In the latest edition, we speak with Liang Zhaosi, Software Engineer at ALTEN China, about what he calls the “10-Year Question”:

As AI-assisted coding has rapidly evolved from a niche tool for tech enthusiasts into a standard part of software development, tasks such as code completion, basic code generation, and automated test creation are becoming increasingly efficient.

But as machines take over more of the coding itself, a more fundamental question emerges:

What will define the long-term value of engineers?

For Liang, this is not simply about upgrading our tools. It represents a fundamental shift in the way engineering work is approached.

From Writing Code to Defining the Paradigm

Over the past decade, software engineering has largely been built on the assumption that people write the code. The next decade of AI-native development will increasingly be built around human-machine collaboration.

The shift from writing to designing, and from implementing to defining, is more than a change in workflow. It is a reset of the way engineers think about their work.

Engineers today are uniquely positioned at the intersection of two eras. They understand traditional engineering practices while being among the earliest adopters of AI tools. They value human judgment, while also knowing when repetitive work can be delegated to machines.

This makes them natural bridges between traditional engineering and the intelligent era.

That bridge carries two important responsibilities:

Passing on engineering knowledge

Turning accumulated industry rules, patterns, and best practices into knowledge that AI systems can understand and apply.

Redefining how value is evaluated

Establishing new standards for balancing efficiency with safety, speed with responsibility, and automation with accountability.

Over the next decade, completing this transition may become one of the most significant contributions of this generation of engineers.

The First Source of Irreplaceability: Defining the Problem

The more capable AI becomes at generating code, the more valuable the ability to define the right problem becomes.

Ambiguous requirements, conflicting business expectations, and hidden non-functional constraints are the reality of engineering in the real world.

The value of an experienced engineer lies in turning this complexity into a clear, logically structured framework that AI can execute.

This goes beyond prompt engineering. It requires a deep understanding of the problem domain: identifying boundaries, anticipating exceptions, and abstracting reusable capabilities.

AI can generate highly effective solutions within a given framework. But who defines the framework? And how do we balance what is “right” with what is “fast”?

These decisions still belong to engineers.

The Next Level: Making System Decisions Under Multiple Constraints

True engineering judgment comes from continuously balancing multiple variables, including performance, cost, maintainability, security, and team velocity.

An architectural decision can shape a technology roadmap for the next three years. A data-storage strategy can determine compliance requirements and business risks.

These decisions cannot simply be automated, because they involve prioritizing values—and that is ultimately a business question, not just a technical one.

Engineers remain accountable for the outcome, including decisions related to safety, ethics, data privacy, and business continuity.

That responsibility is part of what defines the profession—and one of the boundaries AI cannot easily cross.

The Next Shift: Redefining “Mobility” in the Automotive Industry

Working in the automotive industry, Liang sees the next major transformation going beyond range or acceleration.

It is about redefining mobility itself.

Vehicles are evolving from means of transportation into intelligent, autonomous spaces—and potentially into new kinds of companions for people.

This transformation requires engineers to rethink their role: from technical implementers to directors of human-AI collaboration.

The biggest challenge in the coming years may not be learning another new framework. It may be learning to take a “director’s view”—orchestrating AI capabilities, designing interactions, and finding the right balance between machine-driven decisions and human control.

Break the Stereotype: Engineers Are Collaborators First

In real-world engineering teams, engineers are far from the stereotypical image of the isolated “tech geek.” Instead, they are highly collaborative builders: listening closely to the needs product managers may not explicitly articulate, clearly explaining complex logic to testing teams, and working with operations teams to design observability systems. Liang particularly highlights deep listening and the ability to ask questions as some of the most valuable non-technical skills. These skills are not only important at work, but also fundamental to meaningful connections between people.

Being able to ask the right question is often more difficult—and more valuable—than providing the right answer.

AI will not replace engineers. But engineers who know how to use AI will replace those who do not.

What will truly set outstanding engineers apart are still the human qualities that cannot be coded: the insight to define problems, the judgment to make trade-offs, and the ability to communicate and collaborate with empathy.

These are the enduring questions that remain unchanged amid technological transformation.