What Will Happen to the Labor Market in the Age of AI?

I still remember the mainframe developers. I was never one, but they were friends of my dad. They had an awesome job. Then came client–server, and their industry narrowed until many moved to banks and governments.

I still remember the DBAs who approved every schema change. Then came self-service databases, cloud-managed services, and DevOps pipelines, and database changes became part of the developer workflow.

I still remember the massive QA teams, testing each release by hand. I worked with hundreds of great people in Eastern Europe who did it. Then came automated testing and continuous integration, and the need for manual regression testing fell.

I still remember the cold and noisy machine rooms in every office and the IT folks installing servers, swapping cables, fixing printers. I hated that my seat was next to it. Then came the cloud, SaaS, and centralized IT.

Every wave of technology reshaped work. Some people moved early and thrived. Others were caught off guard and took a hit to pay or seniority. AI is another wave. The question is not whether it changes jobs, but how.

What history tells us

History is consistent: technology doesn’t wipe out all work, but it changes the work people do. In each wave, routine and rules-based tasks shrink while problem-solving, creative, and human-driven work grows. Those who adapted their skills quickly ended up with better jobs than the ones they lost. Those who didn’t… didn’t.

You can already see the pattern in past shifts. Clerical and repetitive production roles have been on the decline for decades. In contrast, service roles and high-skill analytical work have grown. The lesson isn’t just “learn new skills” — it’s learn the skills that fit where the work is moving.

What is different about this AI wave

Past automation mostly replaced physical tasks or narrow digital routines. Generative AI automates parts of thinking — drafting, summarizing, coding, analyzing. It reaches into jobs people thought were safe because they were “knowledge work.” The lift isn’t just at the top. In trials, AI tools have made the biggest measurable gains for less experienced workers, because they can borrow the patterns of the best in the field and apply them immediately.

That’s a bigger psychological shock than a new machine in the factory. It’s a tool that can sit in the same chair as you and do part of your job in seconds. Which means the real differentiator becomes what you do with the part it can’t.

What to do now

If you want to stay ahead, start where you are. Map the tasks you do most often. Which ones are repetitive and could be automated? Which ones require judgment, product sense, or deep domain knowledge?

Push yourself toward the second category. Take on projects that use AI, not just in the abstract, but to actually deliver something. Volunteer for the work that involves integrating tools, managing exceptions, designing workflows, or making final calls. Those are harder to automate and put you closer to decisions and outcomes.

And if you see your role leaning heavily on routine work, start shifting toward roles where AI is part of the job already. The more you work with it, the less likely you are to be replaced by it.

A small story

When I worked next to the machine room, it was freezing and loud. The IT team was always there — swapping drives, crawling under desks, keeping the lights on. When the cloud rolled in, the hardware left, and so did most of those jobs. Some of those people jumped early into cloud management and never looked back. Others stayed until there was nothing left to stay for. The work didn’t disappear overnight, but the opportunities did. If you want to look for a new job that fits the future more, I’m inviting you to join the alpha of my new job-landing chatbot. It’s a self-service tool designed for tech professionals to actually land jobs — not just search for them. It covers resumes, applications, interview prep, and market insights, all powered by GenAI.

You can get early access here: Join the Alpha