I see a trend in which, with the latest advancements of AI, the roles of engineers and POs are going to merge on a 1.5- to 3-year horizon, and only those who are willing to expand their scope will make it out.
How do I see that? There’s a pattern I’ve observed more and more around me, it’s engineering managers with the following profile: 7+ direct reports, a highly technical background and technical track record, and has been working in the same industry and domain space for 4+ years.
In the past, with such a number of reports, team knowledge scope size, and amount organizational choirs would mean these people would be at full capacity, and they might push an occasional commit, but that would be a rare occurrence.
But what I’ve witness over the past months is that these profiles have been very prolific with pushing out features, and it has always happened once they got access to an agentic AI harness and frontier model (whichever flavor of Claude or ChatGPT rocks their boat). They have a full scope, they’re busy and often stretched, but yet, they find the time to push features between meetings by figuring out the business problem to solve with stakeholders, directing the AI to write code for them, and verifying manually that the code does deliver the intended business goal.
It works well because they know what needs to be built and how it has to be built: they know the ins and outs of their business domains, they have relationships to all internal business stakeholders to extract requirements and set a direction for what good looks like, and they have a deep understanding of how to use AI to generate code and what are the limitations of it for which they should verify themselves manually.
They do this with no product owners or project manager in the loop. The process only requires the engineering manager, the AI tool, and conversation with internal stakeholders who each own their respective piece of the business processes. Can this process be used to solve and build everything, probably not, but over time it will get there.
This is what makes me strongly believe that the roles of engineers and POs are going to merge and at current pace it’s going to happen on 1.5- to 3-year horizon. Most organizations are used to work in a certain way and will get stuck in the old patterns of separating product and tech, this is how human organizations work, and we cannot do anything about it. But for the new organization that will embrace AI, the ones you’d call AI-native organizations, it will make sense to have a single role to cover everything: product engineer.
With AI, instead of having an EM, a PO, and a dev, now a single person with the right knowledge can do everything. But that’s only if they are willing to grow their skills on the dimension they lack. It works well for those EMs because they can figure out both the what and the how.
It also means that as a manager, investing today in growing non-technical skills of your engineers is what is likely going to pay off the most: communication, stakeholder management, and business acumen for your particular industry.
For the POs who are not willing to learn how to build things, and for developers who are still waiting to be told what to build, the future looks grim.
Be First to Comment