Speech-to-Speech Model Research at Google DeepMind — Valeria Wu Fon & Tom Ouyang, Google DeepMind
Source
AI Engineer
Author
AI Engineer
Date
Terms in this piece · Glossary
multimodal — A model that works with more than text — reading images, audio, or video, and sometimes generating them too.
Why it matters
Names the core three-way tradeoff, conversational latency, task intelligence, and multimodalA model that works with more than text — reading images, audio, or video, and sometimes generating them too.Full definition →, that governs every speech-to-speech model, including why raising thinking budget can wreck latency.
video"My name is... my name is...": A Linguistic Map for Voice Agents — Midam Kim, ServiceNow
videoAct, Confirm, or Stop? Smarter behavior for AI assistants, wearables & robots — Amit Desai, Roku
videoVoice Agents Can Just Do Things — Charlie Guo, OpenAI
videoYour Voice Agent is Just a Walkie Talkie — Neil Zeghidour, Gradium
