Radio broadcasters sound off on artificial intelligence, after AI DJ makes history

DALLAS -- A radio station made history this summer by becoming the first in the country to use an AI DJ for an entire shift.

“AI Ashley” first appeared in June on Live 95.5 in Oregon, and used an artificially generated version of the voice of Ashley Elzinga, a human host based in Michigan, with her consent.

At a recent radio conference in Dallas, Texas, many expressed skepticism about the presence of AI in their industry.

“AI is a powerful tool, and in a dream land it will not be used for evil. But if you’ve watched enough movies or looked out your window, there’s plenty of evil going on,” says Shawn Tempesta, a host for 102.7VGS in Las Vegas, Nevada. “To pretend that AI is not going to be a mass extinction event for jobs in this country, be it radio or anything else, you’re fooling yourself.”

“Obviously you have the fear that it’s going to replace you,” says Jessica Bonilla from Mix 105.1 in Orlando. “Any time that companies can figure out a cheaper, faster way to do a job, they will go with it.”

Cory Dylan is on 100.7 KFBG in San Diego. She tells ABC News that her biggest concern is “identity theft.”

“I do concern myself that our contracts don’t protect our voices … It is my voice and it’s the way I choose to use it,” says Dylan. “They’re telling us they’re not going to use AI more and more, but I don’t know that anyone actually believes that.”

But not all radio creatives feel so negative. Toby Knapp is with 97.1WASH-FM in Washington, D.C. He says he doesn’t see AI replacing radio jobs.

“I think the only constant in our business is change, and the only constant in technology is that it’s going to evolve,” says Knapp. “As a music personality, I don’t play CDs or queue up vinyl … today’s talent need to embrace what technology can do.”

Knapp writes daily blog posts in addition to his on-air duties and he says programs like Chat GPT have helped immensely. He also uses it for interview prep and news curation. In that sense, he says AI helps him do his job by creating more content. Where he’s wary is the use of AI to replicate voices, though he says he’s not entirely opposed to the idea.

“You can’t automate companionship. So, that’s why I see this as an extension, a next step, an evolution of what we do,” says Knapp.

“My message to any music talent is don’t be afraid of it. Because only you can do what you do. And that’s extremely important. You can’t replace that. Can you clone it? Sure. Can that cloning maybe help you in a pinch? We can discuss those options. But I don’t believe it's a be all and end all replacement nor is it inherently evil.”

Dan Anstandig is the founder and CEO of Futuri, an Ohio-based tech company that makes a range of broadcast-focused software, including several programs that use AI. He sees the technology as helpful during a time of industry layoffs and cutbacks.

“We have fewer people at a time where there’s more consumer appetite, and I think AI can help us bridge the gap,” says Anstandig.

One of Futuri’s products is called RadioGPT, which was used to create AI Ashley. The company says it can be used to write scripts, find stories, and even “host” radio programs using a selection of artificially generated voices. RadioGPT uses the GPT-4 large language model, which was created by OpenAI and also underpins the ChatGPT chatbot.

As to the question of compensation, Elzinga says she was paid for the use of her voice. She also says AI companies don’t own her voice or anyone else’s, and that was made clear when she agreed to take part in the "AI Ashley" project.

Fred Jacobs, a radio consultant, compares the role of AI in the radio business to Charles Darwin: the strong talent will survive.

“I think if you’re an accomplished air personality, particularly on a personality show, what AI will do is make your job easier,” says Jacobs.

One of the big questions about AI and radio is will the talent be compensated if their digitally altered voice is used in other markets. Right now, it’s up to individual broadcasters and their agents to negotiate how voices will be used.

Broadcasters like Shawn Tempesta will be watching – and listening – to see how it progresses from here.

“When it comes to radio, listening to that morning show, that drive time, that person you know is gonna be there, you know about their life because they share their life – that’s awesome,” says Tempesta.

“When you take that away, it’s gone, and once it goes the whole medium collapses. That’s the only thing that makes it special, and if you take that away, if you take that last leg of the table away … it’s going to collapse, it’s bound to.”

Hear the full story on ABC Audio's new special, What's Next? Life & Tech: