The Most Valuable Engineers Are Staying Closer to the Problem
AI Is Not Removing the Need for Engineers. It's Changing Where Their Judgement Creates the Most Value.
Implementation is becoming less scarce
For much of the industry's history, engineering skill was closely associated with implementation ability. As more of the implementation process becomes assisted, that balance is shifting.
The engineer's scope is expanding
Engineers are able to spend more cognitive effort on understanding problems, shaping outcomes, refining systems, and responding to how software behaves in real environments.
Faster implementation demands more clarity
Unclear intent can generate large volumes of low-quality change very quickly. The faster implementation became, the more valuable intentionality appeared to be.
What Changed in Practice
At its core, engineering has always been about problem solving. Implementation is part of that process, but it was never the entire job.
We started noticing this shift quite clearly inside our own teams.
One of the changes we made internally was ensuring that engineers had much clearer visibility into the external roadmap of the company and the outcomes we were trying to achieve for clients. Importantly, the roadmap itself was framed around customer problems and operational outcomes rather than simply lists of features.
The goal was to ensure engineers understood not only what they were building, but why it mattered and what problem the capability was intended to solve.
Engineers Embedded With Clients
We also started embedding engineers more directly with clients as capabilities were being adopted.
As implementation became faster and easier to iterate on, it became increasingly valuable for the engineers who built the capability to see how it behaved in real environments themselves.
This created much tighter feedback loops.
Engineers could observe where users struggled, where assumptions broke down, or where small workflow improvements would meaningfully improve adoption. Because implementation changes had become significantly cheaper and faster, subtle refinements that previously might never have justified prioritisation could now be implemented and deployed very quickly.
That changed the economics of refinement quite substantially.
Historically, improving a workflow after observing real user behaviour often required moving back through lengthy prioritisation and planning processes. As implementation accelerated, that trade-off became less severe. Small improvements became inexpensive enough that teams could respond much more quickly to what they observed in practice.
What We Increasingly Value
This changed what we increasingly valued in engineers themselves.
Implementation skill still matters significantly. Strong technical understanding, systems thinking, operational awareness, and architectural judgement remain essential.
What changed was where engineers could apply more of their attention. We found ourselves placing increasing value on engineers who could:
- Understand customer problems clearly — not just requirements
- Shape intent before implementation — defining what success looks like
- Identify operational consequences — thinking beyond the happy path
- Refine workflows based on feedback — the last mile of adoption
- Connect technical decisions to product outcomes — the bigger picture
- Iterate effectively in response to real usage — continuous learning
Why Discipline Matters More
This is one reason we increasingly believe that AI-assisted development does not reduce the need for engineering discipline. In several areas, it appears to increase it.
Faster implementation places more pressure on:
- Systems understanding — knowing how things fit together
- Operational ownership — caring about what happens in production
- Architectural clarity — making intent explicit
- Intentionality — building the right thing, not just building fast
- Product judgement — understanding what matters to users
- Shared understanding — keeping the team aligned
The most effective engineers we have seen are not simply the engineers producing the highest volume of implementation. They are increasingly the engineers who combine strong technical understanding with strong product awareness, operational context, and the ability to continuously refine systems based on real-world feedback.
