Look, I’m three bourbons deep and trying to make sense of this new research about AI and music. Some professor named Jeffs is telling us not to worry, that artificial intelligence won’t destroy music. Real comforting stuff when you’re staring down the barrel of another hangover and wondering if robots will be writing the next summer hit.
Here’s the thing about music that these researchers sometimes miss while they’re busy coding their next algorithm: it’s meant to be messy. It’s supposed to have rough edges. But try telling that to the venture capitalists throwing money at AI music startups faster than I throw back shots on a Tuesday night.
The good news? Live music isn’t going anywhere. You can’t replicate the experience of being packed into a sweaty club at 2 AM, the bass thumping through your chest while some guitarist who learned their craft in a garage somewhere makes mistakes that somehow sound better than perfection. No amount of neural networks can capture that magic.
But here’s where it gets ugly: all those side gigs musicians use to pay rent? They’re vanishing faster than my paycheck at happy hour. Background music, commercial jingles, game soundtracks - AI’s coming for all of it. And the real gut punch? It’s already happening.
The researcher mentions a-ha’s “Take On Me” and how its grammatically incorrect English somehow made it better. He’s got a point there. If some AI language model had “fixed” those lyrics, we’d have lost something quintessentially human. Sometimes the whiskey-soaked poetry of bad grammar hits harder than perfectly structured verses.
What really gets me is this whole prompt engineering thing. These tech evangelists are saying tomorrow’s musicians won’t need to learn instruments - they’ll just need to write good AI prompts. That’s like saying you don’t need to know how to drink to be a bartender, you just need to know how to read recipes. The art is in the doing, in the failing, in the calloused fingers and the countless hours of practice.
Sure, AI might democratize some aspects of music production. It might help independent artists handle their tour logistics or clean up their podcast audio. But there’s a darker side to this democratization: when everyone has access to perfect production, everything starts sounding the same. It’s like when every bar starts serving the exact same IPA - technically good, but boring as hell.
The heart of this whole debate isn’t really about whether AI can make music - it’s about what we lose when we remove the human element. Those imperfect vocals, those slightly off-time drums, that guitar string that’s just a little out of tune - that’s where the soul lives. That’s what makes you feel something when you hear a song at 3 AM in an empty bar.
The researcher’s optimistic, and maybe he’s right. Maybe AI won’t kill music. But it might just smooth out all the rough edges until everything sounds like it was made by committee in a sterile lab somewhere. And in my book, that’s almost worse.
Remember: good music, like good whiskey, needs character. And character comes from imperfection.
Source: ‘Artificial intelligence is not a threat to music,’ says researcher