Voicemod has created 20 “AI humans,” digital characters that allow gamers to speak in voices other than their own.
A Spanish artificial intelligence company has launched a batch of “artificial intelligence humans” that allow people to modify their voices in real time in video games.
Valencia-based Voicemod released 20 humanoid characters on Wednesday, ranging from women in their 20s to older men. The voices are trained from the voices of professional voice actors.
In a demo of the technology on a call with CNBC, Voicemod CEO Jaime Bosch showed himself speaking regularly and changing it into several different voices, from a high-pitched female voice to a baritone.
Players can download the app on their website apple Mac or Windows PC and incorporate Voicemod’s technology as a “virtual microphone” sitting between the microphone applications they use to start speaking through alternate characters.
With talents from leading technical universities in Valencia and Barcelona, Voicemod has been working on speech synthesis and interactive audio features since 2014, with many employees specializing in music technology and audio.
Voicemod’s technology is used by more than 40 million people, and people on the social app Discord can talk to each other in voices other than their own while playing games.
“We have an amazing creative community that’s mostly about having fun, having fun with friends, and getting a sense of belonging,” Bosch told CNBC.
“One of my biggest favorite use cases is shy people — some people have written to us saying I can’t really socialize with people who can do that right now.”
This is a milestone in the field of conversational artificial intelligence. While many of today’s AI algorithms allow people to submit text and receive AI-generated content, ensuring this feat is done in real time is much more difficult.
The technique requires a lot of computing power. The production and patenting of the proprietary algorithmic models behind it require significant investment and talent.
To that end, Voicemod has raised $23 million in cash from venture capitalists including Leadwind and Bitkraft Ventures.
Bosch is no stranger to the risk that the technology could end up being misused — voice-changing technology, for example, could be used to impersonate major political figures or trick people.
“It’s something I think about every day, and it’s something our company thinks about every day in terms of creating sound,” he said.
The company is nearing completion of a “watermarking” solution that can identify whether a sound was generated using a modified system. It is also discussing standardization of such systems with other companies and ensuring the safe release of the voice-changing technology.
“One thing is that clear laws will come,” Bosch said. “We know that Europe is struggling with this. The reality is that, often times, companies act faster than the law.”
watch: Cannes Lions Lions 2023: AI won’t make the market lazy – TBWA CEO