‘I’m used to strange in the studio,” says Holly Herndon. Even so, the experimental electronic composer has never ventured anywhere quite this strange before. For her forthcoming third album, Proto, she has teamed up not just with fellow musicians, programmers and members of the public, but a “baby” called Spawn.
Spawn is quite advanced for a newborn. She can mimic, interpret and develop musical ideas, often revealing elements in Herndon’s compositions that she was unaware of. That’s because Spawn has actually been built from artificial intelligence – Herndon collaborated with the AI expert Jules LaPlace in her endless quest to find fresh new sounds. But don’t fret that the robots are taking over just yet.
“I know I’m known as ‘laptop girl’,” she says, laughing, “but I’m always asking myself: where does the human performer fit in to this? How do we continue to develop without automating us off the stage? This frees us up to be more human together.”
It turns out that this technological tamperer is, at heart, a people person. Born in Tennessee, Herndon is now based between Berlin (where she is rooted in the city’s restless music scene) and California (where she has just completed a PhD in composition at Stanford University). After her 2015 breakthrough, Platform, and a stint supporting Radiohead on tour, Proto is her next step into the unknown.
We meet in a peaceful square in west Berlin, and Herndon is sunny and open, eager to dig into her music’s complications with a breezy “let’s unpack that!” Over the next few hours, she speeds through everything from obscure musicology, the ethics of hologram tours and the difficulties with DJ sets in a streaming-based economy (“it’s basically a neoliberal hellscape”). We start, though, with the unusual composition methods behind Proto: Herndon would write a score, record it with the ensemble, process the results in her studio and then feed it to Spawn before returning its output to a group comprising LaPlace and her partner Mat Dryhurst, among numerous other collaborators. The results see plaintive folksong-style singing fed through electronic production to create cavernous choral soundscapes.
Spawn was introduced to the world via last year’s single Godmother – a juddering, abrasive track made with the Indiana producer Jlin, in which the AI babbles like a baby raised in Berghain. Indeed, the technology is still in its infancy. Herndon likens it to the birth of recording technologies in the 1900s, and explains that the first-born metaphor stuck because “artificial intelligence doesn’t have context – much like a child. We’re encoding our values [within it].”
Much of the research into AI and music is designed to benefit big corporations – for example, Warner Music recently acquired Endel, an app for algorithmically composed mood music. “They’re like: how do I get the system to compose a Hans Zimmer score for me, so that I don’t have to pay an artist?” she shrugs. Herndon is bemused by such unimaginative use of world-changing technologies. “I don’t want to recreate music; I want to find a new sound and a new aesthetic. The major difference is that we see Spawn as an ensemble member, rather than a composer. Even if she’s improvising, as performers do, she’s not writing the piece. I want to write the music!”
These changes in creative ownership are set to up-end some of music’s oldest power imbalances. “There’s often this extreme hierarchy between composer and performer; it’s a traditionally masculine, lone-genius, top-down view of creation. I’m not saying this is non-hierarchical – my name’s on it, I’m choosing which performances land on the record – but ideas aren’t generated in a vacuum. The idea of one person being the entirety of something is just really limited.” The album’s cover is an indistinguishable face-meld of almost all the collaborators involved.
On Proto it’s near-impossible to tell if the voices are human or non-human, singular or plural. One sparkling song called Fear, Uncertainty and Doubt sounds like a constellation of disjointed, disembodied voices, but this time it’s all Herndon – well, kind of. “I’m singing through a system I’ve made. I can morph between human and animal and digital. I can sing through plants.” She patiently explains that this doesn’t require shrubbery in the studio, rather a way to manipulate field recordings in real time, using her voice.
Thanks to her advocacy around the unifying potentials of tech, Herndon has a reputation as a techno-utopian optimist that she feels is undeserved. “It makes me sound like a Silicon Valley shill! I think it’s pretty clear that we have a critical approach to technology.”
She gives an example of the potential problems technology can cause: “We haven’t yet figured out how to deal with intellectual property [in music] and AI is like if a sample could sprout legs and run. It is recording technology 2.0, and we don’t have an ethical framework.”
Forget problematic hologram tours – neural-net voice models could soon make it possible for musical heroes to record wholly new songs from beyond the grave. Holly quotes Miles Davis’s fear of “artistic necrophilia”, meant in the sense that every generation should redefine sound for themselves. “Otherwise we’ll get this recursive feedback loop,” she says, “where we can’t imagine a future that’s different because we’re always regurgitating the past.”
Instead, Holly’s vision of the future is to make the human visible within the machine. On Swim, the last song completed for the album, the human and non-human members of the ensemble are at their most seamlessly, serenely unified. “They really occupy the same space,” she beams. It’s the pinnacle of years of research, and has already transformed their expectations for further projects. “Working with AI has made me appreciate the human body; we’re such amazing sensors,” Holly laughs. “Our eyes and ears and all this stuff you can’t encapsulate in a media file … it really makes you appreciate your own meat sack.”
