A few years ago I gave a short talk at my alma mater about computational thinking and culture. I presented on my theory about the algorithmic nature of James Brown’s “Cold Sweat”, which is has been cited as the first true Funk song. It is the first recording in which Brown calls for a drum solo (“give the drummer some”) from Clyde Stubblefield, beginning a tradition of rhythmic “breaks” that would form the foundation of sampling.
The next day we pulled up in front of King Records studio, got off the bus, got in the studio, set up, and I went over the rhythm with the band. By the time we got the groove going, James showed up, added a few touches — changed the guitar part, which made it real funky — had the drummer do something different... He put the lyrics on it. The band set up in a semicircle in the studio with one microphone. It was recorded live in the studio. One take. It was like a performance. — Pee Wee Ellis
The reason I compare this to computational thinking is because of the following elements or concepts:
- Decomposition — In order to produce the song, co-writers Brown and Pee Wee Ellis had to break it down into smaller, more manageable parts such as the rhythm, the groove, and the guitar part.
- Pattern Recognition — Brown had mastered the process of identifying patterns or connections between different parts of the song. He was quickly able to identify the details that were similar–or different–as well as build a continued understanding of the complexity of the song.
- Abstraction — Brown was able to extract the most relevant information from each part of the song, which helped to define, or generalize what, exactly needed to be done to compose an entire song (through repetition or looping certain parts).
- Algorithmic thinking — Brown and his band used the process of defining a step-by-step solution to composing a song (problem) that can be replicated for a predictable and reliable outcome.
I keep coming back to this and no just “Cold Sweat” but funk music, in general, because the same polyrhythmic patterns you hear in funk music you can see visually in works such as African textiles and quilts:
This process (making “Cold Sweat”) informs my thinking about generative artificial intelligence, especially the process of using prompts to create images. In my Generative AI Tutorial I explore the anatomy of the prompt, in which you see some of the computational thinking concepts mentioned above such as decomposition or breaking down a prompt into parts:
Whether in the computer science industry or in everyday life, computational thinking can be an extremely helpful tool in defining and solving complex problems. Through experimentation, I created a series of prompts or parts of prompts that can be replicated for predictable and reliable outcomes. Like sampling, I use prompts to bring in elements such as embroidery and Kuba cloth patterns.
The best part of my Georgia Tech talk was seeing a computer science professor nod his head in agreement. He was excited and I figured there must be some truth to the theory about “Cold Sweat”, so there may also be some truth in being purposeful about prompting using generative AI. Like James Brown, we can add touches (images and text) and combine or remix elements to create something entirely new.
Note: Taking this approach to explaining computation (computer science) may also engage historically marginalized and underrepresented groups who don’t see themselves in the field or discourse. This is why I keep coming back to the process in my own work.