The paper begins with an explanation of my approach toward analyzing the media coverage of the release of “Where That Came From,” followed by a mostly chronological overview of the reporting about its production and release. After this descriptive overview, I will analyze four themes that run through the media coverage: the promise of “giving back his voice” to Travis, the way in which the relationship between human and technological labor is presented in this story, its framing as a human-interest story, and finally, its broader implications in terms of the strategic and financial interests that are at play.

This announcement of 15 May 2024 raised my curiosity. I found a review for the concert reporting that the CBS Sunday Morning segment on the song was featured on a big screen in-between songs (Anderson). Subsequently, Dupré “performed ‘Where That Came From,’ pausing to allow the audience to take in the rebirth of the Grammy winner’s soothing sound for a portion of the track” (Anderson).

The media coverage of “Where That Came From” during the two-week period spanning the song’s release follows a clear and consistent pattern, and closely follows the narrative arc provided by the promotional strategies of Travis and Warner Nashville. In this dominant framing, the making of the song is reported as a human-interest story about how Travis was, after years, “given back” his voice, thanks to cutting-edge AI techniques. As I argue with the help of studies of similar dynamics in relation to technological advances in the field of biotechnology (Dow; Estreich), such a framing of the story as a human-interest story contributes to the normalization of AI, as it provides readers and viewers with a familiar point of reference for a new and potentially disruptive or controversial technology.

For anyone familiar with Travis’s life story, the significance of this pledge was immediately apparent, considering the singer had been left with irreparable damage to the parts of the brain that are needed to sing or speak after a serious stroke in 2013. “Where That Came From” marks his first release of newly recorded music in over a decade.

Other concerns remain implicit in the reporting of the case; for instance, how this initiative is used by the record company to position itself as a key player in the development of ethical and responsible uses of AI, or how it may be integrated into lobbying efforts for further regulation of AI. In other words, we are witnessing here what Crawford has argued to be true for AI in general: that artificial intelligence involves an interplay of “technical and social practices, institutions and infrastructures, politics and culture” (8), but also a tendency to downplay the social and political interests that are at stake. With the help of work at the intersection of disability studies and sound studies (Friedner and Helmreich; Mills), I have argued that we are likely dealing with a case where a technology, which was originally presented with the promise of helping to compensate for a disability, will end up preparing the ground for other uses of the technology that are stripped of such lofty promises. To return to Mara Mills’s notion of an “assistive pretext” to describe this type of dynamic, one may wonder what this specific assistive pretext of helping the singer who had lost his voice to sing again, may turn out to be a precursor to. In the long run, after all, it seems unlikely that use of similar applications of AI will be limited to recreating the voices of singers who have lost their ability to sing, but not their ability to make public appearances affirming their consent.

In the social media campaign leading up to the release of the song in early May 2024, Travis reassured his fans: “In the meantime, just know that when it comes to me singing songs for you, there’s always more where that came from” (qtd. in Bartiromo and Petrulis)

In developing my analysis, I will draw on insights from the academic fields of sound studies and popular music studies, concerning ongoing developments and challenges related to the application of AI tools to the production of music (Keith et al.; Sterne and Razlogova; Bonini and Magaudda), the use of sound in assistive technologies for people with disabilities (Friedner and Helmreich; Mills), as well as notions of aesthetics and authenticity in country music (Peterson; Pecknold; Stimeling; Malone). Additionally, I will make use of scholarship in the field of science and technology studies (STS), which have studied similar dynamics in relation to new and controversial technologies not only in artificial intelligence (Gray and Suri; Crawford), but also in domains such as biotechnology (Dow; Estreich).