Magic Tools

My dad is a structural engineer—he spent 50 years of his life designing buildings, bridges, and railroads. On a long car ride, we had a conversation about no-code, AI, and how the process of building software is changing.

Software engineering and structural engineering are of course not the same—but it feels like at this moment many people would benefit from a parallel.

My dad started his career drafting building plans with pen, pencil, and large sheets of drafting paper—doing calculations by hand in the margins. Over the course of his career several "magic tools" were introduced that vowed to completely change his profession.

• Calculators that spit out the correct answer to any equation, every time
• Personal computers
• AutoCAD

With each came many questions about how these new tools would affect the job of the structural engineer. When I asked him what impact these tools ultimately had on his job, his answer came quickly:

"Geoff, these things only made the engineers who could do the math themselves, who understood the underlying concepts, who understood physics that much more valuable. Suddenly a new generation of engineers was making decisions that would dictate whether a building would stand or fall based on a number spit out of a machine—with little ability to assess whether that number even made sense. I made a career out of catching what could have been catastrophic mistakes because I had done the work myself and understood the engineering."

I think of my dad's comments often as I scroll through a Twitter feed that declares no-code to be dead, that muses that the job of the software engineer will soon be extinct, and that is far too quick to usher in the era where "anyone can build anything."

I can build a lean-to in the woods; does that mean I can build a skyscraper? Not any more so than dishing out a dose of Advil makes you a doctor. And the commonality between people making such declarations is clear—they're not experienced engineers. Building an internal tool or a job board rapidly is awesome, but it's a far cry from building a more serious piece of software. We need a little more self-awareness about where we stand.

Calculators, computers, and AutoCAD all changed how my dad did his job—these magic tools made him more effective. But he excelled in his profession because he knew the engineering—in the absence of these magic tools, he could still do the work.

We should keep trying to build even better magic tools—it's important that we do. But let's get a grip on all the silly declarations we make about them.

The rewards will go—as they always have—to those who treat magic tools as a force multiplier rather than expecting them to act as a wizard.