AI Engineer releases its World's Fair 2026 Voice Track on YouTube
AI Engineer frames the track around a thesis: voice agents “do not fail at talking. They fail at listening.” Its announced lineup includes speakers from Google DeepMind, OpenAI, Plivo and Roku.
TLDR
AI Engineer announced on September 15 that its World's Fair 2026 Voice Track was live, sharing a YouTube link. The listed talks include speech-to-speech model research at Google DeepMind, five early voice-agent failure modes from Plivo, and Roku's “Act, Confirm, or Stop?” The organizer emphasizes listening and the case for agents that know when not to speak.
Combined views
—
1 Source, first seen ago
