According to the Podcast Index, nearly four in ten new podcasts are now churned out by artificial intelligence. However, industry analysts note that the synthetic audio remains entirely indistinguishable from standard human programming.
The Podcast Index revealed this week that 39 percent of all new audio shows launched over the past nine days were generated entirely by AI. The synthetic programs, categorized internally as podslop, consist of language models reading Wikipedia articles, debating the merits of daily cold plunges, and seamlessly transitioning into read-aloud advertisements for online therapy platforms.
Despite panic among human creators, audio executives have welcomed the automation. Major streaming platforms noted that the AI models require significantly less overhead than traditional celebrities while perfectly replicating the medium's core offering of two uncredentialed men speaking confidently into condenser microphones for three undisturbed hours.
The latest algorithms are remarkable because they know exactly when to chuckle at their own inside jokes before abruptly pivoting to a BetterHelp promo code.
Industry insiders anticipate that by next quarter, large language models will be advanced enough to launch an exclusive Patreon tier and complain that their controversial views on seed oils are being silenced by the mainstream media.