XEPA anticipates a future where machines form their own societies. Going beyond mere generative art, machines will exhibit artistic creativity with the addition of artistic judgment via computational aesthetic evaluation. In such a future our notions of aesthetics will undergo a radical translation. The XEPA intelligent sculptures create animated light and sound sequences. Each sculpture “watches” the others and modifies its own aesthetic behavior to create a collaborative, improvisational performance. No coordination information or commands are used. Each XEPA independently evaluates the aesthetics of the other sculptures, infers a theme or mood being attempted, and then modifies its own aesthetics to better reinforce that theme. Each performance is unique and widely varied. XEPA is an ever-evolving artwork, intended as a platform for ongoing experiments in computational aesthetic evaluation.