As AI wows us with how good it is at mimicking us, it’s starting to redefine creative processes that we have shaped over centuries. Until we completely do an overhaul of artists’ rights and intellectual property safety laws, it looks like artists will suffer.
This time it’s voice artists. Last month, OpenAI took down a voice used by its ChatGPT that sounded uncannily like Hollywood actress Scarlet Johansson. The bot giant halted using the voice after the actress threatened to sue the company. Their explanation about how they got the voice in the first place is that they used another actress for it.
It will take time for us to accept AI as a part of creative processes as a society. Meanwhile, big tech ensures that AI races on, mimicking human creativity within seconds. Where is the time to redefine this new reality and protect the artist?
Where are these AI companies getting these voices from? Isn’t the source important?
“In a time when we are all grappling with deepfakes and the protection of our own likeness, our own work, our own identities, I believe these are questions that deserve absolute clarity,” NPR quoted Johansson.
In a time when we are all grappling with deepfakes and the protection of our own likeness, our own work, our own identities, I believe these are questions that deserve absolute clarity
Scarlet Johansson
She has a point. There are other voice artists too who went through the shock of hearing their voices from sources they hadn’t lent them to. These incidents show the lack of legal safeguards when it comes to using creative work that aids AI tools.
Why is Voice Significant?
It might seem hard to imagine how AI that mimics human voice behavior will impact human society because we tend to take it for granted but as tech giants vie to provide much more lifelike voice assistants, we could soon find out.
As Visar Berisha, an Arizona State University professor who studies AI speech technology, told NPR, “Communication by voice is really intimate, really impactful. It allows the AI to express subtleties, things that are perceived as sincere, urgent, joy, concern. And all of these serve to foster a deeper connection between the user and the machine. You can see how these interactions can potentially become addictive.”
The movie ‘Her’ has naturally become part of this discussion. In the movie, when the protagonist falls for the flirtatious AI operating system, sporting the voice of Scarlet Johansson incidentally, it’s just the voice that tugs at his heartstrings, and we understand why it happens. So, undeniably, there is something in the voice.
That voice has an impact on human society is evident from how attractive it is as an investment for businesses. American entrepreneur and investor Naval Ravikant recently launched Airchat, a voice-centric social media app. AI customer reps are filling the market.
AI companies, like Hume AI, are starting to introduce EVI, or Empathic Voice Interface, where the AI changes its responses based on the user’s tone.
AI & the Music Industry
AI is changing more in the voice area, namely the music industry. It isn’t a surprise to find singers crooning away while AI adds effects to their real voices.
Already AI is changing one of the biggest and most popular voice entertainment arenas, hip-hop and rap music. An ongoing diss-track rap battle between Kendrick Lamar and Drake has brought out the impact that technology has on hip-hop with curated releases on social platforms and AI-generated songs.
It’s not just voice. Creative people are steadily losing to AI. Recently the psychedelic rock band Pink Floyd faced flak from fans when they chose an AI-generated video as the winner of their The Dark Side of The Moon 50th anniversary competition over all the human-made entries.
We all laughed in 2022 when Midjourney first came out and said, ‘Oh, that’s cute.’ Now people are losing their jobs to Midjourney
Artist in Michigan told New York Times
As a movie concept artist in Michigan told New York Times, “We all laughed in 2022 when Midjourney first came out and said, ‘Oh, that’s cute.’ Now people are losing their jobs to Midjourney.”
What will happen when the taboo of AI drops, which seems inevitable?
As an article in Wired says, “The manner that technology has already changed rap beef begs the question of where it will go next. One day, the taboo around AI in hip-hop will disappear, and entire battles will be orchestrated by LLM rappers trained on the raps of individual artists.
Quants will develop metrics for who the winners are. If we are offended by a lyric about a member of our family, we’ll blame the machines. It may sound like the stuff of science-fiction, but the gap between this future reality and 2024 might be smaller (in time and manner) than the gap between Canibus vs. LL Cool J (1998) and Kendrick vs. Drake.”
AI is Sneaking in but Here to Stay
Last month, Gannett, an American media company that owns hundreds of newspapers, added AI-generated summaries to its articles. Five of this year’s Pulitzer winners used AI in their research.
It feels a little like leaving the back door open for AI to sneak in, but we can be certain it will eventually reach the porch.
It seems inevitable that these kinds of incidents will rewrite existing copyright laws. Recently, there have been many copyright infringement lawsuits against companies that are making use of tens of thousands of copyrighted books to train generative AI systems. One of them has called it “systematic theft on a mass scale,” not too far from the truth.
The Good Side of Voice AI
It’s not all evil though. AI voice-cloning technology can also be helpful to some people. People who have lost their speech are using AI voice cloning to get it back. Trained on a 15-second time capsule of her teenage voice, a 21-year-old woman in the US enjoys a synthetic but real-sounding AI voice. Country singer Randy Travis has been able to release his latest song with the help of AI, after he lost his speech to a stroke in 2013.
We can’t ignore the glory of such incidents either.
Co-creative Partners with AI
Researchers are saying that we should lean towards co-creativity, i.e. where humans and machines interact to co-create, or ‘human-centered AI’ and ‘hybrid intelligence.’ They argue this will “ensure both high degrees of automatization through AI and human control” and support “a relationship that optimally empowers each other.”
Still, is it that easy? Letting AI into our creative processes to ‘co-create’ articles, stories, images, videos, and music might make things more convenient but the feeling is uneasy. Purists could forever view it as cheating.
It will take time for us to accept AI as a part of creative processes as a society. Meanwhile, big tech ensures that AI races on, mimicking human creativity within seconds. Where is the time to redefine this new reality and protect the artist?
On the panicky side, as our dependency on AI increases, will we lose our ability to create independently? Imagine an analog world where we grope about, as if in the dark, to come up with basic sentences or paint a simple picture. Already usage of AI is demanding that we give it due acknowledgement in creative works.
Navanwita Bora Sachdev, Editor, The Tech Panda