We are proud to host the 5th International Conference on Live Interfaces at NTNU in Trondheim. The local organizing committee is constituted by artists, researchers and students from music technology, music performance, fine arts, acoustics, art and media studies, and computer science, all at the Norwegian University of Science and Technology. Further, the support from NTNU Art and Technology (ARTEC) lend us a strong and multifaceted platform. Last, but not least, the collaboration with Meta.Morf and the theme “Digital Wild” lend us the opportunity to speculate on (among many things) the continuum from the organized and controlled to the wild and free.
The theme of Live Interfaces this year is “Artificial Intelligence – Artistic Intelligence – Automated Emotional Intelligence”. With the notion of artificial intelligence(s) becoming ubiquitous in our society, we find it relevant to ask how it affects the conditions for human expression. Where is the art in artificial intelligence? Do we understand it on a sufficiently deep level that we dare allow it to mediate our deepest thoughts and emotions? Then again, can we afford to neglect the effort of trying to understand it
In an attempt to simplify: What does A.I. really do? For one, it provides a form of automation. Automation of tasks that would otherwise be impractical or impossible to complete for different reasons. By helping us probe areas that would otherwise be unattainable, A.I. serves as an interface. A means to interact with what exist on the other side of the barrier. While A.I. in itself is an interface, it is also used to build other kinds of interfaces. This nested structure complicates understanding.
What is an interface anyway, and what makes it live? We could say it is something that allows an action on one side to have an equivalent effect on the other side. So . . . an interface is a kind of an equal sign is it? In the late 1990’s I did some work with interactive dance together with choreographer Susanne Rasmussen. In providing dancers with sensors, I had the romantic idea that I could capture the expressive qualities of their movements and translate these without loss to sound and music. Perhaps a naive approach, as the richness of combining these artistic expressions may lay just as much in their opposition.
In utilizing the different possibilities of each medium rather than striving for a direct translation between them. Then, the interface does not equal? Looking at the not-equal sign (=), there is a trace of equality, as the symbol = is still there. It is just protruded by a disturbance. Automatic translation programs have shown us the delicacies of translation between languages, but even with careful human authoring, we can’t really say the exact same thing in two different languages. Claiming that “I love you” means the same as “Ich liebe dich” or “Jeg elsker deg”, helps us approach a common understanding. But when translating a text, we more often have to rewrite the whole thing to let the text have meaning in the other language. When we say that language is an interface for human communication, also translation between different languages is an interface. Translation significantly alters the content of the message, and give it new dimensions of meaning, reflected from the culture in which each language has been developed and used. For this reason, we can hardly look at the interface as something separate from those entities it connects.
The theme for a previous ICLI (Lisbon in 2014) was INTER-FACE, in which I read something happening between two faces. The connection point, where communication is made across a border. Thinking of it like a face, a human face, makes it so much easier to include all that lies behind the face.
The interface concerns first what happens in the meeting point, how events from one side is translated into actions on the other side. But in interfacing two environments, it also makes sense to think about the characteristics, constitution, . . . , in short: The nature of those two worlds. The piano keyboard is an interface between (usually human) limbs and the hammer that strikes a string, making it vibrate. Yet, the nature of the action being done here is not merely the physical action of producing vibrations in the string. Usually, when this action is done, we think of it as an act of making music. Making music is related to playfulness, to conveying emotions, building relations between sounds, telling a story, and many other things. It usually means that an idea is formed by one human being, and that it somehow is contained in the musical expression then made by this being, and subsequently received, (hopefully) appreciated, and (perhaps) decoded. In terms of communication theory, we have significant scope for signal loss. Noise. Misunderstandings. Reinterpretations. When we talk about a musical instrument as an interface, all these things also are entangled in the conversation.
The field of art and technology is a meeting point of very unequal values and cultures. The technology part is often also quite naturally bound to science, to the development of new technologies. The methods and values of science meet and intermingle with the methods and values of art. In many ways, we face similar challenges in the field of artistic research. This also, is a hybrid, where the values and methods of research (sometimes confused with science) meet those of artistic exploration and expression. A potential pitfall in this meeting of cultures is the language (interface) used in the reflection, dissemination and validation of results.
Science and technology are commonly concerned with formalization, and the successful formalization of a result is measured as part of its validity. The reflections and results of artistic research (and artistic production and activity at large) are commonly less compatible with such formalizations, but when different fields and cultures meet it is all too easy to give precedence to those with the more clear-cut and unambiguous statements. The field of communication and information theory forms a basis of development for our now ubiquitous computing technologies.
One much-cited paper is Harry Nyquist’s “Certain Factors Affecting Telegraph Speed”. The first sentence in the abstract reads “This paper considers two fundamental factors entering into the maximum speed of transmission of intelligence by telegraph.” Not intending to downplay the value of this research and this field, the use of terms could be noteworthy of a comment. The plurality of meanings associated with the term intelligence show some root of the problem of understanding artificial intelligence today. It doesn’t really help that the field of AI is firmly based on the scientific use of the term intelligence in the military sense, while our expectations often stray to another and more empathic interpretation of the term. Hubert Dreyfus wrote on what computers could not do in 1972, and still could not do in 1992. AI advances in statistical machine learning have since Dreyfus’ critique been successful in overcoming some of the psychological assumptions of earlier AI. It still relies, as far as I can see, on the formalization of knowledge.
What parts of cognition and intelligence can be formalized, or to aim higher, what parts of human behavior can be formalized? This philosophical question has also been researched in psychology and anthropology. Eleanor Rosh’s theories of categorization via prototypes and embodied cognition is one example. Lucy Suchman’s situated cognition is another, where human behavior is understood in dynamic interactions with the material and social worlds. Modern deep leaning techniques attempts to incorporate these approaches by way of learning from examples. For the most part, the algorithm is still blind, and can only use what it is explicitly given. Part of human nature is also curiosity. Can we formalize that?
What would you call an artificial intelligence that is not intelligent? With regards to the complexity of simulation, we could say artificial intelligence popularly refers to some piece of technology that we don’t yet fully understand. Once we can fully understand it, it becomes a mere algorithm, a tool that we can use mechanistically for a given purpose.
This is also why this year’s Live Interfaces attempts to combine automation and emotion, artificial and artistic. As in a hadronic collision, we hope that the photons produced may shed some light on the matter.
The contributions from all the artists and researchers to this year’s conference prods these questions and many more, untangling, exploring, submitting to and conquering the transmission point, the face where worlds meet. We are indebted to your work of keeping it live.
Øyvind Brandtsegg
On behalf of the local ICLI 2020 organizing committee