Lately, there has been a lot of panic about the software industry being dead, supposedly threatened to its core by the rapid rise of artificial intelligence. But is it really over? As a fifty-year-old "dinosaur" solution provider who has spent decades in the trenches, I can tell you one thing: software isn't dying. It is simply doing what it has always done: evolving.
The short version
While AI tools are incredibly powerful, they only solve technical friction, not conceptual confusion. Just like the visual development boom of the 1990s and the low-code hype of the 2000s, the current AI wave will ultimately increase the demand for skilled engineers who understand the complex "micro-rules" of business processes.
Why Previous Automation Panics Failed
I wrote my very first line of code when I was ten years old. Today, hundreds of thousands of lines of my code are running in factories all over the world. This isn't the first time I've seen the industry panic over its own demise. In fact, these cycles of anxiety are as old as computing itself.
Take the 1950s, for example. When high-level languages like Fortran first appeared, people argued that machine language was obsolete. They claimed that because anyone could write commands in English-like syntax, the need for specialized programmers would vanish. Of course, that didn't happen.
I witnessed a similar shift myself in the 1990s with the rise of Rapid Application Development (RAD) and visual tools like Visual Basic, Delphi, and Power Builder. Suddenly, we went from writing code line-by-line to dragging and dropping visual components. The skeptics immediately asked: "If an accountant can design their own form and slap a couple of lines of logic behind it, why do we need computer engineers?"
But businesses soon learned a hard lesson. Well-meaning amateurs built unscalable, fragile systems full of spaghetti code. When these systems inevitably broke, companies came running back to professional engineers, practically begging them to untangle the mess. A similar cycle repeated in the 2000s with low-code platforms, under the promise that business analysts would build their own applications. It didn't work out that way either.
The Reality of Our Recent AI Experiment
Tools can solve technical friction, but they cannot solve conceptual confusion. In fact, every new tool adds its own layer of mental complexity. The hardest part of programming has never been typing the code; it is knowing exactly what you want to build. It requires analyzing deep business processes and managing thousands of tiny, delicate rules.
To see how this plays out with modern artificial intelligence, we recently ran an experiment at our company. We asked a colleague from another department, someone with absolutely no prior experience in our specific domain, to build a simple project using an AI agent. We gave them a task description and a couple of sample projects to guide the tool.
The result? It was shockingly, almost terrifyingly good. The AI didn't just complete the task; it identified missing details in our prompt, learned from the sample projects, took notes, and even flagged inconsistencies in older code, offering to fix them if ordered to do so. That is the scary part.
Why the Human Element Still Matters
But here is the catch: the colleague who ran the AI agent has no idea what the generated code actually does. While they have a broad overview, they don't understand the hyper-specific "micro-rules" that govern our department's operations. These micro-rules aren't just industry-wide standards; they vary from company to company, and even department to department.
As AI-generated "synthetic" code grows exponentially, so will the silent collisions between this code and those unwritten micro-rules. Because there will be far more code in existence, these clashes will happen more frequently than ever before. When they do, businesses will once again find themselves calling out for experienced developers to rescue them from the chaos.
The software sector isn't ending. For the old wolves of the industry, this is simply a time to sniff the wind, peer through the fog, and guide the way forward. It won't be easy, but we've been here before.
By keltox
Ekşi Sözlük Contributor
What are your thoughts on this topic?
Every article is an open conversation. Whether you have a counter-argument, a local example, or a different perspective based on your own experience, your contribution makes this space better.
0 Comments