SATURDAY 9 NOV 14:00-17:00
Panel: Benoit Baudry, Björn Bjurling, Roberto Bresin, Ludger Brümmer
Moderator: Henrik Frisk

What are the artistic implications of AI?

Articial intelligence (AI) in music production has a long history, and specic technologies have been created for diferent reasons and with diferent purposes. AI is already used in many areas of music production: audio processing plug-ins, composition tools, analysis tool and many other types of software explore diferent kinds of machine intelligence in order to assist musicians to better perform their tasks. To distinguish what part of the result is actually the outcome of machine, rather than human, intelligence is not a distinction that is easily drawn, nor is it necessarily an interesting one.

Furthermore, it is a feature of connectionist strategies for articial intelligence that the structures that guide them are hidden. The use of technology in artistic practices, however, can have diferent objectives. First, from a general standpoint one may argue that art should engage in available technologies for the simple reason that this contributes to our understanding of its social and cultural impact. Though this general notion is sometimes contested, most famously by Heidegger who instead warned about the ways in which technology frames the human capacity, it may be considered equally reasonable to claim that AI should be a eld for artistic exploration (Heidegger, 1954). 

Secondly, from the point of view of innovation, it has been seen in the past that artists’ use of technology push the boundaries for what is possible (Harris, 1999, e.g.). Although this has arguably been true, the resources that the multinational technology and media industry nowcontrol, along with the increased complexity of the technology, make it more dicult for an independent artist, or even an institution or a university, to invent methods of production that can compete with the powers of these companies. The artistic qualities in and of themselves may be uncontested, but as far as creative power is concerned the playing eld has changed.

Thirdly, the challenge to create a machine that is able to compose music with the same level of integrity, witfulness and inspiration as that of a great musician may be seen as a test of the potentiality of the technology. Amachine that can compose music is likely to be able to perform a whole range of tasks with a high level of prociency. However, unlike, say, a self driving car, it is dicult to fully grasp the (economic) value of such an achievement.

There is an underlying ethical dilemma that has impact on the whole eld of AI, also in the eld of music. The emergence of AI depends on designers and programmers that create the systems and as has been shown through studies such as (Snow, 2018) technology as a valueless blank slate may be contested. With an increased use of intelligent technology in artistic practice, there is a broadening of the practice that may include not only the programmer of the software but also the designer and manufacturer of the hardware. And following this there are at least two intriguing questions that arise:

1. What, if any, are the artistic and ethical responsibilities in the development of AI in music? What are the implications on aspects such as copyright?

2. If musical machine intelligence is truly possible, to what extent are we prepared for it to develop its own musical aesthetics? After all, a body less musician is likely to develop a diferent musical conception than a human embodied musician. Or, to summarize: What are the artistic implications of AI?

Prof. Ludger Brümmer: Creativity versus Reproduction - Artificial Intelligence in Music Composition

Since computing machines have been used in music the role of the composer and musician as the creative initiators in the chain of artistic creation had to be discussed. This lecture highlights the gradual evolution of artificial intelligence, analyzes the aspect of creativity in algorithmic composition strategies, and discusses aspects of metadata versus sound as a source for electronic composing. After an historic embedment of machine learning in the field of music composition, current activities of the ZKM | Hertz-Lab are presented that are related to artificial intelligence and music.

Benoit Baudry: Dissecting software execution

Benoit Baudry is a Professor in Software Technology at the KTH Royal Institute of Technology, where he leads the CASTOR software research center. He received his PhD in 2003 from the University of Rennes and was a research scientist at INRIA from 2004 to 2017. His research
interests include dynamic code analysis, software testing and
diversification and software art.

Professor in Software Technology, KTH Royal Institute of Technology https://softwarediversity.eu
Director of the CASTOR center for software research

Roberto Bresin
The design of AI-based applications for music creation needs information from basic research.

In this presentation I will talk about the importance of basic research in the field of sound and music computing and how results can eventually be used for the design of new AI-based applications that can support the creative process of composers and performers.

I will briefly present some results from our research on emotions in music performance and on sonification of movements and speculate how these could be used to inform the design of new AI-based applications, which in turn are limited by the results from basic research.

Roberto Bresin is Professor of Media Technology at the Division of Media Technology and Interaction Design (MID), School of Electrical Engineering and Computer Science (EECS) at KTH. He is also Director of Studies for the Degree Programme in Media Technology at KTH, and director of NAVET, a new KTH center for research art, technology and design.

His research interests cover among others the fields of expressive music performance, emotion in sound and music performance, sound in interaction, and sonification. He has published several research papers in peer-reviewed journal and conferences. At the moment is principal investigator of a project funded by the Swedish Research Council, SONAO – Robust Non-Verbal Expression in Virtual Agents and Humanoid Robots: New Methods for Augmenting Stylized Gestures with Sound.

He is member of the steering committee of the Sound and Music Computing Network.

Björn Bjurling, Senior Researcher, PhD

In my talk I will first give a short overview of my favorite AI topics. Then I will give some examples of applications of AI in industry. Finally, I will share some personal impressions regarding AI and music.

Bjorn Bjurling earned his PhD from the Department of Computing at Imperial College London in 2006.

From 2006, Bjorn has been working as a senior researcher at the Swedish Institute of Computer Science (SICS). (From 2018, SICS is a part of RISE AB). Bjorn has worked with applied AI research in close collaboration with industrial partners in the areas of Telecom, Automotive, Civil Defense, and Health care.

Bjorn's main research interests are in the use of logic and automated reasoning for formalization of industrial problems. In recent years, Bjorn has also been involved in projects with more focus on statistical machine learning and computer vision.

Prior to Imperial, Bjorn studied composition at the Royal Conservatory of Music in Stockholm and he has had a brief career as composer.