PolyAI's CTO Shawn Wen said voice AI hasn't reached its ChatGPT moment yet, despite recent full-duplex model advances. He emphasized the need for faster reasoning to make conversations feel natural. "We have reached the milestone of developing full-duplet models.
The next challenge is to make reasoning very fast, so that the models can fetch answers quickly and the conversation feels natural," Wen said at the HumanX conference last month.
Wen also noted that AI agents in customer service should not sound robotic and should build caller confidence.
"I think the next stage will be slightly different because once the voice is good enough, like, and the customer is willing to engage with them for the first two or three turns, they start to build confidence, and over time, they will feel like I probably don't have to talk to a human if the agent can solve my problem," he added.
Alex Gay, CMO of Otter, said speaker identification, intent capture, and typing with organizational knowledge are key steps for enabling automation. He also mentioned the importance of emotive expressions in digital twins representing people in meetings.
"If you think about the meetings that you're in right now, the best conversations that you have are where you can have debate, and strategic discussions, and when you feel like there’s a relationship that underpins it.
If you aren’t able to have that with an avatar, then it’s just a q and a chatbot," Gay noted.
Voice AI models have improved, but issues like missing keywords in ASR and inaccurate transcriptions remain. Wen said ASR models often miss important keywords, creating context issues. Gay agreed, adding that Otter continues to work on improving transcription accuracy.
"For Otter, you know, transcription was never the end point. It was just the layer that we could start to drive some of the productivity gains on the back of.
But if your original transcription didn’t have the accuracy that you needed, all the follow-up actions that you have become flawed," he said.
Transparency is also a concern with new voice tools. Otter wants to instill trust by notifying meeting participants that the meeting is being recorded.
Wen also emphasized the importance of informing enterprise call participants that they are speaking to an AI. "With the new voice tools, there is also a question of transparency.
Tools should declare to customers that they are being recorded or talking to AI," Otter said.
PolyAI and Otter have not yet addressed specific limitations in their voice AI systems, and both companies continue to focus on improving accuracy and natural interaction. The industry is still in the process of refining voice AI to match the conversational fluency of ChatGPT.
Source: techcrunch
