This series listens to the political, gendered, queer(ed), racial engagements and class entanglements involved in proclaiming out loud: La-TIN-x. ChI-ca-NA. La-TI-ne. ChI-ca-n-@.  Xi-can-x. Funded by an Andrew W. Mellon Foundation as part of the Crossing Latinidades Humanities Research Initiative, the Latinx Sound Cultures Studies Working Group critically considers the role of sound and listening in our formation as political subjects. Through both a comparative and cross-regional lens, we invite Latinx Sound Scholars to join us as we dialogue about our place within the larger fields of Chicanx/Latinx Studies and Sound Studies. We are delighted to publish our initial musings with Sounding Out!, a forum that has long prioritized sound from a queered, racial, working-class and  “always-from-below” epistemological standpoint. —Ed. Dolores Inés Casillas

—

DOI: https://doi.org/10.59350/x88kj-tmw60

Busco la colaboración universal donde todos los Benitos puedan llegar a ser Bad Bunny. –FlowGPT, TikTok

In November of 2023, the reggaetón song “DEMO #5: NostalgIA” went viral on various digital platforms, particularly TikTok. The track, posted by user FlowGPT, makes use of artificial intelligence (Inteligencia Artificial) to imitate the voices of Justin Bieber, Bad Bunny, and Daddy Yankee. The song begins with a melody reminiscent of Justin Bieber’s 2015 pop hit “Sorry.” Soon, reggaetón’s characteristic boom-ch-boom-chick drumbeat drops, and the voices of the three artists come together to form a carefully crafted, unprecedented crossover.

Bad Bunny’s catchy verse “sal que te paso a buscar” quickly inundated TikTok feeds as users began to post videos of themselves dancing or lip-syncing to the song.  The song was not only very good but it also successfully replicated these artists– their voices, their style, their vibe. Soon, the song exited the bounds of the digital and began to be played in clubs across Latin America, marking a thought-provoking novelty in the usual repertoire of reggaetón hits.  In line with the current anxieties around generative AI, the song quickly generated public controversy. Only a few weeks after its release, ‘nostalgIA’ was taken down from most digital platforms.

The mind behind FlowGPT is Chilean producer Maury Senpai, who in a series of TikTok responses explained his mission of creative democratization in a genre that has been historically exclusive of certain creators. In one video, FlowGPT encourages listeners to contemplate the potential of this “algorithm” to allow songs by lesser-known artists and producers to reach the ears of many listeners, by replicating the voices of well-known singers. Maury Senpai’s production process involved lyric writing, extensive study of the singers’ vocals, and the Kits.ai tool.

Therefore, contrary to FlowGPT’s robotic brand, ‘nostalgIA’ was the product of careful collaboration between human and machine– or, what Ross Cole calls “cyborg creativity.”  This hybridization enmeshes the artist and the listener, allowing diverse creators their creative desires. Cyborg creativity, of course, is not an inherent result of GenAI’s advent. Instead, I argue that reggaetón has long been embedded in a tradition of musical imitation and a deep reliance on technological tools, which in turn challenges popular concerns about machine-human artistic collaboration.

Many creators worry that GenAI will co-opt a practice that for a long time has been regarded as strictly human. GenAI’s reliance on pre-existing data threatens to hide the labor of artists who contributed to the model’s output. We may also add the inherent biases present in training data. Pasquinelli and Joler propose that the question “Can AI be creative?” be reformulated as “Is machine learning able to create works that are not imitations of the past?” Machine learning models detect patterns and styles in training data and then generate “random improvisation” within this data. Therefore, GenAI tools are not autonomous creative actors but often operate with generous human intervention that trains, monitors, and disseminates the products of these models.

The inability to define GenAI tools as inherently creative on their own does not mean they can’t be valuable for artists seeking to experiment in their work. Hearkening back to Donna Haraway’s concept of the cyborg, Ross Cole argues that

Such [AI] music is in fact a species of hybrid creativity predicated on the enmeshing of people and computers (…) We might, then, begin to see AI not as a threat to subjective expression, but another facet of music’s inherent sociality.

Many authors agree that unoriginal content—works that are essentially reshufflings of existing material—cannot be considered legitimate art. However, an examination of the history of the reggaetón genre invites us to question this idea. In “From Música Negra to Reggaetón Latino,” Wayne Marshall explains how the genre emerged from simultaneous and mutually-reinforcing processes in Panamá, Puerto Rico, and New York, where artists brought together elements of dancehall, reggae, and American hip hop. Towards the turn of the millennium, the genre’s incorporation of diverse musical elements and the availability of digital tools for production favored its commercialization across Latin America and the United States. 

