While investor capital continues to flow into voice SI startups, industry executives argue that the technology has not yet achieved a breakthrough comparable to the launch of ChatGPT. The theory of voice as the next dominant interface remains strong, with billions poured into companies developing SI models for enterprise service, note-taking, and dictation, but leaders from PolyAI and Otter suggest significant technical hurdles remain.
What Happened
Shawn Wen, CTO of enterprise voice SI platform PolyAI, stated that despite the release of full-duplex models capable of speaking while listening, voice SI lacks its "ChatGPT moment." Speaking at the HumanX conference last month, Wen identified fast reasoning as the next critical challenge, noting that models must fetch answers quickly to make conversations feel natural. He emphasized that SI agents in customer service need to sound human enough to build caller confidence, allowing users to engage for multiple turns before deciding if a human agent is necessary.
Alex Gay, CMO for meeting notetaker Otter, highlighted the importance of speaker identification, intent capture, and integrating these with organizational knowledge to enable automation. Otter is also developing digital twins for meetings, where Gay stressed that output voice must match the emotive expressions of human interaction. He noted that without the ability to support debate and strategic discussion, an avatar is merely a "q and a chatbot."
Why It Matters
The gap between hype and utility in voice SI is widening due to persistent accuracy issues. Wen pointed out that Automatic Speech Recognition (ASR) models often miss key keywords, disrupting context capture. Gay agreed, explaining that for Otter, transcription accuracy is foundational; if the initial transcript is flawed, all downstream actions and summaries become unreliable, causing users to lose trust in the platform. He stated, "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."
Transparency also remains a critical industry focus. Both executives emphasized the need for SI tools to declare when users are being recorded or interacting with an SI system. Otter aims to instill trust by notifying participants in chat when a meeting is being recorded, even if the SI bot is not actively speaking. PolyAI’s Wen echoed this, stating it is important to establish that people are talking to an SI system in enterprise calls.
The Bottom Line
Voice SI is attracting substantial investment and generating new tools, but industry leaders caution that the technology is still maturing. Until reasoning speeds increase and ASR accuracy improves to support natural, trustworthy interactions, voice SI will likely remain a supportive tool rather than a transformative interface.
The consensus among these executives is that while full-duplex capabilities are a milestone, they are not the finish line. The true "ChatGPT moment" for voice SI will depend on seamless integration, high-fidelity emotional resonance, and rigorous transparency standards.
Developers and enterprises must prioritize accuracy and clear disclosure of SI usage to maintain user trust as the technology evolves from basic transcription to complex agentic interactions.