Today’s conversation about AI in software development is largely centered around productivity: generating code faster, accelerating reviews, and helping developers work more efficiently. While these advances are significant, they represent only the first step of a much larger transformation. They still assume a fundamentally human-centered software engineering process in which AI acts merely as a tool.

This talk explores a different future: one in which AI agents become active participants in software engineering teams. Inspired by the emerging field of Agentic Software Engineering, we examine how software engineering changes when humans and AI agents collaborate as peers within a shared socio-technical system.

In this new paradigm, AI agents do more than generate code. They analyze requirements, investigate incidents, propose architectural solutions, coordinate development activities, and interact with human experts whenever judgment, creativity, or accountability is required. As a result, the role of software engineers evolves from executing tasks to designing, guiding, and governing networks of collaborating agents.

The implications reach far beyond developer productivity. Team structures, engineering processes, software architecture practices, governance models, and even the definition of a software engineering organization will need to be reimagined.

This session offers a forward-looking perspective on the future of our profession and discusses what software architects can do today to prepare for a world where software is no longer built solely by humans, but by hybrid teams of humans and AI agents working together.