The imitation of previous artists has been embedded in the fabric of reggaetón from a very early stage. Some of the earliest examples of reggaetón were in fact Spanish lyrics placed over Jamaican dancehall riddims— instrumental tracks with characteristic melodies. When Spanish-speaking artists began to draw from dancehall, they used these same riddims in their songs, and continue to do so today. A notable example of this pattern is the Bam Bam riddim, which is famously used in the song “Murder She Wrote” by Chaka Demus & Pliers (1992).

This riddim made its way into several reggaetón hits, such as “El Taxi” by Osmani García, Pitbull, and Sensato (2015).

We may also observe reggaetón’s tradition of imitation in frequent references to “old school” artists by the “new school,” through beat sampling, remixes, and features. We see this in Karol G’s recent hit “GATÚBELA,” where she collaborates with Maldy, former member of the iconic Plan B duo.

Reggaetón’s deeply rooted tradition of “tribute-paying” also ties into its differentiation from other genres. As the genre grew in commercial value, perhaps to avoid copyright issues, producers cut down on their direct references to dancehall and instead favored synthesized backings. Marshall quotes DJ El Niño in saying that around the mid-90s, people began to use the term reggaetón to refer to “original beats” that did not solely rely on riddims but also employed synthesizer and sequencer software. In particular, the program Fruity Loops, initially launched in 1997, with “preset” sounds and effects provided producers with a wider set of possibilities for sonic innovation in the genre.

The influence of technology on music does not stop at its production but also seeps into its socialization. Today, listeners increasingly engage with music through AI-generated content. Ironically, following the release of Bad Bunny’s latest album, listeners expressed their discontent through AI-generated memes of his voice. One of the most viral ones consisted of Bad Bunny’s voice singing “en el McDonald’s no venden donas.”

The clip, originally sung by user Don Pollo, was modified using AI to sound like Bad Bunny, and then combined with reggaetón beats and the Bam Bam riddim. Many users referred to this sound as a representation of the light-heartedness they saw lacking in the artist’s new album. While Un Verano Sin Ti (2022) stood out as an upbeat summer album that addressed social issues such as U.S. imperialism and machismo, Nadie Sabe lo que va a Pasar Mañana (2023) consisted mostly of tiraderas or disses against other artists and left some listeners disappointed. In a 2018 post for SO!, Michael S. O’Brien speaks of this sonic meme phenomenon, where a sound and its repetition come to encapsulate collective discontent.

Another notorious case of AI-generated covers targets recent phenomenon Young Miko. As one of the first openly queer artists to break into the urban Latin mainstream, Young Miko filled a long-standing gap in the genre—the need for lyrics sung by a woman to another woman. Her distinctive voice has also been used in viral AI covers of songs such as “La Jeepeta,” and “LALA,” originally sung by male artists. To map Young Miko’s voice over reggaetón songs that advance hypermasculinity– through either a love for Jeeps or not-so-subtle oral sex– represents a creative reclamation of desire where the agent is no longer a man, but a woman. Jay Jolles writes of TikTok’s modifications to music production, namely the prioritization of viral success. The case of AI-generated reggaetón covers demonstrates how catchy reinterpretations of an artist’s work can offer listeners a chance to influence the music they enjoy, allowing them to shape it to their own tastes.

Examining the history of musical imitation and digital innovation in reggaetón expands the bounds of artistry as defined by GenAI theorists. In the conventions of the TikTok platform, listeners have found a way to participate in the artistry of imitation that has long defined the genre. The case of FlowGPT, along with the overwhelmingly positive reception of “nostalgIA,” point towards a future where the boundaries between the listener and the artist are blurred, and where technology and digital spaces are the platforms that allow for an enhanced cyborg creativity to take place.

—

Featured Image: Screenshot from ““en el McDonald’s no venden donas.” Taken by SO!

—

Laurisa Sastoque is a Colombian scholar of digital humanities, history, and storytelling. She works as a Digital Preservation Training Officer at the University of Southampton, where she collaborates with the Digital Humanities Team to promote best practices in digital preservation across Galleries/Gardens, Libraries, Archives, and Museums (GLAM), and other sectors. She completed an MPhil in Digital Humanities from the University of Cambridge as a Gates Cambridge scholar. She holds a B.A. in History, Creative Writing, and Data Science (Minor) from Northwestern University.

—

REWIND!…If you liked this post, you may also dig: 

Boom! Boom! Boom!: Banda, Dissident Vibrations, and Sonic Gentrification in Mazatlán—Kristie Valdez-Guillen

Listening to MAGA Politics within US/Mexico’s Lucha Libre –Esther Díaz Martín and Rebeca Rivas

Ronca Realness: Voices that Sound the Sucia Body—Cloe Gentile Reyes 

Echoes in Transit: Loudly Waiting at the Paso del Norte Border Region—José Manuel Flores & Dolores Inés Casillas

Experiments in Agent-based Sonic Composition—Andreas Pape