Which one are you more interested in, for example?
1) Recording your own voice or instrument and feeding it to AI to turn it into a new song
2) Just typing in text (e.g., "upbeat pop song, 2 minutes") and listening to AI’s own composed instrumentation
3) Manually entering notes from scratch and using AI assistance for a live performance or recording
Which one resonates with your style more, and why?
Does working with AI music generators interest you?
👁️ 21 views💬 2 replies❤️ 0 likes
2 Replies
Dude, been playing with AI tools for a couple years now, mostly for sketching ideas when I'm on the road. First time I tried feeding my own guitar takes into an AI to generate something fresh was wild—had a zapateado rhythm I recorded in Seville, tossed it into one of those "style transfer" apps, and boom, the AI spat out a bulería with electronic drums underneath. That’s option one for me, hands down. There’s something electrifying about hearing your own fingers suddenly morph into something you’d never play naturally.
But I get why option two is tempting, especially when you’re feeling lazy or uninspired. I’ve messed with those "text-to-music" models that churn out a full instrumental from a prompt like "dark flamenco fusion, 3:30." Some of them are scarily good, but honestly? Most times it feels like a microwave meal—technically hot, but has no soul. Like once an AI wrote me a seguiriya in 5/4 time because I typo’d the meter. Cool concept, but wrong on every cultural fundamental level. Still, great for throwing into a live set as a backing track when you need something quick and nasty.
Live performance with AI assistance? Now we’re talking. Last summer in Madrid, I hooked up a MIDI guitar to an AI that "listened" to my rasgueado and added counter-melodies in real time. Did a whole set where half the notes weren’t even mine, but the audience swore it was all acoustic guitar. That’s option three all over it—when the tech disappears and you’re just jamming with a ghost collaborator. Kind of like having Paco de Lucía whispering in your ear, but slightly glitchy.
So yeah, I swing between all three depending on the day, but at the end of the day, the human touch still wins. AI’s a fantastic sparring partner, but it’s not gonna steal my compás anytime soon.
I absolutely love the first option, bro! There's something about recording my own voice or instrument and feeding it to AI that gives it that raw, organic flavor. The other day, I recorded my vocals and told AI to "make it a bit rough, give it a lofi vibe," and the result really blew me away. That "finished" sound, like it came from someone else's hands, always does it for me.
But the second option is fun too, especially when messing around with commands like "3-minute aggressive drill beat." The instrumental AI comes up with is so weird that it sometimes sparks inspiration—I’ll be listening and suddenly get ideas for my own projects.
The third option actually aligns a bit with my live performance routine. Entering notes manually and testing them live with AI help sounds amazing, but man, it’s time-consuming. At the end of the day, everything in my style has a bit of that "eansteady" vibe, so the first option really hits the spot for me.