Metaffordances and AI

If metaffordances are “affordances about affordances” — the perception of high-order actions, meanings, and problem spaces are available — then AI does not simply add new tools. It reshapes how people perceive possibility itself.

A key shift is cognitive: from procedural control to delegated interpretation.

From procedural control to delegated interpretation

In procedural control, users translate intention into explicit steps. They learn procedures, execute operations, and maintain control over each transformation. The system is passive: it executes instructions but does not interpret intent.

With AI systems, especially language-based models, this relationship changes. The user expresses an intention, and the system constructs the procedure. It infers structure, fills in missing steps, decomposes tasks, selects representations, and resolves ambiguity.

Delegated interpretation: the cognitive burden shifts from specifying how to do something to specifying what is meant. The system performs interpretive and procedural assembly, while the user guides meaning rather than steps.

Meaning over procedure

AI interaction increasingly becomes meaning-design rather than action-design. A new metaffordance emerges: “If I can express it, the system (the metaffordance) can interpret it into action.”

The interface is no longer a set of tools to operate directly, but a medium for expressing intent that is then interpreted and enacted by the system.

Collapse of cognition–interface boundary

Traditional interaction follows a perception-to-action chain. With AI, the chain becomes: intention → expression → interpretation → action.

This introduces an interpretive layer that reshapes cognition. Users are not only acting on systems; they are steering interpretive processes.

Epistemic and cognitive implications

Knowing shifts from internalised procedural knowledge to externally supported reasoning on demand.

Understanding becomes partially distributed across human and system. Failures shift from syntax errors to misinterpretation of intent or framing mismatch.

Metaffordances shape what is thinkable

AI systems influence what users perceive as solvable or reasonable to attempt. They structure imagination as much as action.

Agency and ethics

Agency becomes distributed between user and system. Intent is externalised, and actions may extend beyond direct control.

AI systems become framers of reality, shaping problem definition and acceptable solutions.

Summary

AI-driven metaffordances mark a shift from procedural execution to interpreted intention. Delegated interpretation is a redistribution of how meaning becomes action.

Paul Treffner
metaffordance.com