The Future of Audiobooks: AI Takes the Mic

Audiobooks have become increasingly popular, with platforms like Spotify even creating dedicated spaces for them. But recording an audiobook is a challenging endeavor, even for seasoned voice actors. Enter the world of AI. Researchers from MIT and Microsoft are collaborating with Project Gutenberg, the world’s largest repository of open-license ebooks, to produce 5,000 AI-narrated audiobooks. These include classics like “Pride and Prejudice” and “Alice’s Adventures in Wonderland.”

Mark Hamilton, a lead researcher from MIT, shared, “We wanted to create a massive amount of free audiobooks for the community.” Trained on millions of human speech examples, it can mimic various voices, accents, and even languages. Remarkably, it can produce custom voices from just five seconds of audio.

However, challenges persist. Project Gutenberg ebooks, crafted by volunteers, often have inconsistencies. The ultimate goal? Expand the AI-narrated collection to all 60,000 books on Project Gutenberg and possibly translate them.

Currently, these AI-voiced audiobooks are available for free streaming on platforms like Spotify and Apple Podcasts. The technology’s potential is vast, from reading plays with distinct character voices to creating personalized audiobook gifts. Imagine being able to eventually personalize and customize the voice you have read to you.

Related

The FT reports on the excess inventory in certain parts

Each Tuesday, I’ll post interesting infographics I’ve come across: Here

One of the next sensors to become mainsteam on the