ChatGPT Speaks and Controls Your Computer by Voice

When Your Voice Becomes the Keyboard

Interacting with AI is no longer limited to typed text. OpenAI has announced a significant update to the ChatGPT desktop application, giving users the ability to direct the app and manage complex tasks simply by speaking — no keyboard or mouse required. This shift doesn't just reshape how we interact with our computers; it opens entirely new horizons in what digital productivity can mean.

What's New in This Update?

The update is powered by OpenAI's new family of voice models, launched under the name GPT-Live — models designed to handle real-time voice interaction with an advanced level of fluency and responsiveness. The most notable features of this update include:

• Voice control of AI agents: Users can direct multiple agents operating within ChatGPT Work or Codex using direct voice commands.
• Multi-step task execution: The app can process compound instructions encompassing a sequence of consecutive actions all at once.
• Browsing and app interaction: The model can access websites and desktop applications within the context of the requested task.
• Screen reading on macOS: Through the Appshots feature, the app can analyze on-screen content — including alternative text for images — deepening its level of understanding and interaction.

A Real-World Example from a Development Environment

To demonstrate the update's capabilities, OpenAI published a demo video showing a developer issuing a single voice command asking ChatGPT to create a new thread, open a pull request, and identify the root cause of a code bug — all at once. What once required minutes of navigating between tools and windows is now achievable with a single spoken sentence, reflecting the genuine advancement this update brings, particularly for developers.

Integration with Mobile Devices

This feature isn't limited to the desktop. The company clarified that users can access Codex by voice from the iOS app via remote connection, making the experience a seamless extension across devices rather than something confined to a single platform.

Competition Heats Up: Anthropic Responds with Its Own Update

In the same vein, Anthropic hasn't stood by idly. The company announced a comparable update to voice mode in its Claude assistant, which now supports all three of its models — Opus, Sonnet, and Haiku. Notably, this update allows Claude to interact with popular external apps such as Gmail, Google Calendar, Slack, Notion, and Canva, making it an effective tool in everyday work environments rather than just a conversational assistant.

Voice: The Gateway to the Next Generation of Productivity

This accelerating trend among leading AI companies signals that the voice interface has become a genuine strategic bet — not merely an added feature. The ability to issue complex commands by voice, receive real-time feedback, and control an entire digital workflow without touching a device is redrawing the boundaries of what it means to be productive in the age of AI. The question is no longer whether these interfaces will become widespread, but how quickly we will adapt to them.

✦ بقلم فريق دروب أيديا