ChatGPT’s new GPT-Live voice model can speak and listen at the same time (www.techradar.com)

OpenAI has announced an upgrade to ChatGPT's voice mode, introducing the new GPT-Live voice model that allows for simultaneous speaking and listening, significantly enhancing conversation flow. This advancement addresses previous limitations where the AI would interrupt users and lacked the depth of its text-based counterpart. The model integrates with GPT-5.5 for answering complex queries, creating a seamless interaction experience by notifying users when it is retrieving information, thereby maintaining a natural dialogue. Additionally, the new voice mode features simultaneous translation, enabling real-time translations in major languages as users speak, similar to professional interpretation settings. This capability could transform how users engage with AI in multilingual environments. The rollout of GPT-Live-1, with both standard and mini versions for different user tiers, marks a pivotal shift towards making voice interaction a primary interface for AI, a step that may redefine user experiences with AI technology moving forward.