"Is coding dead?"
A Computer Science graduation student I coach asked me this today.
I am also the stakeholder for her graduation project, so we regularly talk about what she is building, how she approaches it, and what becoming a software engineer might look like in the coming years.
Behind her question was obviously another one:
Am I preparing myself for a profession that may disappear?
I gave her the most honest answer I could.
I think coding as pure execution is slowly dying.
If your main value as a developer is that someone gives you a technical design and you turn it into code, AI will become very difficult to compete with. Claude, Codex and other coding agents are already becoming remarkably good at exactly that part.
But I don't think the software engineer is disappearing.
I think the job is moving up a level.
I gave her the example of an orchestra conductor.
A conductor needs to understand the instruments. He needs to know what strings can do, what percussion can do, how everything fits together and when something sounds wrong.
But he doesn't need to be the best player of every instrument.
His strength is in orchestrating them.
I think the software engineer of the near future will increasingly work in a similar way.
You may enter a company with your own arsenal of agents, tools and workflows. One explores the codebase, another builds, another tests, another checks security. You give direction, connect them, challenge assumptions and audit what is actually happening.
That doesn't mean you no longer need to understand code. I told her the opposite.
You need to understand it well enough to know when the machine is wrong.
And that requires architecture, technical design, security, judgment and a much deeper understanding of the problem you are solving.
So if I were starting Computer Science today, I would absolutely still learn how to code.
I just wouldn't prepare for a career where coding is the final destination.
I would learn how to conduct the orchestra.