Testing Finally Begins: ChatGPT Invites Select Users for New Real-time Voice Conversation Mode

At a press conference held previously by OpenAI, the company introduced a new feature for ChatGPT: real-time voice conversation mode. However, only now has OpenAI started to invite a small number of users to test this feature.

The editor of the technology website @TestingCatalog received an invitation. The feature, dubbed Advanced Voice Mode, supports natural voice conversations with real-time responses, allows for the adjustment of emotional tone and intonation, and enables real-time visual detection through camera activation.

These functionalities, as described on the invitation interface, mirror what was announced during the press conference. Inviting a very limited number of users for testing allows OpenAI to collect data and make improvements.

It's important to note that if users accept the invitation and enable the new voice conversation mode, all input data, including but not limited to text, voice, or content captured by the camera, will be shared with OpenAI for model training purposes.

Therefore, users should avoid disclosing confidential information to ChatGPT and refrain from capturing sensitive content with the camera, to mitigate potential data leakage risks.

The real-time visual detection feature deserves a special mention. Upon enabling this feature, ChatGPT utilizes the camera to capture live images, which are then analyzed by cloud-based AI to provide relevant responses.

For instance, the popular task of selecting watermelons on social media platform X, which initially required taking photos and marking the watermelons, can now be simplified with the real-time voice conversation mode, allowing ChatGPT to identify and respond to the watermelons directly. This method is indeed simpler.

However, it might take a few more months before this new mode is officially launched. During this period, ChatGPT users can look out for the invitation interface to try out this novel feature.