Approaching artificial intelligence (A.I.) much like any other art form — with a hint of doubt and a focus on its origins — could potentially reduce the sense of threat associated with its creations.
The year 2023 brought forth a myriad of intricate emotions, thoughts, and experiences. It marked significant milestones such as my daughter’s initial steps and words, a standstill in film and television productions due to artists advocating for new rights in a constantly evolving industry, and a somewhat uncertain future. Amidst these events, we grappled with the moral implications of a novel technology that, while unable to replicate my daughter’s first words, is perceived by many as an existential menace to human essence, encompassing our artistic expression: artificial intelligence.
Being a composer and pianist, my primary method of delving into, processing, and articulating complex emotions revolves around music. The arts possess the remarkable ability to translate the inexplicable into something comprehensible to both the gut and the heart. Even in moments of linguistic inadequacy, engaging in art creation or appreciation remains an immensely profound — and inherently human — experience. However, as we transition into 2024, I find myself pondering on how we can navigate the intersection of technology and the arts, particularly the emotionless, data-driven creative process of A.I.
Let me clarify: I harbor a deep appreciation for technology. Increasingly, I find myself inadequately informed about its functionalities, its impact on me through learning, listening, or tracking, and the extent to which it intertwines with my daily life. Nonetheless, I cannot envision my life devoid of it. Technology plays a pivotal role in my work as a composer, necessitating the creation of a fully digital rendition of a film score before its orchestral recording (which invariably pales in comparison to the live performance). As an enthusiast of music and film, I voraciously consume every available score, recording, movie, and interview accessible on the internet. The ability to peruse an analysis of a Mahler symphony and uncover the samples Timbaland used in the “Work It” beat within a brief five-minute Instagram break continues to captivate me.
It is this insatiable quest for knowledge, along with other mental acquisitions, that I draw upon when composing music. Naturally, I gravitated towards ChatGPT shortly after its latest version release last year to explore its potential for facilitating learning and assimilation. What I encountered was simultaneously exhilarating and unsettling.
The excitement stemmed from requesting recommendations for ten ocean-inspired classical music pieces and iteratively refining that request to unveil nearly twenty previously unheard classical compositions. Conversely, the disquiet arose when soliciting names of Black composers, only to have white composers listed in the generated compilation. Or when tasking it to craft a scene mimicking a specific writer’s style, and despite the peculiar phrasing and anomalies typical of A.I.-generated content, discerning a disconcerting sense of familiarity in the scene’s ambiance. I began to question its reliability. How could I place trust in an information source devoid of fact-checking or research? If we were to entrust A.I. with preserving our history for posterity, what narratives would it reshape? Would poorly documented tales that lack a digital footprint fade into obscurity?
After a period of contemplation, I revisited ChatGPT with a fresh perspective, influenced by my early professional experiences.
My journey commenced as a composer for documentaries, during which I ventured into independent filmmaking. Immersed in this realm, I developed a fondness for the cinéma vérité style, characterized by observational filmmaking with minimal narration, striving to capture events without the filmmaker’s intrusion into the narrative. I was enamored by the rawness of the cinematography and the decision to simply observe. The protracted, uninterrupted sequences of individuals in their natural state exude a stark beauty. However, as I delved deeper into the medium’s nuances, I became intrigued by the concept of “truth” in cinema. The camera operator selects what to film, the editor determines what to excise, and the director ultimately shapes the narrative from the footage. How can these individuals entirely eliminate their inherent biases from the creative process?
This notion resonated with jazz improvisation. As improvisers, we remain receptive to the present moment, drawing inspiration from the bassist’s note or the drummer’s rhythm that emerges spontaneously. Yet, even the most innovative improvisers retain their distinctive sound. They exhibit inclinations, a unique touch and tone on their instrument that persist regardless of the occasion. I conjecture that a similar phenomenon applies to every artist practicing cinéma vérité. Certain cinematographers prefer intimate shots, while specific editors gravitate towards a particular editing pace or rhythm. Our creative output is inevitably influenced by our backgrounds and how we perceive the world. Even within an art form denoted as “truthful cinema,” the notion of “truth” remains filtered.
The same principle extends to A.I. Initially, I approached ChatGPT anticipating an all-knowing fount of information, only to encounter a program’s approximation of responses based on its internet-derived database. Instead of an educational mentor, I encountered a product shaped by misinformation and copious textual data, capable of instructing an algorithm on writing but lacking the capacity to comprehend. As an artist, I found solace in the realization that A.I. cannot decipher how we translate life experiences and profound emotions into written words or musical notes. An A.I.-fabricated imitation of an artist will invariably retain traces of the internet’s randomness and fallacies.
Acknowledging this reality and viewing A.I. as simply another artistic form — information sourced from diverse perspectives and filtered through various lenses — transforms it into a compelling wellspring of inspiration. As a society, it is imperative that we equip these algorithms with precise, inclusive, and extensively reviewed information. As an artist, I await with anticipation the evolution of A.I. and intend to respond to its creations akin to any other artist: with a degree of professional skepticism and a keen eye on the origins that influenced it.
Kris Bowers, an Academy Award- and Emmy-nominated composer, filmmaker, and musician, has composed scores for upcoming films such as “The Color Purple” and “Origin,” in addition to previous works including “Green Book,” “King Richard,” and the Netflix series “Bridgerton.”
