We Are All Tech Leads Now

The era of the "coder" is over.

We are witnessing the "TechLead-ification" of the software engineer.

With AI agents handling the implementation details, we are all shifting toward work usually reserved for Staff Engineers and Tech Leads: system design, architecture planning, code reviews, and aligning technical roadmaps.

Initially, I was skeptical. I considered AI to bo just a fun toy. I made silly mistakes. It hallucinated a lot. I told myself what we all told ourselves: "Clearly, this will not replace me."

But the technology evolved faster than my scepticism could hold out. The models improved, the tooling matured, and I finally had to admit: This stuff is good.

I rarely write code the "normal" way anymore. Line by line, function by function. That approach feels increasingly archaic.

Now, I define scope, expected outcomes, and test cases, and then I release the agent to do its thing. It feels less like programming and more like managing a Junior Engineer, one who never sleeps and never complains. One who always tells me I'm absolutely right.

The Great Squeeze

I see my colleagues experiencing a similar shift. The definition of "coding" has fundamentally changed. With our herd of AI agents, we are not just more efficient in writing code. The whole idea of what "coding" means for us is different.

We are entering an era where the "Junior Engineer" is software, not a person.

The Missing Junior

Historically, software engineering advances by abstraction. High-level languages abstracted away assembly; frameworks abstracted away boilerplate. Now, AI agents are abstracting away the implementation itself.

Across the industry, coding agents like Cursor and Claude are reshaping engineering roles by effectively absorbing the tasks traditionally assigned to juniors. With AI handling implementation details, humans are being pushed to become full-stack generalists.

When daily coding is being handled by automatons, the centre of gravity shifts from writing to reviewing. The challenge is no longer syntax; it is ensuring that the AI's output aligns with high-level architecture, security standards, and the product vision.

Successfully managing this "AI workforce" requires experience and intuition that junior engineers simply haven't built yet.

The result is the devastation of entry-level roles.

Big Tech companies reduced hiring of new graduates by 25% in 2024, building on a 50% drop since 2019. The math is brutal: An AI tool costing $20/month that boosts productivity by 30% is far more attractive than hiring a junior developer who requires a salary, benefits, and months of mentorship.

We are moving toward a "hollowed org chart": Senior architects at the top, AI agents handling the grunt work at the bottom, and a missing middle where juniors used to learn the ropes.

With shrinking team sizes and the disappearance of junior roles, the complex scaffolding built to manage large engineering organisations loses its purpose. The meetings, the coordinators, the layers of alignment. They were there to manage human communication friction.

With that friction gone, they now look like an expensive bureaucracy.