What’s new in AI and Music? A whole lot!…
What’s new in AI and music? It’s 2023, and it feels like there’s now a new ChatGPT version every week. And it’s not just Twitter that’s aflutter over these powerful new tools; my family group chat had an informed debate about these and other robots just yesterday! AI is here to stay, and musicians and producers are taking notice. In today’s Part 1 post, we explore some music AI tools available, and ask the question: how does it work, and how can I try it?
Clearly lots is going on at the intersection of music + AI. We’ve all heard that Drake/Weeknd demo-sounding track by Ghostwriter, and Grimes stepped into the ring the other day to announce that she is IN for AI voice modeling! Debates around copyright, sampling, and likeness laws abound. If all of this is news to you, check the TL;DR from Water and Music to get caught up.
So how does it work?
First off, disclaimer: let’s be clear that you should not be releasing music containing AI voice models without the artist’s consent, and i-wave does NOT get down with THAT! Got it? Good.
Now, as an artist and producer myself, my first question after hearing that Ghostwriter track was: how did they do it, and how can I try it? Are these tools free and readily available for music makers to run amuck?
Well, yes and no. Like all things AI, things are evolving QUICKLY, so here’s where I landed in my research: if you’re willing to do a little digging, you can find a website to make you sound like Drake or Kanye pretty easily. Beyond that, there’s a little more to it, but I’m sure it will be a quick & easy process (maybe a paid one) in no time as companies seem to have officially entered the ring in their race to be the first & best AI voice model generators for consumers.
So how do I make myself sound like Drake rapping “Princess Diana”?
There’s the manual way to create AI voice models, which is not user-friendly or straightforward; Reddit is probably a good place to get started if you’re interested in going the nuts & bolts route.
However, companies are already emerging to make it easier for anyone with a web browser to do “WAP” in Eminem’s voice (OG Slim Shady or modern day Marshall? Take your pick!)! The one that caught my eye this week was Uberduck, who also jumped on the Grimes announcement hype to create a song contest with a $7k reward. Nice marketing move — it definitely got me messing around on the website to see what else I could do (using their AI voice models is currently just a free account creation away).
Record your voice directly on their website, or upload a vocal file; click Convert and within seconds, you too can sound like Champagne Papi. Check out WTF KYLE’s breakdown of Ghostwriter’s probable methods to see what that could look like.
Stay tuned for Part 2 next week where I’ll explore other music AI tools beyond voice modeling, and maybe share a few of my own AI creations. 😈
xo C