A.I., Machine Learning, Culture & The Creative Process

From the 2003 exhibition, “The Quilts of Gee’s Bend.”

It started with an algorithm. A few years ago, I taught AP Computer Science Principles to students at a high school for the visual and performing arts and I was looking for ways to link art and computation. To explain conditional statements I referenced Meghan Trainer’s notion of Sojourner Truth’s famous speech as an if-then statement that tells a computer program to execute a certain section of code only if a test is shown to be true. For example:

void slavery() {
    // the "if" clause: person bears children and does as much work as any man
    if (isWoman) 
        equalRights--;
}

As part of my research, I discovered many instances of “coding” to hack systems (of oppression). Songs, speeches, quilts, etc. are all ways in which forcibly displaced, dispersed and enslaved people have used to disrupt systems and invent new realities. This work led to my interest in culturally relevant computational thinking.

I challenged the high school students to find algorithms in everyday life such as in quilting with Professor Audrey Bennett who explores and shows how students can intellectually harness the computational thinking concepts embedded in the cultural traditions of tiling geometric shapes into meaningful, rhythmic patterns that take form as quilts.

Professor Audrey Bennett guiding my AP CSP students at Boston Arts Academy

The AP CSP students’ original quilt-based designs using the Quilting CSDT.

Computational thinking (CT) has been described as the use of abstraction, automation, and analysis in problem-solving. Thinking computationally draws on the concepts that are fundamental to computer science, and involves systematically and efficiently processing information and tasks.

Animate CSDT (circa 2011). Courtesy of Libby Rodriguez and Ron Eglash.

Algorithm design is creating an ordered series of instructions for solving similar problems or for doing a task. The quilting tool (CSDT) the students used allowed them to computationally make an original quilt using a traditional pattern. Later that year, I showed the students DeepDream, a program that uses machine learning algorithms to create images.

BAA student “Travora” + Deep Dream Generator

Machine learning is a branch of artificial intelligence (AI) based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention. Machine learning is a specific subset of AI that trains a machine how to learn. To create the DeepDream image (above) I used the Gatys’ Gram-matrix based Neural Style Transfer, an image stylization algorithm.

My NSynth Super AI music making machine

Emerging technology combines visuals, sound and artificial intelligence (AI). AI can be used to create new ideas by producing novel combinations of familiar ideas, exploring the potential of conceptual spaces, and by making transformations that enable the generation of previously impossible ideas. I wrote an essay for a book titled “Glitched: Spacetime, Repetition & The Cut.” I shifted my focus to music and discovered NSynth Super, a Raspberry Pi and machine learning powered instrument for generating new unique sounds that exist between different sounds. I was so excited that I made one (above).

My goal with NSynth Super is to explore the possibilities of sound, specifically building on the hip hop production I was involved in years ago (i.e. sampling, cutting, remixing). I want to know what sounds exist between samples and can NSynth usher in a new era for hip hop/electronic music. Like any technology, machine learning devices are tools. The creativity (artist’s hand) is still human.