The article discusses the use of generative AI in transforming a longform journalistic piece into a multimedia story, specifically the story “The Making of Edgar Matobato” which covers the Philippines’ recent history of drug-related killings. To reach a wider audience, the writer sought permission from the original story’s author and used AI tools to create images, illustrations, voice-overs, animations and music.
For the voice-overs, the writer chose Eleven Labs, a company specializing in AI voice generation. By subscribing to the Creator tier, the writer had access to diverse voice options including those with appropriate accents for the story. The chosen voice for the main narrative was “Francesca”, described as a young female voice with a mild Filipino-American accent. The story was divided into chunks before converting them into audio using the Project feature, which is designed for longer form content like podcasts and audiobooks.
The writer faced challenges in getting the synthetic voices to correctly pronounce certain words, especially proper nouns. To overcome this, they listened to the generated voice and adjusted the spellings until the AI could pronounce the words correctly. For example, the spelling of “Samar province” was changed to “Sam-mar” and the surname “Failon” to “F