Electronic Chaos Oracle #4
This is another edition from the installation series, the Electronic Chaos Oracle. The Electronic Chaos Oracle is an ongoing artistic research project by Marcel Schwittlick and Ramin Soleymani. With this work, the artists are investigating possibilities of creating an artificial intelligence systems to inspire it’s observers. The Electronic Chaos Oracle opens up new questions by continuously feeding back the current topic of the conversation into the next question or statement it’s proposing. The project plays with the notion of automatically generating novel ideas by simulating creativity. The edition #4 is an augmented machine learning conversation, generated from millions of statements taken from artbooks. See this list for a full list of the credits: mrzl.net/eco_pdf/pdf_credits.pdf.
One main idea behind the project is to close the circle of training machine learning systems. The situation of feeding vast amounts of data in our new machine learning and artificial intelligence systems was lacking the purpose. The Electronic Chaos Oracle’s objective is to give humans the chance to learn from their own data, which was used to train the intelligence behind the installation. The Electronic Chaos Oracle (ECO) is an interactive installation and an embodiment of an ancient oracle in the computer age. With ECO you can collaboratively co-author texts, ideas and theories around any topic you like. ECO distinguishes itself from it’s chatbot predecessors by having the premise of being a partner in the process of developing creative and expressive writings. What you can co-develop with ECO ranges from philosophical, theoretical-aesthetic, media-theory, culture critical, feminism and visual art, to conceptual art, digital culture, internet theory, ecology, political theories, cybernetics, virtuality and the algorithmic revolution amongst much more. ECO’s knowledge about these topics are based on a curated database of scientific research publications. One elevant aspect of the nature of ECO is that is can act as a feedback system for learning purposes. One tendency of recent machine learning developments are the continuous feeding of algorithms with data, the machine learns. ECO is attempting to close this cycle by constituting a framework that helps us, humans, to learn from machines.
Read the full story here: mrzl.net/eco_pdf/eco4_story.pdf
Taking into account that the population on our world is growing exponentially. There are not enough teachers for things to learn. There is too much information and knowledge to be taught, that we are having problems to continue learning from our past. ECO is a necessary reaction to this phenomena. A machine to learn from ourselves in an exploratory fashion.
Marcel Schwittlick & Ramin Soleymani