Artificial Intelligence & the Artist’s Hand: Blues Edition
The book Abstracting Craft: The Practiced Digital Hand by Malcolm McCullough has been very relevant to my research and artistic practice, especially his exploration of “craft.”
When the tools are complex, when the artifacts produced are abstract, or when tools provide the only means of access to the medium (all common conditions in high technology), it can be difficult to say where a tool ends and a medium begins. But we can say that under skilled practice even these tools become transparent, and that a sense of a medium eventually emerges.
I’m making the argument that medium, as described by McCullough has a correlation in the blues (music), specifically in the call-and-response interaction between the performer and the instrument. A call-and-response is a succession of two distinct phrases usually written in different parts of the music, where the second phrase is heard as a direct commentary on or in response to the first. Take, for example, Skip James who I recently discovered. James was an American Delta blues singer, guitarist, pianist and songwriter.
Call-and-response patterns can also be found in deep neural networks, especially neural style transfer, which uses neural representations to separate and recombine content and style of arbitrary images, providing a neural algorithm for the creation of artistic images. In A.I. music, the system interpolates between any two sounds, to produce a new sound. Listen closely and you can hear it in James’ vocalizations that respond to the sounds from his guitar. I kept this in mind when I produced new Deep Dream images using photos of musicians who take inspiration from blues music.
This quality of engagement is personal. Something must draw your interest. This might be the pleasure of handling a material. It could be the concentration required not to ruin a piece of work. At a more abstract level, it could be the intricacies of solving a problem, whether technical or conceptual. It could be the anticipation of a finished product. It might be the ambition to succeed, or the fear of failure. Or it could simply be the calming effect of routine, based on soothing motions, habitual expertise, and a sustaining commitment to practice.
In the comments section of a post about my work, researcher Aaron Hertzmann noted that my results “don’t just look like the usual automatic output, they’re better and they show the artist’s hand in making them fully realized, even though the elements of the algorithms are still visible.” This is what brought me back to McCullough’s “practiced digital hand.” He writes that for a medium to be engaging it must also be dense (deep). He notes that the medium must “surround us in possibilities” and it is through immersion that we can “coax the medium from one state to another.”
Density supports engagement not only through continuity but also through variety. Only countless subtle differentiations of conditions will yield a heightened, satisfactory practice. A rich medium offers such an extent of possibilities that no one author or piece can incorporate them all, and only this is enough to sustain continued exploration.
In addition to density, McCullough notes that there should be enough states of the medium to create a sense of a continuum of possibilities. Continuity depends on the condition that between any two states there exists still another. It also means sensations of states cannot be disjoint: a neighboring state must appear, and feel, nearly like the present state. Only such density will produce continuous behavior that can be worked with continuous hand-guided processes, like coaxing a material.
Having reached my goal of producing at least one Deep Dream image I liked per day for one year, I still find that there is more to discover and learn in the process. I used the same styles to create the images above but the results were different based on the decisions I made, as well as the source images I selected. Many of my choices were in response to what the machine was doing, much like the blues singer/musician who improvises with sounds.