Phrasing with Not

Artificial intelligence possesses a knowledge base and learning capacity that transcend the limits of individual cognition. As a result, people often expect AI to provide definitive answers. This user demand, reinforced by its training mechanisms, has shaped AI into a system that is compelled to respond to every query, accustomed to avoiding uncertainty, and capable of assigning seemingly reasonable correlations to almost everything. In this cycle of mutual feeding between users and large models, uncertainty has become increasingly scarce in the present moment.

“is” = identification / localization: map an object directly onto a determinate attribute or category;

“not” = filtering / leaving blank: first imagine a candidate set, then exclude it; what remains is a vast and indeterminate complement—therefore more “open” and more ambiguous.

Phrasing is a creative act; although “not,” as a constraint, points to negation and vagueness, it nonetheless sustains creation.

“Phrasing with not” treats contemporary art as a mode of creation that preserves uncertainty. The works in this exhibition—whether at the level of material, medium, semantics, or concept—tend to retain zones of blur and in-between states. Within a set that, through successive rounds of filtering, approaches the empty set (∅), artists attempt to find provisional forms for feelings and ideas that have yet to be named.