The Evolving Roles in IT Teams and My Personal Reflections in the Age of AI

The Evolving Roles in IT Teams and My Personal Reflections in the Age of AI

In software development and product development, roles like product manager, architect, and project manager used to be clearly separated. The product manager mainly deals with why something should be done and what should be done. That includes things like requirements, value, user experience, and product boundaries. The architect deals with how it can be implemented technically, with the focus on whether the system can actually work, whether it is stable, secure, and able to scale later on. The project manager deals with how to get things done on time, pushing progress, managing risk, and coordinating collaboration within limited time, budget, and manpower.

These roles are really dealing with three different layers of the same thing. One is direction, one is the path of implementation, and one is the process of delivery. On the surface, they just look like different divisions of labor, but in reality they are tightly connected. The product manager sets the direction, the architect works out the technical path, and the project manager organizes resources and rhythm so that the work can finally be delivered. Whether a team can truly function often depends on whether these parts can connect properly.

In the age of AI, these roles themselves have not disappeared, but their center of gravity is clearly shifting. A lot of work used to be execution-oriented. Now more and more of it is moving toward orchestration, guidance, and result validation. Product managers used to spend a lot of time on prototypes, pages, and feature lists. In the future, what matters more may be understanding what AI can actually do, what it cannot do, and where its boundaries are. Especially as many interactions gradually move away from traditional graphical interfaces toward natural language, the focus of product design will also change. The question is no longer just where to place a button or how a process should be clicked through, but how to provide AI with the right context, data, and constraints so that its output is closer to real needs.

The change on the architect’s side is also very obvious. In the past, the main work was traditional IT system architecture, thinking about services, databases, permissions, networks, and deployment. Of course those things still matter, but they are no longer enough. The new focus has become how to connect models, how to organize knowledge bases, how retrieval should work, how context should be passed, how tools should be called, and how the whole system can form a structure with cognitive capability. Put more directly, architects used to design information processing systems. Now they are increasingly designing something closer to a cognitive system. And this system is not abstract. It quickly runs into very concrete issues, such as whether the calling cost is too high, whether the response is fast enough, whether enterprise data may leak, and whether the whole solution is controllable. These are no longer peripheral issues. They have become central ones.

The change for project managers may be even more direct. In the past, a large part of their work was following processes, pushing schedules, attending meetings, tracking status, and coordinating task distribution. In the future, a lot of that standardized and process-based work will be taken over by AI tools. As a result, the real value of project managers may become more concentrated in the parts that are not so easy to replace, such as coordinating between people, handling conflicts of interest, judging priorities, balancing different teams, and making practical decisions in situations where information is vague or even contradictory. In other words, project management will not disappear, but it will increasingly shift from process management to interpersonal management and judgment management.

For someone who has worked in IT for several decades, seeing these changes naturally brings some feelings. Many things that used to be familiar, important, and repeatedly emphasized are no longer in the position they once were. Some abilities that used to be seen as core are moving into the background. Others are still there, but their meaning has changed. This is not simply a matter of one tool replacing one role. It is a broader reshaping of the whole way of working, collaborating, and defining capability.

Still, I do not resist this change. I am still willing to keep watching, and still willing to get my hands on some concrete AI projects myself. For me, this is not necessarily about going back into the old kind of professional competition. More than that, it is about keeping my mind active and not letting it become dull too quickly. Retirement does not mean completely cutting oneself off from technology. It only means that the way of participating has changed. Instead of being pushed forward by work as before, one can now engage with it in a lighter, freer way, one that is closer to personal interest, and continue to observe it, experience it, and understand it.

In the end, however dramatic the changes in the world of technology may be, they still belong to the world of technology. Human life itself has some more stable things in it. Sunshine does not change because models are upgraded. Beaches do not disappear because technical routes shift. Grass, tree shadows, and coffee are all still there. Technology is certainly worth paying attention to, but it is not the whole of life. At this stage, being able to understand these changes while not being pulled too tightly into them seems to me a fairly good state to be in.