Ai, Music, and the (Missing) Human Touch

There’s a strange sense of unease creeping into the music world lately, or I’m hoping there is, given the noise around all this Ai gubbins being talked about. A slight robotic hum beneath the melody that isn’t Kraftwerk’s doing, a feeling that something’s not quite right. Maybe you’ve felt it too. It’s that moment when you hear a track (usually background music at the moment), everything’s in place—the beat is right, the chord progression is nice enough, the voice is pitch-perfect, but somehow, it feels a little… soulless. You can’t quite put your finger on it, but it’s there. And just perhaps, what you’re hearing wasn’t written by a person at all. Perhaps not, perhaps it’s just formulaic crap. For the record, I don’t mind a bit of formulaic crap once in a while. I’m writing this after watching something on Auntie Beeb about Artificial Intelligence creating utterly believable movies from scratch – though humans put all the info in their (prompts). 

We’re living in the age of Artificial Intelligence (Ai). What was once the stuff of dystopian sci-fi is now quietly (or not so quietly in the case of music) embedded in our everyday lives. From autocorrect on your phone to algorithms that know what you want to listen to before you do (I’ve ranted about this before), Ai is everywhere. And now, it’s making music.

Ai isn’t just helping compose; it’s creating tunes. Companies like OpenAI, Google, and even mainstream platforms like Spotify and TikTok are by all accounts investing heavily in systems that can generate original music from scratch – melodies, lyrics, vocals, the lot. Entire symphonies have been “composed” by neural networks trained (that’s how the “things” learn) on the likes of Bach, Miles Davis, or Radiohead – I believe for the sake of humanity that they aren’t trained on Coldplay. There are Ai-powered tools that can write a pop song in the style of Taylor Swift in seconds, right down to the heartbreak and catchy chorus.

Now, I’m not about to be the old man shouting at clouds (again) and declare all this the end of music as we know it. There’s something quite fascinating about what technology can do. There’s no doubt Ai has its place. It can assist musicians with writer’s block, generate backing tracks for bedroom producers, or help remix an old tune in a completely new way. Some people are making genuinely creative and interesting music with Ai as a tool, not unlike how synthesisers or samplers once revolutionised sound  – and were suitably hated at the time by some. I recall someone, perhaps on the box, saying “You just press a button and the synth does it all for you!” – they didn’t.And I’m all for innovation. HiFi has always embraced the cutting edge of technology to a greater or lesser extent, depending on your degrees of ludditery.

But here’s where I start to feel the robotic rub. When Ai begins to replace, not assist, creativity, we lose something fundamental. The crack in the voice, the struggle behind a lyric, the imperfection that makes a performance human – all that gets ironed out. It’s all a bit too perfect. And for me, that’s where the soul starts to slip away. Imagine a Neil Young guitar solo that I’ve always found to be right on the edge of falling apart being done perfectly – it just wouldn’t be the same, and that’s what makes it genius.

I think what bothers me most is the creeping idea that music could become just another output of data processing. Feed the machine a few million hit songs and let it spit out the next big summer anthem or Xmas number 1.  I guess it might well chart and It might go viral. But will it mean anything to humanity – to the listener and to the artist?

One argument often thrown around is that people can’t tell the difference. I reckon in blind tests many listeners genuinely wouldn’t know whether they’re hearing a human-written song or one generated by an Ai. Fair enough. We’ve all heard ultra-polished pop that sounds as though it was written by committee, and, increasingly, it probably was – look at a Chemical Brothers album and look at the personal and writers credits. But here’s my counterpoint: just because we can’t always consciously identify the difference, doesn’t mean we don’t feel it.

The thing is, music isn’t just about sound. It’s about connection. Emotion. Context. The story behind the song. I bang on about this a lot! That old Elvis tune recorded in shed with a single microphone has a gravity to it, a sense of rightness. You can hear the pain, the cigarette smoke, the half-empty bottle in the corner of the room, the chopped up lines. That can’t be manufactured. It’s human. It’s real! 

I believe we’re heading towards a breaking point. Right now, there’s a novelty to Ai music. It’s clever, uncanny, and impressive. But I suspect, maybe hope is the better word, that we will eventually start to push back a bit. Humans have a habit of rebelling against artificiality. Look at the resurgence of vinyl, of analogue synthesisers (despite what I said earlier), of acoustic sets. For every mega-produced, Autotuned pop song, there’s someone in a pub playing their heart out to an audience of ten people and a dog (cat’s don’t get involved in this kind of thing), and that audience is feeling everything – the emotion, the mistakes…the humanity!

I’ve always believed that music is about humanity reaching out across time and space and saying, “This is what it feels like to be me.” It’s flawed. It’s messy. It’s beautiful. And I don’t think a machine, no matter how sophisticated, can truly replicate that. It might be able to mimic, but it can’t truly replicate it.

Now, don’t get me wrong, I’m not suggesting we ban Ai or go back to recording everything on wax cylinders. That ship has well and truly sailed. But I am saying we need to be careful about what we lose in the name of progress. As Ai becomes more integrated into the music industry, we risk mistaking quantity for quality. Just because we can make a thousand songs a day doesn’t mean we should. The same goes for my other passion, photography.

There’s also an ethical thing here, too. Who owns AI-generated music? The programmer writing the prompts? The Ai? The data it was trained on – thousands of tracks written by actual artists, living or dead? Where does inspiration end and thievery begin? And if anyone can create a convincing song with the click of a button, what happens to musicians trying to make a living from their craft? We lose humanity. We lose the human touch. We lose connection with our fellow human beings. 

In the end, I think we’ll see a divide. Mass-produced Ai music will become background noise – the sonic equivalent of fast food, though look how popular that crap is. Easy, accessible, and everywhere. It’ll fill shopping centres, social media clips, lifts (elevators for our American readers) and algorithmically generated playlists (another thing Ive ranted about in the past). But just like food, there will be people who crave something more nourishing. Something real. Something human! 

And that’s where the real-life human artist comes in.

There will always be a place for music made by people…for people. Music that carries intention, vulnerability, and presence. Music that wasn’t just generated, but felt. I believe we’ll start to value human creativity more, not less, as Ai becomes more prevalent. There will be a kind of renaissance, a renewed appreciation for the flaws and fingerprints of the human touch. Try to imagine Fleetwood Mac’s Rumours (yeh, I know I’m an old duffer) without the very public break-ups and the songs that came out of that situation.

So maybe that unease I mentioned at the start isn’t such a bad thing. Maybe it’s a kind of internal warning light (which I’m sure many will ignore just like we ignore a warning light on the dashboard of our cars), a reminder that, deep down, we do perhaps know the difference. Even if we can’t always articulate it when we subconsciously recognise it. Even if the machines get better and better. 

Because at the end of the day, music isn’t just something we listen to, it’s something we share with our fellow human beings. And no matter how clever the code, no Ai can know what it means to be heartbroken, in love, lost, or found. Not really. Not with any kind of human “feeling”. It’s artifice. It’s false. But it’s out there and perhaps it’s here to stay. 

Music is a human story to tell. And long may we tell it. Warts and all! 

Stuart Smith

What do you think about the rise of Ai-generated music? Join the conversation over on our socials.

Read more Sunday Morning Thoughts