The Craftsman and the Code
To explain what’s happening with AI in software development, it helps to look at how a developer’s work is changing - how it was done, how it’s being done now, and where it’s heading.
Instead of computers, let’s talk about building furniture. It’s not a perfect analogy, but it works.
At some point in history, if you wanted furniture you needed to talk to a skilled craftsman. Anyone who has taken a shot at making simple furniture can attest that it takes more than hammering a few boards together. There are specialized tools and materials, each with their own strengths and weaknesses.
You have to know about the individual parts, how to make them, and how to put them together. There may be a dozen ways to do any particular step. Learning which materials to use, how to manipulate them, and how to design and make things from the beginning is a craft that can take most of a lifetime to master.
If you want anything particularly nice, like intricate carvings or special shapes (curved legs on your table or chair), then you need even more tools, and even more knowledge and practice.
Even simple things are a lot more difficult than they seem. Sawing a board into two pieces isn’t that easy. Hammering in nails is not as easy as you might think by watching. Plenty of nails get bent during the process.
It takes a lot of practice to develop any proficiency with those tools. If a woodworker wants to make furniture for other people, they have to learn and practice even more so that they can deliver what a customer wants.
This discussion will gloss over the mass-manufacturing stage of making furniture. There are some parallels with software development but this is already long enough.
The Rise of the Machine
Enter the CNC machine. These are machines which take CAD drawings - digital representations of shapes - and use those drawings to guide machinery to carve the shapes out of wood (or plastic, metal, and more).
Someone with a CNC machine doesn’t need to learn the intricacies of how to manipulate a jigsaw around a complex curve, or how to measure wood down to 1/16th of an inch. The machine does all of that for you. Also significant is that it can do it over and over again with perfect replication.
A CNC replaces a lot of tools, as well as the knowledge of how to use those tools.
That doesn’t mean it’s all automatic. You still need to know how to design those shapes, and how to put them all together. You still need to have a good understanding of the materials. Additionally, you have to know how to use CAD to design those shapes to begin with. CNC machines have their own considerations which you have to learn.
Ultimately, though, using a CNC machine is faster and more accurate than doing things by hand. A lot of traditional woodworking skills are no longer as important.
AI as a Digital CNC
With AI, software development is going through a similar process.
Until just a couple of years ago, writing software required a lot of manual labor. A project that might have taken months of effort can now be completed in a fraction of the time. A large, complex feature can be built in about one-third of the time it would have taken previously. That’s because AI - the CNC of software development - can do the manual work.
That doesn’t mean that AI can do all the work. The developer still has to know a lot about how things should be put together. It’s less effort, but it isn’t zero effort. Just because someone knows how to use a CNC machine doesn’t mean they’re going to know how to design the right pieces or how to properly assemble them.
On the other hand, work that used to require a team of four or more people can now potentially be handled by one.
The Skill Doesn’t Vanish, It Transforms
The core message is that powerful tools don’t eliminate the need for skill; they just change where the skill is applied. A CNC operator may not need the muscle memory to saw a straight line, but they must master CAD, understand material tolerances, and know how to assemble the final product. The craftsmanship moves from the hand to the mind.
In the same way, a developer using AI writes less manual code but takes on a more critical role as an architect and a quality gatekeeper. They must design the overall system, guide the AI, review its output, and integrate the final components. The skill is less about the act of typing and more about the vision for the final product.
A powerful tool in the hands of a novice will only help them make a mess faster. In the hands of an expert, however, that same tool allows them to achieve results that were previously impractical or impossible. Both woodworking and software development still belong to the skilled craftsperson, even if their tools look very different than they did a decade ago.
