AI is making broad engineers more useful. Hiring still rewards specialists.

I've spoken to several engineers recently whose work has become much broader.

Some are using AI tools to move between product, infrastructure and application code. Others are working across hardware and software, or running several coding agents while they concentrate on architecture, requirements and review.

They don't necessarily describe themselves as generalists. It is more that the amount of ground they can cover has increased.

One engineer I spoke to maintains two versions of his CV: one for general engineering roles and another for forward-deployed engineering. The underlying experience is the same, but he has to package it differently depending on the job.

Another described what he called a “specialisation gap”. He could work across a broad range of problems, but was still losing roles to people who had already operated at the required level in that specific area.

I've now seen the same tension from the company side.

One client recently interviewed an engineer with a handful of years of overall experience, only a fraction of it directly related to the particular area they were hiring for.

The client decided that he was still too junior for what they needed. His broader experience didn't compensate for the depth they wanted in the fundamentals of that domain.

That wasn't an unreasonable decision. They were making an early hire who would help set the technical direction of the company. The cost of getting it wrong was high, so they wanted somebody who had already solved closely related problems.

But it does show how the two things can be true at once.

AI tools may allow an engineer to contribute across a much wider technical surface area. Hiring processes can still favour the person who has already spent several years doing the closest possible version of the job.

Why hiring leans specialist

I can see why.

Breadth is difficult to evaluate. It can represent someone who understands how the parts fit together and can take ownership of a problem from beginning to end. It can also represent shallow experience across a collection of tools.

Exact prior experience is easier to assess. If somebody has already built the type of system you need, at roughly the scale you need, that feels like a safer hiring decision.

Job descriptions reflect this. They are usually built around the immediate problem a company wants solved: inference, observability, Kubernetes, compilers or a particular part of the stack. The broader engineer may be capable of solving it, but the specialist is easier to recognise.

Not every company hires this way.

I've worked with one company that has been willing to consider strong candidates across several different roles. Rather than asking whether someone matches one job description exactly, they look at the person's underlying capability and work out where it might fit.

I've also spoken to an engineering leader who deliberately avoids screening too heavily by title or exact skill set. He is more interested in intelligence, adaptability and whether somebody can create useful business outcomes.

Those examples are why I wouldn't say that AI simply rewards generalists, or that specialisation is becoming less valuable. I don't think the evidence supports either claim.

What I think may be happening is narrower.

AI is making some broad engineers more useful, faster than companies are becoming comfortable hiring them for that breadth.

The day-to-day work can change quickly. An engineer can adopt new tools, work across more of the lifecycle and become capable of delivering things that would previously have needed several people.

The hiring process moves more slowly. Companies still need legible evidence, particularly for senior or high-risk hires. The easiest evidence remains that somebody has already done almost exactly what the company needs.

That leaves some engineers in an odd position. Their practical capability is getting broader, but they still have to present themselves as a specialist to get through the door.

I don't yet know whether that gap will close. Companies may get better at assessing broader capability. New types of technical interview may emerge. Or exact experience may become even more valuable as AI makes surface-level competence easier to produce.

For now, I'm seeing engineers become broader while hiring briefs remain narrow. That feels less like a contradiction than two parts of the market changing at different speeds.