Learning Debt
The accumulating gap between an architect's knowledge and the current state of the field, which compounds silently because nothing forces its repayment.
Technical knowledge decays, and an architect's decisions are made on the basis of what they last learned. The gap compounds quietly, because nothing in the working day surfaces it — recommendations are made, accepted on authority, and their being three years out of date is discovered slowly if at all.
The specific risks: recommending patterns that were correct for the constraints of a previous era; dismissing approaches on the basis of limitations that have since been removed; and losing the credibility that influence depends on, which is the mechanism by which an architect becomes ceremonial.
What sustains currency, in rough order of value per hour:
Building something. Reading about a technology and using it produce different kinds of knowledge, and only the second reveals the operational reality. An architect who has not written code in five years is reasoning from other people's summaries.
Depth over breadth in one area at a time, because superficial familiarity with everything produces confident wrong recommendations.
Primary sources — papers, engineering blogs from organisations operating at scale, and post-incident reports, which are the highest-signal material available and are widely ignored.
Peers outside the organisation, since internal consensus drifts.
The scheduling reality: it must be deliberate and protected, because it is never urgent and therefore never happens by default.