This paper examines the epistemological transformation of the creative and design process across the transition from analogue to digital practices and, ultimately, to generative artificial intelligence, focusing on the relationship between the materiality of knowledge and the immateriality of algorithmic processes. Where traditional practice grounded the production of meaning in a progressive sequence of actions — a direct, embodied link between gesture, technique, and support (Sennett, 2008; Adamson, 2007) — AI introduces a structurally different logic, which this research defines as the inverse process: the designer no longer constructs form from intention, but receives an already-formed output and must retrace its logical, aesthetic, and ethical coherence backwards. This inversion transforms the very nature of design competence, shifting it from constructive to curatorial. On the aesthetic level, generative models tend to produce an algorithmic pastiche — a recombination of pre-existing elements structurally analogous to the postmodern pastiche theorised by Jameson (1991), yet statistically determined by the mass of training data rather than by a conscious authorial choice (Manovich, 2019; Somaini, 2023). The result is a convergence towards uniform, preconfigured aesthetics that the research interrogates critically: do they reflect an authentic desire of the designer, or are they a condition accepted because it is favoured by algorithmic tools? In response to this twofold transformation — of process and of aesthetics — the paper proposes a model of critical design education articulated around three components: Limited Knowledge Models (LKM), trained on situated and contextual knowledge (Haraway, 1988) and capable of preserving local knowhow as an alternative to large universal models; constraint-based pedagogy, understood as the recovery of material limitation as a cognitive device that reactivates the embodied dimension of making; and the figure of the designer-curator, a profile that privileges ars interrogandi over algorithmic delegation, maintaining ethical responsibility throughout the process. Through a literature review and the mapping of case studies, the research documents how these components may integrate into a hybrid pedagogical model capable of forming designers who are critically aware of the algorithmic logics that traverse their work, without surrendering the critical depth that design culture has historically produced.
Il processo inverso: Conoscenza incarnata, estetica algoritmica e formazione critica nel design dell'era dell'IA
Jacopo Battisti
;Filippo Maria Disperati;
2026
Abstract
This paper examines the epistemological transformation of the creative and design process across the transition from analogue to digital practices and, ultimately, to generative artificial intelligence, focusing on the relationship between the materiality of knowledge and the immateriality of algorithmic processes. Where traditional practice grounded the production of meaning in a progressive sequence of actions — a direct, embodied link between gesture, technique, and support (Sennett, 2008; Adamson, 2007) — AI introduces a structurally different logic, which this research defines as the inverse process: the designer no longer constructs form from intention, but receives an already-formed output and must retrace its logical, aesthetic, and ethical coherence backwards. This inversion transforms the very nature of design competence, shifting it from constructive to curatorial. On the aesthetic level, generative models tend to produce an algorithmic pastiche — a recombination of pre-existing elements structurally analogous to the postmodern pastiche theorised by Jameson (1991), yet statistically determined by the mass of training data rather than by a conscious authorial choice (Manovich, 2019; Somaini, 2023). The result is a convergence towards uniform, preconfigured aesthetics that the research interrogates critically: do they reflect an authentic desire of the designer, or are they a condition accepted because it is favoured by algorithmic tools? In response to this twofold transformation — of process and of aesthetics — the paper proposes a model of critical design education articulated around three components: Limited Knowledge Models (LKM), trained on situated and contextual knowledge (Haraway, 1988) and capable of preserving local knowhow as an alternative to large universal models; constraint-based pedagogy, understood as the recovery of material limitation as a cognitive device that reactivates the embodied dimension of making; and the figure of the designer-curator, a profile that privileges ars interrogandi over algorithmic delegation, maintaining ethical responsibility throughout the process. Through a literature review and the mapping of case studies, the research documents how these components may integrate into a hybrid pedagogical model capable of forming designers who are critically aware of the algorithmic logics that traverse their work, without surrendering the critical depth that design culture has historically produced.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


