During a Wednesday livestream, OpenAI unveiled GPT-Live, a new voice model powering ChatGPT Voice, designed to make talking with AI feel more like talking to a human by allowing the assistant to listen and speak simultaneously instead of waiting for a clean pause. The assistant will acknowledge the user with intermittent "mmhm," "yeah," and "got it" while the human is still talking. In one demo, a user asked ChatGPT to fact-check a meeting date while also checking weather and traffic, and ChatGPT responded with short acknowledgments like "hmm" and "sure" before continuing through the request. OpenAI also showed off ChatGPT's ability to translate language in real time, with the full-duplex assistant keeping pace naturally as a conversation unfolds. Greg Brockman, OpenAI's president, described the update as a "much more natural way of interacting with your computer." In May, Thinking Machines, the AI lab led by former OpenAI chief technology officer Mira Murati, teased similar technology, stating its interaction models handle input and output continuously across audio, video, and text.