– A former OpenAI researcher has raised $40 million for a startup that will build artificial intelligence models that can communicate emotionally with humans through voice, the company announced Monday.
The startup, WaveForms AI, is backed by Andreessen Horowitz and valued at $200 million, CEO Alexis Conneau told Reuters.
Alexis Conneau co-created the voice mode capability on OpenAI’s GPT-4o model, which when released earlier this year demonstrated the ability to respond in real time and without delay, as well as handling interruptions – two features of realistic conversations that have so far eluded AI voice assistants.
After the launch, OpenAI CEO Sam Altman posted the word “her” on social media platform The story of a man who falls in love with his AI assistant, played by Scarlett Johansson.
OpenAI’s launch of GPT-4o sparked controversy when Scarlett Johansson accused the startup of copying her voice in the film. OpenAI said the voice was not an impersonation of Johansson and that the company had used another actress.
Mr. Conneau and Coralie Lemaitre, co-founder of WaveForms AI, said the startup will use the funding to train new audio AI models that will solve the problem of having a voice conversation with an AI robot that is indistinguishable from one of a human. He added that audio is key to making interactions with AI feel “deeply human.”
“Audio…conveys emotions and provides emotional responses to users,” he said.
Conneau said the company plans to create consumer software focused on the emotional connection between the user and AI, but declined to elaborate.
“The idea is to create new, more immersive experiences with AI, more enjoyable experiences,” he told Reuters. “Many companies are currently focused on superintelligence, but not really on the quality of human-computer interaction.
