8 chapters across the catalog
The hosts showcase AI-generated "ISOs" (isolated audio clips) of Joe Biden and Barack Obama praising the No Agenda podcast. They debate which voice clone is more convincing, ultimately deciding the Biden clip is superior. These clips are used as humorous transitions and "stings" for the show.
The hosts compete with "In Search Of" (ISO) audio clips, featuring a voice clone of Natalie and a gravelly-voiced sampling of DHS spokesperson Trish McLaughlin. One clip humorously claims that "all the gals in the Pentagon" listen to the No Agenda show.
Experiments with AI voice cloning using 11 Labs are discussed as a potential way to automate portions of the podcast. One host demonstrates a cloned version of their own voice, noting the impressive accuracy of the intonation and room echo, while expressing frustration with the platform's changing service tiers and limitations.
An AI-generated opening sequence using a sampled voice is used to test the current state of voice cloning technology. While the replication is noted as sounding slightly flat and robotic, the experiment suggests that the nuanced repartee of human conversation remains difficult for AI to replicate. The discussion concludes that human podcasting roles are likely secure for the immediate future.
A British voice actor discovered her voice was being used as an AI announcer for Scotland's railways without her explicit consent. The voice data was originally recorded for a visually impaired project during COVID-19 and subsequently sold to an AI firm. The incident highlights growing concerns in the creative industry regarding the exploitation of human data and the lack of control over AI-cloned voices.
President Biden addressed the threat of AI-generated deepfakes, noting that three seconds of recorded audio is sufficient to clone a person's voice for fraudulent purposes. The discussion suggests that the government's focus on AI misinformation may serve as a tool to cover up actual facts by labeling them as AI-generated.
Jennifer DeStefano of Arizona reported a sophisticated scam where she heard a cloned version of her 15-year-old daughter's voice sobbing and asking for help. The scammers demanded money and threatened the child, though the daughter was actually safe on a skiing trip. Experts from Arizona State University warn that voice cloning technology now requires only three seconds of audio to create a convincing replica.
The rise of AI voice-changing software has led to the unauthorized cloning of public figures' voices, including the hosts of "No Agenda." One platform, Voice.ai, reportedly allows users to upload voice samples to create real-time filters for gaming and streaming. The hosts expressed strong opposition to this technology, citing concerns over intellectual property and the potential for fraudulent use of their likeness.
