This is the problem with the podcast: there are too many.
More than 4 million, at least, according to the Podcast Index database. In the last three days alone, nearly 113,000 episodes have been posted online, an avalanche of audio content so voluminous that listeners they will never run out of options. You could spend the rest of your life with the existing catalog of the genre true crime in Apple Podcasts or Spotify sports talk shows and end up dying of old age while the episodes survive you.
However, in the midst of the current generative artificial intelligence gold rush, opportunistic entrepreneurs are looking to break into even the most saturated markets. A wave of startups, including ElevenLabs, WondercraftAI, and Podcastle, have introduced easy-to-use tools to generate AI voices in minutes. So, just like that, the podcast of artificial intelligence are already here, it does not matter that we have not asked for them.
In these early days, no one keeps track of how many listeners this strange new genre of music has. podcast. Major platforms like Apple Podcasts and Spotify don’t have separate lists for robot hosts. However, there are some of those podcast created with AI that have clearly found their audience, at least in their first batch of episodes.
“This is good enough for me”
The first AI-generated podcast to take off cheated a bit: it used the cloned voice of the human presenter of podcast most popular in the world: Joe Rogan. Joe Rogan AI Experience is a series of simulations of Rogan conversing with (equally fake) guests like Sam Altman, the CEO of OpenAI, and former President Donald Trump. Shortly after the first episode was posted, the real Rogan tweeted a link to it: “This is going to get really slippery guys,” he wrote.
On YouTube, imitation racked up more than half a million visits. Some listeners didn’t even care that it was artificial intelligence. “this is good enough for me. Good material” wrote one.
Joe Rogan AI Experience was created by a Rogan fan named Hugo. He refused to give WIRED his full name because he doesn’t want to be professionally associated with the project. He has a Patreon account for people to support the production of the show, and he recently turned on monetization on YouTube, but he doesn’t expect to get any real income from it, especially since he’s aware that You do not have consent to use Rogan’s voice or imageand that the platforms of podcasting they may end up banning this type of impersonation.
Hugo created the series because he wanted to show what AI voice tools can do. Although he carefully edits the episodes to make them smooth for listeners (they can take days or weeks to get right), he doesn’t think the conversations themselves are particularly gripping, even if they are reasonably accurate imitations. “Aside from hearing the podcast Because of its technological advances, it doesn’t make sense,” Hugo confesses. “It’s a waste of time.”
A waste of time
It’s unclear if the audience will stick around, or if they just wanted to see something unusual and new; Hugo has published four episodes, and each subsequent installment has drawn a smaller audience than the previous one.
WIRED spoke with other creators of AI-powered podcasts who echoed Hugo’s opinion. They enjoyed playing with the technology, but consider the end results to be a byproduct of experimentation. Israeli sound engineer Lior Sol, for example, created a podcast called Myself, I Am and That (Myself, I am and That) with the tools of ElevenLabs. She made a clone of her voice and then a clone of that clone in conversation: “I definitely have fun with it,” he says. But that doesn’t mean it’s going after big audiences. Right now, his listeners number in the dozens. His friends like it, he likes it: it’s an artistic project and an opportunity to tinker with new technologiesIt is not an attempt to do something commercial.
