From Hacker Poets to Cybernetic Poets

By Terry Trowbridge and Joseph Alexander Brown (Twitter)

Generative poetry and narrative programs have become mainstream projects for all levels of programmers and even gamer culture. Professional programmers design small scale poetry bots in their spare time. Undergraduate Computer Science students are frequently assigned little projects that test their skills with combinatorics and dictionaries. The ability to code a generative algorithm to mimic the vocabulary or style of a poet is a standard way to put the “STEM” in a “Science Technology Engineering Arts and Mathematics (STEAM)” curriculum. Literary criticism is quickly catching up, with more journals of literary studies asking for more generative poetry theory each year. Casual literary criticism in newspapers and blogs have moved away from asking “Is it art?” to more substantial reviews.

What is the next step for generative poetry? In a fast-paced world with short attention span, how does generative poetry resist a swift death as a fad, while also avoiding the slow stagnation of becoming more of a cliché than a genre? What justifies generative poetry as experimental instead of just reiterations of a standard toolkit? How is a poetry generator moving beyond just a developer of optimized literal strings of characters and into something of an art?

We propose that generative poetry needs to start working its way into live poetry readings, beyond the framework of production of poetry as an artifact and into the idea of poetry as performance. In particular, we believe that poetry programmers need to take their poems to the open mic sessions that bookend most community-based poetry readings. This aim is as much of a tactical maneuver, for recognition of the contribution they make to art in their local community, as it is a living laboratory to test their mettle like any human poet in the postmodern, intertextual, twenty first century.

It is time for hacker poets to find out if they can “hack it,” pun intended, in open mic poetry readings. Someone who codes generative poetry algorithms is contributing to the literature of their neighbourhood, city, and nation. They deserve some stage time among the other poets. We also suggest that while some poetry is cliché, all poets are eccentric and have a unique voice. All poets are cybernetic, in the sense of Wiener’s definition of “the scientific study of control and communication in the animal and the machine.” This creates a new definition of the Cybernetic poet who not only uses the tools but is able to present the expression.

Some read their poetry off of their smartphones. Other poets incorporate sound effects like looping and distortion in their readings. Others add images to their stagecraft that doesn’t appear on the page or poetry blog. Some poets have artificial limbs, speak through vocoders, or simply change the settings on the microphone/amp setup on stage. All manner of performance poets utilizes the live venue to create unique artistic experiences.

We assert that generative poetry is equally as cybernetic as the other forms of poetry that happen on

stage, especially when computer generated poems are performed through the voice, body, and presence of a human. We will not know what that means for each AI poet, however, until they take the stage. The Turing test is not enough to examine a poem and only looks at the ideas development method, and not the importance of the expression of the ideas on stage.

All live performances are somewhat rehearsed, in that they are imagined beforehand and then altered by the stage, the lighting, the acoustics, and a host of variables in the performance. Because performances are shared with an interactive audience, the audience brings their own rehearsed role, whether that means silently paying attention, kibitzing the stage, or being bored and texting rudely. Each poetry reading series has a culture of regulars who form cliques, and one-shot audience members. Since the poet is surrounded by so many variables, their literature is subjected to strong forces of improvisation. These forces, we contend, are the ancient cultural codes that constitute living generative features. They change the form and the content of otherwise static text into dynamic new artwork, at the moment.

Therefore, we should not, as hacker poets, underestimate the opportunity to treat open mic as a place to engage comrade poets of all stripes in their literary criticism. Perform our guts out in front of the poetry cliques and listen to their responses. Pay attention to their stage presence and consider what each of their presentation styles could do to alter or challenge our own computer-generated verses.

These responses are feedback into how the next bot should be designed and how to push forward a new expression. The expression now becomes the domain of the poet and computer, they work together as a team, the generator and the expressed presented work.