradesmen wasn’t just their skill, it was their relationship with their tools. These weren’t people who had read about carpentry or plumbing in a book and then shown up to work. They had learned by doing, which is exactly why apprentice programs exist. You simply cannot read your way into this kind of expertise.
But here’s what really caught my attention: every skilled worker had their own hammer. Not just any hammer, but a specific weight, style, and claw type that worked for them. They would invest in these tools, take care of them, and genuinely cherish them. When a tool broke or went missing, work stopped until they could get a replacement.
These workers also had a no-nonsense approach to getting things done. Sure, there was joking around and camaraderie, but the underlying philosophy was clear: work before play. Get the job done, and get it done right.
The knowledge work parallel
Fast forward to today, and I see the exact same dynamic playing out with AI tools in knowledge work. These tools are fundamentally enabling technologies that help us achieve outcomes and deliver results. The frontier models are advancing seemingly every week, and the capabilities are genuinely amazing.
But here’s where the construction analogy becomes powerful: it’s not about having the shiniest, most expensive hammer. It’s about how you wield the tool and, crucially, understanding the craft you’re trying to practice.
The most effective knowledge workers I know who use AI aren’t the ones obsessing over which model has the highest benchmark scores or the latest features. They’re the ones who have developed fundamental knowledge through actually doing the work. They understand their craft deeply enough to know when the AI is helping and when it might be leading them astray.
The craft still matters
This reminds me of learning math in school. Teachers would always say “show your work,” even when you could get to the answer faster without it. At the time, it felt tedious and unnecessary. But showing your work meant you actually understood the underlying principles, not just the shortcuts.
The same principle applies to knowledge work with AI. I’m using technology right now to help me format and refine this very article. But that doesn’t mean I don’t know how to write without it. The underlying craft knowledge, the ability to structure thoughts, develop arguments, and communicate clearly, those skills remain essential.
When you have that foundation, AI becomes like that perfectly weighted hammer in the hands of an experienced carpenter. It amplifies your existing capabilities rather than replacing your need to understand what you’re building.
Beyond efficiency: the joy of craft
There’s something else I learned from those construction projects that applies here: there’s genuine joy and artistry in manual craft work. The satisfaction of framing a wall square, or seeing how an experienced plumber solves a tricky routing problem, that’s not just about efficiency. It’s about taking pride in work well done.
I think we risk losing something important if we view AI purely as a way to shortcut or eliminate the learning process. Yes, these tools can handle much of the heavy lifting, and they’re getting easier to use all the time. But there’s still value in understanding your craft deeply enough to do it by hand when needed.
The best knowledge workers I know treat AI like those skilled tradesmen treated their tools: as carefully chosen instruments that amplify their existing expertise, not as replacements for developing that expertise in the first place.
Choosing your hammer
So what does this mean practically? Instead of chasing the latest AI tool or trying to find the “best” option, focus on developing your underlying craft. Learn to write well, think clearly, solve problems methodically. Understand your domain deeply enough that you can recognize when an AI suggestion makes sense and when it doesn’t.
Then, like those experienced carpenters, find the tools that work best for how you think and work. Invest time in learning them properly. Take care of them. But remember that the tool is only as good as the craftsperson wielding it.
What’s your hammer? And more importantly, how are you developing the expertise to use it masterfully?
What parallels do you see between traditional craftsmanship and knowledge work in your field? I’d love to hear about the tools and practices that have become essential to your work.