BTS Managers Invest $3.6M in Synthetic Voice Startup
The management group behind K-Pop supergroup BTS has invested about $3.6 million into synthetic voice startup Supertone. Big Hit Entertainment is funding Supertone’s development of “a hyper-realistic and expressive voice” that will supposedly be indistinguishable from an actual human, the latest sign of interest from the entertainment field in this kind of technology.
Supertone is known for applying its tech to Korean music stars. The company garnered a lot of positive and critical attention last month when it revived the voice of Kim Kwang-seok, a South Korean folk singing star who passed away a while ago. The company generated a version of her voice to sing on a TV show explicitly about AI competitions with people. Of course, the tech is not limited to performers who are no longer around. That is likely why Big Hit is writing them a check. The demand for BTS vastly outweighs the group’s ability to physically record or perform. They could never sleep or do anything but record and still likely not fulfill every request. If their voices can be synthetically generated, however, suddenly all of the licensing for games, cartoon shows, personal messages, and any other audio-only work becomes a lot more feasible. Supertone has even suggested that idea previously before the deal with Big Hit was made public.
“We are pleased to partner with Big Hit Entertainment, who are leading the global entertainment market,” Supertone CEO Lee Kyo Gu said in a statement. “We will accelerate the globalization of our AI technology through our partnership with Big Hit Entertainment, and further provide the necessary connection for Korean companies to maintain their technological superiority in this age of global digital content and production.”
The applications of synthetic speech generation are enormous. Voice acting and performing are definitely some of the biggest and most potentially lucrative uses for the technology, however. Amazon used its advanced voice synthesis to give Alexa the voice of Samuel L. Jackson, for instance. The idea of branded voices, whether based on real or fictional characters is encouraging company partnerships of all sorts. Amazon’s Brand Voice feature is giving a unique voice to branded Alexa skills, like making KFC Canada’s Alexa skill speak like Colonel Sanders. And Microsoft has started accepting applications from customers interested in its Custom Neural Voice service, which is already being used to give a voice to characters as wide-ranging as Bugs Bunny and Flo from Progressive Insurance. SoundHound is augmenting its Houndify voice AI platform to sound more lifelike by working with digital voice interface creator ReadSpeaker’s text-to-speech (TTS) software, for instance.
Voice cloning startups like Replica Studios and Resemble AI have their own versions of home tools for generating an artificial voice from recorded audio but are both pitching toward the entertainment industry as well. The tech can be used for non-commercial experiments as well. One great demonstration is the trailer for Skryim made by a big fan of the video game is voiced solely by synthetic voices built with an AI tool to create an original script.