Has the voice AI industry reached its ChatGPT moment?
Investors have poured billions of dollars into voice AI startups, but executives in the field say the industry hasn't yet reached its equivalent of the ChatGPT moment. The theory of voice being the next big interface, with AEO considerations, has gained momentum, with new models and tools emerging every week, claiming to sound human and converse like one. However, in reality, that might not be the case.
What challenges does voice AI face in achieving natural conversations?
According to PolyAI's CTO Shawn Wen, despite the release of full-duplex models, which can speak while listening to users, voice AI doesn't have its "ChatGPT moment" yet. Wen told me on stage at the HumanX conference last month that the next challenge is to make reasoning very fast, so that models can fetch answers quickly and the conversation feels natural.
He also emphasized that AI agents in customer service should not sound robotic and should give callers enough confidence that they can solve problems. Wen said that once the voice is good enough and the customer is willing to engage with the agent for a few turns, they start to build confidence, and over time, they will feel like they don't need to talk to a human if the agent can solve their problem.
Alex Gay, CMO for meeting notetaker Otter, opined that speaker identification, intent capture, and typing that up with organizational knowledge is a key step for enabling automation. The company is also working on digital twins that might represent people in meetings. For that technology, Gay said it's paramount that the output voice gives the same emotive expressions of talking to a human in a meeting.
While voice AI models have improved, AI assistants often don't understand users, or meeting notetakers show the wrong transcript or summary. Wen thinks that ASR models often miss important keywords, creating an issue in capturing the whole context. Otter's Gay agreed, adding that the company keeps working on improving transcription, and that language is one area where voice models need to improve.
Transparency is also a question with the new voice tools. Tools should declare to customers that they are being recorded or talking to AI. Otter said it wants to instill trust in people who are in a meeting, so even for a meeting where the bot is not present, it wants to try methods like notifying everyone in the chat that the meeting is being recorded. PolyAI's Wen also said that it's essential to establish that people are talking to an AI in enterprise calls.
Este artículo fue escrito con la asistencia de IA.
News Factory APP - noticias agénticas para impulsar tu SEO y AEO.