How to check if you've got ChatGPT's new voice mode — and what to try first (www.techradar.com)

OpenAI has begun rolling out its latest voice model, GPT-Live, to all ChatGPT users, introducing innovative features like real-time duplex communication, which allows the AI to listen and respond simultaneously. This enhancement promotes a more natural conversational experience and includes functionality for immediate translation of spoken phrases. Users can identify the new model by the "Live" label in voice chats on mobile and web platforms. While free and paid users will have slightly different models, both can interact with the upgraded features, including adjustable voice intelligence levels. The significance of GPT-Live lies in its ability to simulate real human conversation more closely, enhancing user engagement and interaction quality. Additionally, users can now request visual answers, further enriching the experience through graphics related to their inquiries. However, OpenAI has acknowledged early bugs, particularly concerning memory reliability and pronunciation issues in foreign languages, which are under investigation. Overall, GPT-Live represents a substantial step towards integrating more advanced, human-like interactions within AI systems, setting a new standard for voice communication in AI applications.