In 2023, thanks to the rapid development of AI adoption, from the launch of ChatGPT to AI DJ, this technology is fundamentally changing conventional practices and impacting consumers, creators, and industry leaders.
Similarly, AI technology is also fundamentally changing the entertainment industry, including podcasts. Perhaps the most controversial part of this revolution is AI voice generation, which has raised legitimate concerns among creators who fear that their voices will be used without permission.
However, podcast creators have a particular opportunity to monetize their voices. New tools allow them to generate ads, introductions read by the host, and even entire episodes based on text, reducing the time spent in the studio and the need for expensive recording equipment. But AI voice generation is a double-edged sword, bringing both convenience and risks to creators as their voices may be misused by unauthorized users, leading to the spread of misinformation.
New opportunities: AI in audio advertising
Advertisers are already using AI technology to fine-tune podcast ads. In February of this year, iHeartMedia, the largest audio media company in the United States, announced a partnership with audio intelligence platform Sounder to provide new brand safety technology for podcast ad clients, allowing the meaning and context of podcast content to be evaluated and explained before ads are placed, helping brands make brand safety and suitability judgments.
- MPC+AA is the path that Crypto wallets with a billion users must take for Mass Adoption.
- Facing 200 million dollars, 19-year-old Euler hacker hesitates for three weeks due to conscience.
- DeFi leader Compound doubled in a week – what happened behind the scenes?
According to reports, iHeartMedia uses Sounder’s audio data cloud technology to provide this service for more than 750 original podcast programs, including some of the world’s largest podcasts and creators. These programs have more than 392 million downloads and listens per month, more than the second, third, and fourth-ranked podcast copyright holders combined.
In addition, podcast platform Acast has launched a new conversational positioning tool using AI technology that can place ads in the most appropriate position based on the content of the episode, thereby enhancing the effectiveness and brand safety of ads.
If a brand wants to promote its environmental philosophy, the conversational positioning tool can place its ads in the section of the podcast that discusses environmental issues or sustainable development. This can increase the relevance and appeal of the ads while avoiding conflicts with content that does not conform to the brand’s values or sensitive topics.
Some companies are exploring how to use artificial intelligence technology to create complete audio ads. For example, audio advertising technology company Adswizz is developing an AI-based voice generation tool designed to help podcast advertising clients more accurately place ads. These tools are developed by Adswizz’s technology innovation lab, which uses AI and machine learning algorithms to interpret podcast content and target ads based on brand requirements and goals, eliminating the cost of hiring voice actors.
However, ads read by podcast hosts are considered more effective because podcast listeners trust their hosts and are more sensitive to the relevance of ads than the general public. This provides an opportunity for hosts to use AI voice models themselves.
TechCrunch reported that Spotify has been using AI technology in its AI DJ feature to create podcast ads by training an AI model to mimic the voice of Xavier “X” Jernigan, head of Spotify’s culture department and podcast host. Now, Spotify may apply this technology to the advertising field.
Spotify executive and podcast host Bill Simmons revealed that Spotify is testing AI technology that will enable podcasts to generate ads using their own voices in real-time. This allows creators to produce ads that sound like the host is personally reading them in less time, increasing the number of ads and revenue.
In addition, this technology can also help target audiences and translate ads into various languages, provided that the host allows their voice to be copied. This tool may attract more advertisers and creators to join Spotify, which is consistent with Spotify’s recent shift towards serving independent creators.
However, Spotify is not the only platform dedicated to this type of feature. For example, Revoice, an AI voice cloning tool launched by Podcastle, can create a digital voice similar to a real host using AI technology.
The use of Revoice is simple, users only need to record themselves reading 70 different sentences, as well as a legal disclaimer to protect the user’s voice safety and privacy. Then, Podcastle’s AI algorithm will process and generate the user’s digital voice within 24 hours. Users can enter any content on the Podcastle platform and let AI generate high-quality audio files with digital voices.
New risk: spread of misinformation
The voice of the podcast may be one of their most valuable assets. For many creators, their voice is their only identity, especially for those who do not produce podcast videos.
Joe Rogan is a famous podcast host whose show, “The Joe Rogan Experience,” invites guests from all walks of life for in-depth conversations. However, in May of this year, someone used artificial intelligence technology to create a fictional podcast, mimicking conversations between Joe Rogan and some guests who have never been on his show.
The podcast is called “Joe Rogan AI Experience” and was created by a creative director from Australia named Hugo using a text-to-speech platform that can clone any voice and a chatbot-based dialogue generator called ChatGPT. Hugo said that he only did this project for fun and did not intend to threaten the real Joe Rogan or other podcast hosts.
Hugo released the first episode of the podcast a month ago, which simulated a conversation between Joe Rogan and OpenAI founder Sam Altman. The conversation eerily touched on the risk of AI-generated false content, which is exactly what this podcast does.
In addition, Hugo has made several other episodes of the podcast, mimicking conversations between Joe Rogan and guests such as Andrew Tate and Steve Jobs. While these conversations can be silly or boring at times, they can also sound realistic or interesting. Hugo said that he hopes to showcase the development and potential of AI technology through this project and also remind people to be aware of the dangers of false content.
Meanwhile, TikTok has also deleted a viral fake ad made with Joe Rogan’s voice.
In fact, news is the main type of podcast, which means that erroneous content can have serious consequences and lead to the widespread dissemination of false information. AI voice generation tools are developing at lightning speed, but regulations to protect their voices will take time to develop. However, if platforms can implement effective measures to prevent abuse and flag unauthorized works, podcasts can even use AI technology to expand their voices into other forms of entertainment.
For example, podcast hosts can authorize their voices for movies and television. In the “Star Wars” series of films, the voice of Darth Vader has always been voiced by actor James Earl Jones, whose deep and dignified voice has left a deep impression on viewers. However, the 91-year-old Jones stated that he wants to retire from the role and let AI take over.
According to Vanity Fair, Jones has transferred the rights to his past voice work to Disney, allowing Ukrainian startup Respeecher to use artificial intelligence technology to recreate his voice in Disney+’s Obi-Wan Kenobi.
Respeecher uses voice clips to “clone” an actor’s voice, allowing production companies to record new lines without the actor being present. When Jones “mentioned that he wanted to retire the character” of Darth Vader, Skywalker Sound’s supervising sound editor Matthew Wood offered him the option, he told Vanity Fair.
After Jones agreed to let Lucasfilm use AI-generated audio, the production company reportedly commissioned Respeecher to make Darth Vader sound like Jones did 45 years ago, presenting that “dark side” villainous quality in Disney+’s Obi-Wan Kenobi. That’s why you may notice that Darth Vader sounds like he did in early films rather than Jones’ actual voice in 2019’s Star Wars: The Rise of Skywalker.
Unlike artists, actors, and other entertainment creators, the most valuable asset for podcasters is their voice – meaning they have more incentive to explore how to work with AI voice technology than to resist it.