Japanese AI device reads stories to your kid with your voice, even if you’ve never read them

Japanese toy maker Takara Tomy’s playthings often have a technological side to them, like a photo sticker booth with a time travel mode that takes you back to the ‘90s or a dancing Pikachu robot. Takara Tomy’s newest creation, though, is especially hi-tech.

The Coemo’s name is a play on words with koe, the Japanese word for “voice.” So what does the egg-shaped device do? It tells stories to your kids, using your voice, even if you’ve never spoken the story’s words yourself.

It does this through an AI system that synthesizes your voice, then matches the inflection and intonation to provide proper emotion to the stories it tells. To set things up, first you use the Koe Station smartphone app and register your voice by reading a sample passage of text. Supplied with the audio sample, Koe Station can create a vocal profile for you, which you then export to the Ceomo app and send to the Coemo device.

The app also lets you select one of the 45 stories Coemo comes pre-loaded with, including both original stores and fairy and folktales from both Japan and abroad, such as “Momotaro/Peach Boy,” “Little Red Riding Hood” and “The Three Little Pigs.” Once you’ve registered your voice and selected a story, the AI generates narration from your voice profile, which plays out of the device’s speakers. You can even register multiple voices with a single Coemo, allowing you to assign grandma, grandpa, or other relatives who’ve created a vocal profile as the storyteller for different tales.

▼ Video of the Coemo in action

Coemo goes on sale this fall, priced at 12,980 yen, with preorders stating June 14. No word on whether or not it’s compatible with Takara Tomy’s robo grandchildren.