AI Johnny Cash singing ‘Barbie Girl’? I won’t be walking that line – and neither should you

Hello, I’m not Johnny Cash.” This is the uncanny opening of one of the latest AI-generated music “covers” to go viral across social media. In this instance, Cash’s iconic baritone is repurposed to sing Aqua’s 1997 dance-pop hit, “Barbie Girl.”

Wade a little further into the world of AI music and you might also encounter Frank Sinatra crooning “Gangsta’s Paradise” or Elvis swinging his way through “Baby Got Back”.

There’s no denying that these so-called covers have comedy value. And to my inexpert ear, they sound almost indistinguishable from the artists they imitate. Still, as with perhaps all things AI, the broader implications are not so cheery.

The whiplash speed of AI development already feels like the defining feature of 2023. Staying abreast of the technology’s latest advancements is a futile pursuit, while the material impact of AI is already beginning to touch normal people.

It’s an unsettling feeling, watching in real-time as governments scramble to manage something which feels less like tech and more like magic. For musicians, AI is the biggest and baddest in a long line of disruptors.

As it stands, artists are already poorly remunerated by streaming platforms. Spotify even offers artists greater exposure via its “discovery mode” in return for a slashed royalty rate.

Then there’s TikTok. The continuous demand for content has positioned earworm song fragments above, well, actual songs. The result is an increasing quantity of music which (as Mitch Therieau recently put it) exists primarily as “audio furniture for viral videos”.

What’s truly scary is that AI has the scope to overshadow every one of these issues. After all, it’s not just deceased music legends who are being puppeteered by AI – but living, breathing artists.

A particularly pivotal moment came last April when a “Fake Drake” track went viral. Released by an anonymous TikTok user, the song “Heart on My Sleeve” featured AI-generated lyrics with vocals styled to sound like Drake and The Weeknd. The song was eventually pulled at the request of Universal Music Group, but not before racking up 15 million views on TikTok.

It certainly sheds some light on what might be to come. At once opening up the potentially messy world of copyright implications, while also proving that there is an appetite for AI music. Forget emaciated royalty fees – streamers could soon forgo humans altogether, machines being famously less fussy about artistic and financial credit.

It’s fitting enough, then, that the Grammys have recently amended their own rulebook to accommodate these shifting sands. The organisation’s new requirements make clear that “only human creators” can receive award nominations. However, that does nothing to prevent AI from being used on eligible tracks, merely banning songs which are devoid of any human authorship.

It’s probably small-minded to suggest that there is no facet of AI which might be creatively used as part of some broader musical endeavour. What concerns me is the treachery of this particular capitalist slope. Can we really trust big studio execs to stay loyal to not only music artists, but the countless people who make the creation of an album possible?

Humans are messy and capricious. Artists especially so. They need time and space to make music. Worse still, they have wills and opinions. What if they don’t want to keep cruising down whichever algorithmically approved stream promises the greatest financial dividends?

Of course, in breadth, this is a conversation which spans the entirety of the arts sector. The lack of AI regulation is one of many reasons the Writers Guild of America is currently on strike. Visual artists are in an arms race over increasingly sophisticated tools that help them to stay one step ahead of plundering AI systems. The fight between data-hungry AI programmers and artists looking to retain their intellectual property rages on.

What often gets lost in this furore is the very reason art exists in the first place. It’s so much more than some amusement to pass our mortal hours, but a direct response to the conscious experience – the very strangeness of being alive. Something that AI, having no personhood, cannot speak to.

Putting aside the legal and ethical implications of dredging up the voices of long-dead artists, do we really want to tune into songs that have no emotional grounding in actual experience? Or hear voices technologically detached from the silk and grit of real human vocal cords?

Music is a genuine force in our lives, a universal art form which can be found in almost every human culture and time period. Johnny Cash drawling about life in plastic might be an amusing gimmick, but to pick AI over real artists is a line I’m not willing to walk.