Synthetic voice startup Elevenlabs presents six classic stories in multiple languages, all read by a believable AI voice.
They include classics like “Winnie the Pooh” and “The Picture of Dorian Gray.” The company’s “big dream” is to make every public book available in any language, with any voice, in high quality, it writes.
The entire process took just one day, according to Elevenlabs. You can listen to “The Picture of Dorian Gray” in the video below.
The quality of the AI voice is so high that it could spark further debate about replacing voice actors with artificial intelligence, a topic that has arisen in the wake of the Hollywood protests and also in the gaming industry. Game studio Ninja Theory, for example, had to defend itself after criticism that it was replacing human voices with AI. It does use AI voices, but only as placeholders in game development.
Microsoft and Project Gutenberg recently unveiled a project to automatically produce more than 5,000 audiobooks with AI voices. Researchers developed algorithms to distinguish the main text of an e-book from unimportant elements such as footnotes and page numbers, and to identify actors, dialogue, and characters so they could narrate with different roles and emotions.
Elevenlabs unveils new multilingual model
Eleven Labs recently introduced its new model, Eleven Multilingual v2, which supports authentic-sounding voices in 28 languages and can automatically convert text to speech.
The release of Eleven Multilingual v2 also marks the end of the beta phase for Eleven Labs. The startup now plans to integrate a feature that allows users to share their voices on the platform to build a global voice community.
Eleven Labs claims to have more than one million users worldwide and counts several media companies, game developers, publishers and independent authors among its customers. More AI-narrated free audiobooks from ElevenLabs are available on Twitter or YouTube.