OpenAI ChatGPT Desktop App Adds Voice Mode for Background Multi-Tasking
OpenAI has updated its ChatGPT desktop application to introduce a voice mode powered by the GPT-Live full-duplex model. This update allows users to initiate, monitor, and adjust multiple background tasks simultaneously using voice commands. This update marks a significant shift toward agentic AI workflows, where users can delegate complex, multi-step tasks via voice and continue conversing while the AI executes them in the background. It enhances human-computer interaction by making AI assistants behave more like proactive, multitasking human collaborators. The GPT-Live model enables full-duplex communication, meaning the AI can listen and speak at the same time to handle interruptions naturally. Users can trigger workflows across Chat, Work, and Codex sections, such as checking calendars and organizing emails, while performing other physical tasks.
## BACKGROUND
Traditional voice assistants operate in a half-duplex mode, requiring users to wait for the AI to finish speaking before responding. Full-duplex voice models, like GPT-Live, allow for bidirectional, real-time conversation where both parties can speak simultaneously. Agentic workflows represent an evolution in AI where models transition from simple chat interfaces to autonomous agents capable of planning and executing multi-step tasks.