Deep Sophia is a three-channel interactive video installation that explores the intersection of artificial intelligence, deep fakes, and emotion capture. Using facial recognition technology, the Sophia Robot is mapped onto the faces of women from historical and contemporary films. Rather than a traditional close-up, Deep Sophia amplifies the way machine learning produces a profoundly uncanny form of human-robot expression. The actor is better at facial expression than the robot, while the robot effortlessly wraps itself around the actor like a garment. The human becomes the machinery for the robot’s artificial emotion, much as a snake takes on the shape of its prey as it swallows it, transforming itself as it digests it.