PolyAI CTO says voice AI has not yet reached its ChatGPT moment
Investors are pouring billions into voice‑AI startups, but executives say the technology still lacks a breakthrough moment. PolyAI’s chief technology officer Shawn Wen told TechCrunch AI at the HumanX conference that while full‑duplex models can now speak and listen simultaneously, the next hurdle is “very fast” reasoning so answers arrive quickly and conversations feel natural. He added that…
Key points
- PolyAI CTO Shawn Wen says full‑duplex voice models exist but fast reasoning is still needed for natural conversation.
- Otter CMO Alex Gay highlights speaker identification, intent capture, and digital twins as next steps for meeting‑note AI.
- Executives stress transparency, urging tools to notify users when recordings or AI agents are active.
Otter’s chief marketing officer Alex Gay emphasized that speaker identification, intent capture, and digital‑twin avatars are essential for turning meeting recordings into actionable insights. He warned that without emotive, human‑like voice output, meeting assistants become mere Q&A bots. Both executives highlighted persistent gaps in automatic speech recognition, noting missed keywords can corrupt transcripts and downstream actions, eroding user trust. They also called for clearer transparency, urging tools to notify participants when recordings occur or when an AI is speaking.
The discussion underscores that voice AI, while advancing, still needs faster reasoning, higher transcription accuracy, and stronger user‑facing disclosures before it can become a mainstream interface.
These execs think voice AI hasn’t reached its ChatGPT moment yet
TechCrunch AI · 11 October 2026
Loading the full article…
This text was published by TechCrunch AI and written by Ivan Mehta. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Anthropic releases free crash course on building AI agents with Opus 5.5 · 1 src
- Meta's Muse AI permits explicit adult conversations, unlike other Meta apps · 1 src
- Elon Musk exposes Lego order details via Grok, users cancel purchase · 1 src
- Docker releases Docker Agent CLI plugin for building AI agents as containers · 1 src
- Open-source REA hits 100,000 GitHub stars as AI agents decompile software · 1 src
Comments
via GitHub Discussions