Voice AI technology has attracted substantial investment and development across various applications, from customer service platforms to meeting transcription tools. Despite these advances, key industry leaders believe voice AI has yet to experience a breakthrough moment comparable to ChatGPT’s impact in conversational AI. Shawn Wen, CTO of enterprise voice AI provider PolyAI, emphasized the importance of enhancing the speed and naturalness of AI responses, noting that while full-duplex models—those that can listen and respond simultaneously—have been developed, the technology still faces challenges in maintaining fluid and confident interactions with users.

For voice AI to achieve widespread adoption, Wen explained that users need to feel assured that these agents can effectively resolve their issues, eventually reducing reliance on human support. Alex Gay, CMO of meeting tool Otter, highlighted the complexities of replicating human-like interactions, particularly the emotional and relational dynamics often present in meetings. Otter is pursuing innovations such as digital twins to embody participants, but Gay stressed that without genuine conversational depth, these digital avatars risk being reduced to simple Q&A bots rather than meaningful collaborators.

One major hurdle voiced by both Wen and Gay relates to the accuracy and contextual understanding of speech recognition technology. Automatic Speech Recognition (ASR) models frequently miss crucial keywords, creating errors in transcription and subsequent summaries which undermine user trust and the effectiveness of downstream automation tasks. Improving ASR remains fundamental, as inaccurate transcriptions compromise not only immediate communication but also productivity tools built upon these outputs.

Transparency around voice AI usage emerged as another critical discussion point. Both Otter and PolyAI prioritize informing users when they are interacting with AI or being recorded, a safeguard essential for building trust. Gay mentioned Otter’s efforts to notify meeting participants about recordings, even if an AI assistant is not actively present, while Wen underscored the importance of clear disclosure in enterprise customer service contexts. Overall, these executives agree that while voice AI continues to evolve, significant advancements are still needed before it reaches its pivotal “ChatGPT moment.”