In the performance “Learning”, the action potentials of real and artificial intelligence are compared by having an actor and a 3D character (spatially visualized through Pepper’s Ghost effect) act in equivalent spaces in parallel.
Both are initially unable to stand and try to learn this during the performance. The actors repeatedly fall to the ground and try to learn from their mistakes.
The digital character is equipped with a genetic algorithm, a programmed intelligence, while the real character has an organic brain.
The avatar learns to stand faster than the human but simultaneously appears ridiculous in its uncoordinated actions, revealing the distance of digital technology to the real human.
In contrast, the human evokes emotional empathy in the viewer through their physical pain and desperate attempts to stand.

Andreas Bayer

Video of the working process, how the AI learned to stand: