Earlier this month, I witnessed a digital miracle. In a press briefing ahead of CES, Nvidia showed off a demo for its Ace microservice, an AI suite capable of generating fully voiced AI characters. I watched in awe as a demoist spoke to an in-game NPC through a microphone, only to have the digital character respond in real time. It was a true sci-fi feat, but there was one question: How did it learn to do that?
Nvidia gave an ambiguous response, claiming there was “no simple answer.” The statement set off a firestorm, as users on social media assumed the worst. Speculation arose that Ace was trained on content Nvidia didn’t have the rights to. Nvidia later claimed it’s only using data it’s cleared to use, but tensions were still high. A mountain of ethical and artistic concerns left gamers skeptical.
Among the spectators watching it all unfold from the sidelines was Purnendu Mukherjee. The software engineer wasn’t another face in the crowd; he created the AI tech at the center of a debate he didn’t start. Mukherjee is the founder of Convai, the generative AI company powering Nvidia Ace. Rather than sitting back and watching someone else try to explain his tool, he was eager to set the record straight.
RelatedSpeaking to Digital Trends, Mukherjee sat down to answer some ethical concerns in a wide-ranging interview about AI tools like his. He offered his thoughts on everything from unemployment fears to worries that AI would sap humanity out of art. For Mukherjee, that’s far from the truth. The Convai founder sees an optimistic future where artists work hand-in-hand with AI to realize their creative visions fully. But when it comes to the hot topic of data usage, his explanation could raise more questions than answers.