AI Pulse · AI Trends Pulse

OpenAI just rolled out ChatGPT Voice for desktop users on macOS and Windows. This is not the mobile voice feature you might already use. It’s a control layer for directing multiple AI agents at once, by voice, inside ChatGPT Work or Codex. The underlying model is called GPT-Live. It’s going out to Plus, Pro, Business, Education, and Enterprise tiers now, according to the official announcement.

The real story is not the feature itself. It’s the shift both OpenAI and Anthropic are making toward the same interaction model: you talk, a fleet of agents does the work. This is not speculative. Both labs are building toward a future where you boss around multiple specialized agents by voice, and the system figures out which one handles what. That’s the default enterprise interface they’re racing to own.

For business owners, this matters because it changes how you think about deploying AI. You’re not just automating one task with one tool anymore. You’re managing a team of agents that can handle research, code, analysis, and coordination in parallel. The question is whether your company is set up to use that kind of system, or if you’re still treating AI like a chatbot you check in on occasionally.

This is exactly the kind of capability we build into an AI command centre at /resources/omni, where you can direct multiple agents across your actual workflows without duct-taping tools together. The frontier labs are moving fast. If you’re still figuring out how to use one assistant, you’re already a generation behind.