As generative AI expands, it’s starting to reach the ads we hear on podcasts and radio.
Startup Adthos this week launched a platform that uses AI to generate scripts for audio ads — and even adds voiceovers, sound effects and music. Customers can optionally enter and edit the ad using a browser-based multi-track editor, a “fully-built” ad that can be played by integrating with radio automation systems and ad servers.
Adtos CEO Raoul Wedel set the stage as a way Advertisers will “efficiently self-serve” and “immediately” produce high volume audio ads – and “support long-tail revenue streams” for publishers by creating self-service ad buying platforms.
“The real game changer is that a small advertiser can run a good quality ad in a minute or two,” Wedel said, pointing to the potential for localization and personalization.
But the platform — like many that embrace generative AI technology — raises ethical questions.
The Adhos platform is built on a mix of in-house technology and third-party APIs. To generate ad scripts, Adthos uses OpenAI’s recently released GPT-4 text processing model. And on the voice side, Addos has developed a library of hundreds of synthetic voices, including – according to Wedel – “Emmy Award-winning” voice actors and talents.
Voice actors are fully aware of the platform and pay a premium or license fee to use their voice based on their preferences, Wedel said. “Adtos has already paid actors hundreds of thousands in royalties,” he added.
But what’s unclear is whether all of these voice actors chose to be included in the Addos library in the first place.
Recently, Vice reported that actors are being asked to sign away the rights to their voices, and clients are being asked to use AI to generate synthetic versions of them that could eventually replace them. The contract language tends to be ambiguous, with voice actors saying some actors are being told they can’t be hired without agreeing to clauses that give them the right to use their voices for synthetic training.
Image Credits: Athos
Wedel did not disclose the terms of the Adhos arrangements. But he reaffirmed that the voice actors on the platform had signed consent and licensing agreements.
“We use real-life voice actors to train synthetic voices. We train and select them based on our experience in marketing and ad creative,” Wedel said.
Wedel added that actors are given some control over how their voices are used on stage, such as banning them from being used in certain advertising categories, such as religious or political ads.
Asked what steps Adtos will take to curate content generated by the platform, Wedel said it will step away from reviewing ads before they are finalized and distributed to clients and publishers.
Generative AI, especially AI that generates text, has a tendency to go off the rails. For example, OpenAI’s AI-powered chatbot, ChatGPT, has been found to reinforce facts and reinforce racial and gender stereotypes. If an Adthos client fully automates the ad generation process, they run the risk of producing toxic, offensive, and ultimately harmful ads—an undesirable outcome for any brand.
Wedel announced that OpenAI performs some filtering at the API level. But – perhaps realizing that this alone isn’t enough – Adthos says it plans to increase content moderation “as the platform evolves”, including words or terms that can’t be used in advertising.
“It’s up to the publisher to make sure that humanity is overlooked,” Wedel said. But we have certain requirements, such as disclosing the AI nature of the ads and using voice capabilities with their consent.
Either way, Adtos says it has a number of paying clients, including Dutch media conglomerate Talpa, and advertisers and agencies such as Group M, Dentsu and Spartradar.
I asked Wedel point blank if there was any concern that the Adhos platform could cost marketing and advertising agencies and companies their work. OpenAI CEO Sam Altman was quoted as saying it was a “deliberate decision” to start conversations in the community about what’s to come regarding AI.
“It will certainly cause jobs to disappear. But the people who develop and embrace AI will thrive and become more productive,” Wedel said. “Those who don’t will lose their jobs.”
