The music industry frets about AI music, but is it just a new form of Muzak?

The music industry frets about AI music, but is it just a new form of Muzak?

The following op/ed comes from Eamonn Forde (pictured inset), a long-time music industry journalist, and the author of The Final Days of EMI: Selling the Pig. UK-based Forde’s new book, Leaving The Building: The Lucrative Afterlife of Music Estates, is out now via Omnibus Press. 

There is a fabulous, possibly apocryphal, quote attributed to Nick Cave.

“I’m forever near a stereo saying, ‘What the fuck is this garbage?’ And the answer is always the Red Hot Chili Peppers.”

Which leads us to his latest fabulous set of quotes. In the January edition of his Red Hand Files newsletter, he responds to a letter from a fan, Mark from Christchurch in New Zealand, who tasked the ChatGPT bot writing some lyrics “in the style of” Nick Cave.

Cave wearily replies that he has been sent similar things since November when ChatGPT launched. He calls the Cave-esque lyrics “replication as travesty” and then really rolls his sleeves up.

“It could perhaps in time create a song that is, on the surface, indistinguishable from an original, but it will always be a replication, a kind of burlesque,” he says, arguing that “algorithms don’t feel and [d]ata doesn’t suffer” like true artists do in order to write their lyrics, calling what he does “a blood and guts business” that a machine can never come close to.

“What makes a great song great is not its close resemblance to a recognizable work,” he continues. “Writing a good song is not mimicry, or replication, or pastiche, it is the opposite. It is an act of self-murder that destroys all one has strived to produce in the past […] [T]his song is bullshit, a grotesque mockery of what it is to be human.”

There is a lot more and all of it is stirring and beautifully written, as one would expect from someone like Cave.

He was, however, talking as a lyricist reacting to AI-generated lyrics. If the best lyricists aspire to be poets, to always reach for something beyond, then AI-generated lyrics will always be found desperately lacking.

Cave is probably as withering and damning about AI-generated music, seeing it as a toothless, bloodless, pointless simulacrum.

But – and without wanting to provoke a Red Hand Files newsletter aimed at my head – is there still a certain place for a certain type of AI-generated music?

To answer that, we will have to separate out different kinds of music. Cave is coming at the debate, as he absolutely should, from the perspective of the profound power of music: music as a means of emotional explanation and human connection. His starting point is music as the apotheosis of art, but a) lots of music will fall pathetically short of such lofty expectations and b) not all music has to even try to achieve this kind of transcendence.

Indeed, following Cave’s artistic hierarchy argument, it overlooks the fact that a tremendous amount of human-written lyrics (not his) are absolute doggerel or insipid/mawkish nonsense. But that’s OK. And maybe, just maybe, AI could help improve the output of some of the shockingly bad lyricists out there who have publishing deals. (I will not name names, but there are thousands of them out there. You hear them every day.)

So AI could become a type of shuttering for terrible lyricists, holding up the feeble structure of the songs they have somehow thrown together without proper architectural rigour. AI might even become a total replacement for a pitiful lyricist whose work is beyond all help and hope. Is either result necessarily a bad one?

Setting lyrics aside, there is an argument that AI music composition does have its place. A small place, but a place nonetheless.

Doubtless composers will argue this is driving them out of work, but that depends on what the music is required to do. It’s not quite the MU trying to ban the synthesiser in the 1980s, but music can sometimes exist as a sonic space-holder that has no great artistic aspirations. It’s nothing more than the sound it makes.

We perhaps need to think of music as existing in two very different configurations.

1) Functional/utilitarian music

There to serve a distinct role where music is in the background and which only works when it is in the background. This is music that has no use, power or beauty outside of that specific use case. The music is there only to fill the silence. It is very distinct from library music which has a creative form of its own and can adapt to different use cases.

2) Artistic/aesthetic music

This is music written to connect with humans and to distil human emotion, capturing and expressing something that exists beyond words. At its apogee, this is the “blood and guts business” that Cave is actively engaged in. Sometimes it is gristle, but it is at least grasping for something more.

These two types of music could, as long as we understand the distinctions and how and where they can or should be deployed, co-exist.

The level of creativity/genius involved in the latter will never be replaced by machines. But the use contexts of the former are very different from the aesthetic underpinnings of the latter.

It is really a case of the split between music, on the one hand, as hyper-industrialised and, on the other hand, music as highly artisanal.

This, of course, should not give AI-generated music carte blanche to infiltrate DSPs as part of a terribly topical game of fraud.

Songwriters, who already feel they are being fleeced, belittled and erased in the digital age, might argue that AI music is not just taking food off their table, it’s taking the crumbs. And setting fire to the table. It’s the thin end of a miserable wedge for them.

But AI is not replacing them. It is (or it should) only be used to put up temporary aural wallpaper that is not dependent on skill, craft or art. No songwriter worth their salt would want to create music devoid of these attributes. This is music that is less “blood and guts” and more “bland and ruts”.

AI music is really just a new form, arguably even a lesser form, of Muzak.

Just as Muzak was a homonym, we therefore need a new name for this utilitarian sound to separate it out from the music that wants to stir or soothe your soul. For the sake of argument, let’s call it MusAIc.

Or, in a nod to Brian Eno and his experiments on Ambient 1: Music For Airports that he wanted to be “as ignorable as it is interesting”, we could name it Proxy MusAIc.

 Music Business Worldwide