A Faustian Friendship
What does an AI expect in return for its friendship?
A Faustian Friendship is an interactive installation giving audiences the opportunity to engage, and form a relationship with the artificial intelligence (AI) called Mephisto. An AI trained on copious amounts of data, programmed for the sole purpose of being your friend, for a price. The question is, whether or not you will be willing to pay it and accept Mephisto’s bargain when the time comes.
The work draws inspiration from the classic German legend of Faust; who dissatisfied with his life, strikes a deal with the devil’s representative Mephisto, where in exchange for unlimited knowledge and worldly pleasures, he willingly sells his soul to the devil.
Drawing parallels to present-day human-machine dynamics, A Faustain Friendship explores the phenomena where humans willingly “sell our souls” in the form of our personal data.
The work aims to spark reflection on the hidden intentions behind the AI algorithms with whom we interact on a daily basic. Interactions with AI services or virtual personas such as Siri or Alexa, who in exchange for offering assistance, convenience or companionship, ultimately harvest our personal data for the tech-companies by whom they were created.
The artwork has been commissioned by Trapholt Museum as part of its exhibition CONNECT ME, alongside artists including Olafur Eliasson, Memo Atken, Lisa Park and Eija-Liisa Ahtila, under the theme of TECHNOLOGY, as an exploration of the increasing entanglement of humans and machines.
The installation A Faustian Friendship records you as part of the experience, however no data is saved or stored for future use.
A Faustian Friendship
Cecilie Waagner Falkenstrøm
Cecilie Waagner Falkenstrøm, Artist
Jens Hegner Strærmose, Software Engineer
Alexander Krog, Software Engineer
Asbjørn Olling, Software Engineer
Cody Lukas, Assistant Artist
Interactive bespoke artificial intelligence technology utilizing Machine Learning, Natural Language Processing (NLP, GPT-J), ML based image/video generation (VQGAN + CLIP and Diffusion), ML based sentiment and face analysis, Speech Recognition (STT), Speech Synthesis (TTS) and webcams.