Design Remains a Cultural Practice

Why Perception, Judgment, and Responsibility Must Remain Analog in the Age of AI

Abstract

The integration of generative artificial intelligence into design practice and education has repositioned analog processes — sketching, modeling, process documentation — as optional or preparatory stages, effectively marginalizing their epistemic function within the design process. This paper argues against that repositioning. Drawing on established research in design cognition, theories of trust and accountability in knowledge production, and detailed analysis of two case studies from editorial information design — the Artemis-2 mission graphics developed by Marco Hernandez and his team at The New York Times, and the scientific reconstruction work of Fernando Baptista at National Geographic — we contend that analog practices constitute irreducible conditions for traceable, responsible, and culturally grounded authorship in design. The analysis reveals that authorship in high-stakes information design is produced through sequences of epistemic decisions — selection, verification, translation, deliberate rejection — that AI-generated outputs structurally obscure rather than document. A second risk, specific to design education, is identified: the substitution of process authorship with output evaluation repositions both students and instructors as consumers of computational production rather than as designers of knowledge. Against this, the paper proposes a framework of process-oriented critique and analog-anchored pedagogy in which AI tools are situated as instruments within — not substitutes for — the human design process. The implications concern not only pedagogical methodology but the conditions of trust and attribution that sustain design as a cultural practice.