Machine Learning and the Arts

2022-23 Visiting Artist

Andreas Refsgaard's Doodle Tunes. Courtesy of the artist.
Andreas Refsgaard's ml5. Courtesy of the artist.

Exploring the creative potential of emerging digital technologies

About

During his two-week visit to the MIT campus, creative coder Andreas Refsgaard shares his perspectives from the field between art and interaction design. 

The centerpiece of Refsgaard’s residency is a talk titled “Machine Learning and the Arts,” providing examples of emerging digital technologies that can support artists and content creators in the arts. Refsgaard demonstrates some of the tools and techniques of machine learning—many of them open-source—that he uses to create his playful, interactive projects.

These include Doodle Tunes, which converts hand-drawn images of instruments into musical tracks; Eye Conductor, an interface designed for people with physical disabilities that allows the user to play music through eye movements and facial gestures; Perler Beads to Landscapes, which transforms simple plastic crafts into realistic-looking photos; and Booksby.ai, an online bookstore that sells science fiction novels authored by artificial intelligence. 

Refsgaard uses a sense of humor and fun to involve his audience in an exploration of the possible benefits, drawbacks, and ethical dilemmas that such technologies could bring to our daily lives.

The residency includes a digital exhibit of Refsgaard’s work at the MIT Lewis Music Library, a campus hub for experimentation in audio production, editing, and creation. The library is also recording an interview with the artist for its longstanding Music at MIT Oral History Project. 

While on campus, Refsgaard is teaching two workshops geared toward MIT students of any major who are working on machine learning technologies with potential applications in the arts. He is also visiting MIT Music and Theater Arts classes to provide guidance on interdisciplinary student projects in such areas as electronic music composition, digital instrument design, and interactive music systems.

Schedule

Events

Doodle Tunes Digital Artwork Exhibition
Opens September 26, 2022
Lewis Music Library, MIT Building 14
160 Memorial Drive, Cambridge, MA


Sounds From The Mouth Soundscape Interactive Installation
Opens September 26, 2022
Lewis Music Library, MIT Building 14
160 Memorial Drive, Cambridge, MA


Workshop: Art, Algorithms and Artificial Intelligence
Thursday, September 29, 2022 / 5:00-7:00pm
Lewis Music Library, MIT Building 14
160 Memorial Drive, Cambridge, MA

In this hands-on workshop, artist and interaction designer Andreas Refsgaard demonstrated how to use machine learning algorithms in playful and poetic ways. Participants collected data and trained their own image classifier, generated algorithmic poetry, and lots more!

The workshop was suitable for people with no prior experience with code or machine learning, who are interested in getting acquainted with the potentials these technologies hold.


Workshop: Machine Learning for Interaction Designers
Wednesday, October 5, 2022 / 5:00-7:00pm
Lewis Music Library, MIT Building 14
160 Memorial Drive, Cambridge, MA

In this workshop, artist and interaction designer Andreas Refsgaard provided an overview of image classification and face, body, and finger tracking tools to support interaction design. Through creative examples and hands-on exercises, he rolled out the many creative and technical possibilities machine learning brings to the field, such as offering new and effective ways of prototyping and testing ideas.

Experience with p5.js, Processing, or similar creative coding frameworks is recommended but not a requirement. Participants provide a functional laptop with a webcam and Google Chrome installed.


Public Artist Talk: “Machine Learning and the Arts”
Thursday, October 6, 2022 / 5:30-7:00pm
The Nexus, Hayden Library, MIT Building 14S-130
160 Memorial Drive, Cambridge, MA

Andreas Refsgaard, creative coder and 2022-23 CAST Visiting Artist, explores ways that emerging digital technologies can support artists and content creators in the arts. Refsgaard will demonstrate some of the tools and techniques of machine learning—many of them open-source—that he has used to create his playful, interactive projects. 

Andreas Refsgaard is an artist and creative coder based in Copenhagen. Working in the field between art and interaction design, he uses algorithms, coding, and machine learning to explore the creative potentials of emerging digital technologies. In his artistic practice, Refsgaard applies a humorous and cheerful approach to the digital tools and their potential applications, purposes, and values. 


Class Visits
21M.080 Intro to Music Tech
21M.380 Music and the Internet
21M.470 FaMLE, the MIT laptop ensemble

Collaborators

Andreas Refsgaard is an artist and creative coder based in Copenhagen. Working in the field between art and interaction design, he uses algorithms, coding, and machine learning to explore the creative potentials of emerging digital technologies. In his artistic practice, Refsgaard applies a humorous and cheerful approach to the digital tools and their potential applications, purposes, and values. His works often consist of unconventional combinations of inputs and outputs that allow people to, for example, play music using eye movements, control games by making silly sounds, or transform drawings of musical instruments into real compositions.


Website: andreasrefsgaard.dk
Social: Facebook | Instagram | LinkedIn | Twitter 


Avery Boddie, MIT Libraries Music and Media Department Head, manages the daily operations of the Lewis Music Library while envisioning and implementing digital initiatives related to music, audio, and multimedia. Boddie is also responsible for the management of the music library’s vast collection of scores, media, audio, and recording technologies. He also works closely with MIT’s Music and Theater Arts department to support research and instruction in all areas of music

Biography: MIT Libraries
Social: Instagram | YouTube