With AI, Everybody Is an Engineer

There is a hiring conversation that captures where this is all going. A candidate for a platform team lead was asked about the on-prem Kubernetes SREs who, famously, do not want to write code. They want to run systems, not be engineers. His answer was one sentence: but with AI, everybody is an engineer. He got the offer mid-interview.

That line is worth sitting with, because it reframes what AI actually does to an organization. Most of the conversation about AI coding treats it as a productivity multiplier for people who already write code. Faster engineers, more output per engineer, the same team shipping more. That is real, but it is the small version of the story. The bigger version is that AI does not just speed up engineers. It changes who counts as one.

The SRE who never wanted to write code was making a rational choice about friction. Writing code is a skill, it takes years, and if your job is keeping systems alive you had every reason to stay on your side of the line. AI collapses that friction. When you can describe what you want and get working software back, the reason to refuse evaporates. The reluctance was never about the goal. It was about the cost of getting there, and the cost just dropped to near zero.

This is why I keep saying that agents are about to break out of engineering. The primitive that used to belong to a walled-off team now shows up everywhere. Sales runs agents. Ops runs agents. The person who would have filed a ticket and waited two sprints now ships the change themselves. The E in SRE always stood for engineering. AI just makes that impossible to ignore.

So the practical question for leadership is not whether to give your engineers AI tools. It is whether your org chart still makes sense when everyone can produce software. If engineering is a small department that everyone else routes requests to, you are structured for a world that is ending. Staff for a world where every function touches code, where review and approval matter more than authorship, and where the bottleneck is taste and judgment rather than who knows the syntax. The candidate saw it in one sentence during an interview. The question is whether your organization sees it before the reorg gets forced on you.