AI can help ask the right questions early and surface industry-specific basics even before the customer explicitly mentions them. I enjoy learning how different industries work, and AI is surprisingly good at covering the fundamentals. Country-specific details still matter, but the core concepts are often the same. From there, it is mostly time and effort to reach the level of domain expertise customers expect.
Another challenge is the size of the technology landscape. No architect can know everything, and projects often include technologies that are new to them. Previously this meant a lot of studying. Now AI can act as a tutor, a sounding board, and sometimes a second opinion. You can ask endless questions, check best practices, and validate ideas. It feels like being part of an architecture team.
This does not mean the architect role gets easier by default. Everyone has access to these tools, so learning to use them well matters. Context management becomes a key skill, since architects work with details that AI models do not know by default, such as internal data, requirements, and new technologies.
AI is also increasingly part of the software being built. In that sense, it feels similar to cloud infrastructure. There are dedicated cloud architects, but almost every architect today needs at least a solid understanding of cloud basics because almost every system runs there. AI is heading in the same direction. Every architect will need to understand the fundamentals, while some will naturally choose to specialize further and focus on AI architecture itself.
The role of the architect is not shrinking. It is expanding. And it now includes knowing how to utilize AI both for the architecture work itself and the end result.
