The online and offline world has been pigeonholed into a constant state of optimization: for attention, resources, and “engagement”. How do these metrics impact the content we create? What should, or shouldn’t we, use them to evaluate?

Attention Economy is a speculative series of continuously-evolving A.I. generated paintings that constantly self-optimize to attract the attention of passerby and viewers. Each installation starts with a random painting, created using a generative deep learning model trained on decades of historic artworks. Using face detection and head-pose analysis to create its own set of “engagement” metrics for each viewer, each piece constantly learns and develops a distinct painting, tailored to the viewers around it.

Note: no data about viewers is saved, collected, or shared in any way. The piece immediately computes anonymous metrics not attached to any identifier, and discards all individual data.