The divide around AI in software is not really about code quality, scalability, or whether a model can write a production application.
Those are surface arguments.
At its core, this debate is about communication. And more uncomfortably, it is about ego, learning, and the shelf life of skill.
I have been working with AI systems since the beginning, long before they were polished, conversational, or impressive to watch. Back then, working with them was an exercise in patience and perseverance. You did not prompt once and get brilliance. You prompted, failed, rephrased, corrected, and tried again. Over and over.
What I noticed early on was strange.
Junior developers loved them. Senior developers often despised them.
AI is not exposing bad developers.
It is exposing bad communicators.
The seniors would curse the outputs. They would cite horror stories. Bad code. Poor abstractions. Systems that would never scale. And you hear the same refrain from writers, artists, and filmmakers. AI is garbage. AI cannot write a real novel. AI will never make a good movie.
But these reactions are deeply subjective. And they reveal far more about the individual than they do about the technology.
This all goes to the heart of what it means to be a constant learner.
And where ego shifts from healthy self respect into something more brittle. A place where acknowledgment turns into resistance. Where the unspoken fear is not that AI is bad, but that one’s existing skills are no longer sufficient on their own.
The Uncomfortable Truth About Expertise
In my own work, I once described the modern developer as both the janitor and the architect as you work with AI. You clean up messes it makes and you design systems it builds. That still largely holds true. But AI adds another role that many people are uncomfortable with.
You are now also a teacher and the student.
I started coding at age twelve. My father had to enroll me in college courses just so I could learn programming because there was no structured path for someone my age. I never took a normal route into this industry. There was no clean ladder. No framework. No best practices handed to me.
Because of that, I was forced early on to articulate what my code was doing. Not just to computers, but to people. I experienced the frustration of working with junior developers who did not yet have the mental scaffolding to see what I saw. I learned quickly that if I could not explain it clearly, the failure was not theirs. It was mine.
Structure was not there in the early days. You had to build it yourself. And you had to learn empathy for those still learning.
Over time, something clicked.
I began to recognize myself in the AI systems I was working with.
They behave exactly like a capable but inexperienced junior developer. They do not lack intelligence. They lack context. They do not lack potential. They lack articulation from the person guiding them.
When I say you must articulate what you want clearly, I am not being poetic. I am being literal. Communication is the difference between generic output and something that feels intentional. It is the difference between mediocrity and real craft.
And this is where I believe much of the hostility comes from.
For the first time, many so called experts are being forced into a room where the only way forward is to explain themselves. Not through status. Not through reputation. Not through seniority. But through dialogue.
The prompt is the room.
The AI is the junior.
And suddenly, the pedestal disappears.
For the first time, expertise is no longer proven by status,
but by the ability to explain yourself clearly.
For people who have coasted on expertise without needing to teach, this is deeply uncomfortable. They do not feel they should have to explain themselves. They do not feel they should have to train something that might eventually replace them. And they resent the implication that mastery includes communication, not just execution.
But I have always approached life with a different philosophy.
I try to replace myself.
That does not make me weaker. It forces me to become better. More precise. More valuable. If I cannot explain what I do, I do not truly understand it. If my knowledge cannot be transferred, it is brittle. And brittle skills break.
This is why the loudest claims about what AI will never achieve tend to age poorly.
It is not that the technology magically improves overnight. It is that the people willing to engage with it improve themselves in the process. They learn how to think more clearly. How to structure intent. How to decompose problems. How to communicate with rigor.
Those who refuse to teach their craft do not stand still.
They fall behind.
AI is not replacing developers, writers, or artists in the way people fear. It is exposing who was already operating on autopilot. It is revealing who understands their craft deeply enough to teach it and who relied on mystique instead of mastery.
The art is not in the output.
The art is in the communication.
And those who learn that lesson early will not just survive this shift. They will define what comes next.