Looking at the painting of the Ideal City of Urbino, though often linked to L. B. Alberti, one perceives an aseptic sensation, a total “anosmia” caused by the absence not only of human presence but of any organic element. In its apparent realizability, it does not seem to be meant for humans, it smells of nothing. Rather, it represents a fusion between political and divine power, a submission to a higher control, to a rigor governed by the exact sciences of geometry. It implies a political intent tied to its time. “Alberti ... recognizes that the structure of the city corresponds to the social and political structure ... The model considered the best is the oligarchic one: the ‘most illustrious and wise’ are entrusted with supreme direction, which they exercise through a solid executive apparatus, and the masses must obey and collaborate.” What would be the result of submitting the image of Urbino’s Ideal City to an automatic “enhance” process using artificial intelligence? Is AI capable of transposing ideality? The algorithm would need to be trained to interpret contemporaneity through a political and socio-economic lens at the very moment the prompt is entered. It would also have to integrate abstract values such as ethics or morality, which are evidently not universal but instead are closely linked to the author's own perspective. AI is not capable of grasping the subtleties embedded in architectural drawings rich in social content, it is inevitable that the result will be merely object-based. The outcome may be formally captivating and technically feasible, but inevitably out of time and impersonal, no longer able to serve as a "medium" without the critical voice we find, for example, in the controversial originality of Exodus, or the voluntary prisoners of architecture.
Zambon et al. (Thu,) studied this question.