I know AI is a divisive thing for a lot of people, especially in the VO community. A lot of folks feel like it’s stealing jobs, and yeah, to an extent, it probably is. Personally, my feelings about it are… complicated.
I’m definitely not a fan of AI music. When I see AI slop on my feed, I just swipe past it or hit “not interested.” I don’t like seeing Hollywood or filmmakers use AI when they could easily make something with real people and real set pieces. And I can’t stand listening to fake podcasts, like that thing Spotify did with NotebookLM for Wrapped that one year.
But I don’t really have an issue with services like ElevenLabs and voice cloning. I’ve listened to audiobooks made with cloned voices, and they sound pretty good. I even cloned my own voice, a professional-grade version, and licensed it out.
If I use my clone to read my own written work, whose job am I stealing? I’ve proven over and over that I can read audiobooks myself. I’ve recorded several. Two seasons of Dark Days of Dorothy Gale, and more recently Noodles Al Dante. Selling the clone for personal use could maybe be seen as job theft, I guess, but I look at it differently.
I’ve sold stock photography through Shutterstock, given it away on Pixabay, and back when I was really into portraiture, I’d often shoot for free because I’ve always felt everyone deserves a good portrait without breaking the bank. Licensing my voice feels the same. People pay a small fee, I get a small percentage, and they can use it however they want, even if their profit ends up way bigger than mine. I’m fine with that imbalance, even if it looks “unfair” on paper, because it gives people an affordable tier. Sure, a big corporation could use my work for cheap, and that’s a bummer. But a small mom-and-pop shop making a flyer can use it too. That’s… not a bummer.
When people argue that a voice clone is just a wrapper on framework of other people’s recordings, or they say “the software can give your clone an English accent, so it’s not really your voice,” I get where that reaction comes from. But to me, the base software is just an engine that understands phonetic rules. Nobody owns the concept of an accent or how language works. Human actors learn accents by listening to people talk, too. The software provides the acoustic framework, but the identity? The timbre? The actual sound? That’s 100% my voice, trained on hours of me performing words that I wrote.
The real issue in the VO community is consent and corporate greed. Scraping people’s voices without permission to replace them. But because I built my own dataset, consented to the clone, and own the asset, I don’t see licensing my own voice as job theft. It’s just another way for me to offer my own work on my own terms.
And of course, some people will disagree, and that’s fine. If you don't disagree, you're welcome to use my voice for your own stuff.
