Can an artificial intelligence recognize itself?
Artificial Intelligence (AI) and Machine Learning (ML) are developing at an astonishing pace. Current technology enables algorithms to match individual areas of human cognition, such as problem solving and reasoning. Current technology enables algorithms to reach individual areas of human cognition, yet developing an Artificial General Intelligence (AGI) that completes a full range of cognitive abilities and is both broad and adaptable is still uncertain. A key feature of human consciousness is self-awareness: the awareness of our existence, our body, our mind. What do we always look at, but never fully understand? This is the question at the core of the human experience.
Narciss, by Christian Mio Loclair, is an AI seemingly reflecting on its own existence. The artwork is built as a machine that analyzes its own physical embodiment with the sole purpose of describing its thoughts while looking at itself. Composed of a bare set of electronic components, namely a GPU, CPU, hard drive, cooler and camera, the sculpture represents the minimum configuration required to execute algorithms of self-exploration. While the machine continuously analyses camera images, it uses AI to translate them into sentences and turns self-portraits of a digital body into lyrical guesses. Narciss portrays these endless attempts of self-interpretations on a screen mounted to the back of its body – inviting us to an endless performance of autobiographical narration and a research experiment on models of synthetic self-awareness.