Post

In the face of generative AI coding agents, the role of human-led architecture becomes even more important in the crafting of software-intensive systems.

10 reposts45 likes

Could agents estimate architectural significance from dependency impact and migration cost before proposing code, rather than after generating it?

All architecture is design, but not all design is architecture. Architecture represents the set of significant design decisions that shape the form and function of a system, where significant as measured by cost of change.

…but they are not particularly good at making decisions that balance the static and dynamic forces on a system in order to yield a reasonably optimal solution: this is where human judgment, human understanding, human knowledge, and human experience shine.

The trap is assuming cheap generation means cheap modification. An agent will happily couple two internal schemas with 200 lines of glue just to turn a local test green. Without hard interface boundaries, systems calcify around whatever glue the model threw together first.

Interface boundaries need a migration rule, not only a wall. Otherwise the agent routes around them through copied schemas and hidden adapters. I’d test cost-of-change by deleting one side’s type definitions: what breaks outside the declared contract?

Exactly. As code generation gets cheaper, the scarce part becomes preserving architectural intent through the change path. An agent can write the patch; the system still needs an explicit baseline, bounded scope and independent verification.

The architecture work also has to be written down where the agent reads it. Boundaries that only live in someone's head get quietly violated by the third generated PR.