Personhood: Iteration 4

Personhood (Iteration 4) relies on state-based branching animation trees formed via AI animations that interpolate strategically between pivotal keyframes. At every given moment, the artwork “decides” which character animation to transition into next. Prior to recent breakthroughs in video generation, branching a complex scene into several alternate paths at this degree of seamlessness would have been close to impossible.

Previous iteration of the interaction used Unity to simulate characters casting shadows in the SF Asian Art Museum. Those procedural techniques yielded uniquely responsive character animations, but at best they moved like NPCs. This iteration foregoes that procedural quality to prioritize human fidelity, with a shadow formed in the likenesses of Xin.

*Curated by So-Yeon Yoon, supported by Cornell Council for the Arts.*