News

ChatGPT desktop is equipped with "human voice": GPT-Live voice is online, and it will work in the background as you speak. OpenAI today made a major upgrade to the ChatGPT desktop application - this client, which has been integrated with Codex, has officially introduced C driven by the GPT-Live model.

2 min read
ChatGPT desktop is equipped with "human voice": GPT-Live voice is online, and it will work in the background when you speak. OpenAI today made a major upgrade to the ChatGPT desktop application - this client, which has been integrated with Codex, has officially introduced the ChatGPT Voice voice mode driven by the GPT-Live model. In the new version, users can directly use voice to initiate, check or adjust tasks in the three sections of Chat, Work and Codex, moving multi-threaded work collaboration from typing on the keyboard to using words. The origin of this line is not short. OpenAI added voice recognition to ChatGPT for the first time in 2024, and then continued to attack voice technology all the way. It was not until the newly released GPT-Live model that the full-duplex capability was truly realized - it can listen and speak at the same time, making the experience of AI conversations closer to real people. OpenAI made its vision very straightforward in its statement: In the future, users only need to verbalize their needs, and ChatGPT will be able to follow your ideas and quickly advance multiple tasks. The most critical feature of the new voice mode is that it allows users to start multiple workflows from a single conversation, and while the agent silently completes the task in the background, you can continue to chat with it. OpenAI took planning a business trip as an example: you can ask ChatGPT to check the calendar to check for conflicts, scan the inbox to see if there are any changes to flights, and prepare meeting minutes conveniently - and all of this happens while "you are making coffee or doing other things." Compared with simply using voice as an input box, this kind of interaction of "you speak, it does things, without delaying each other" is exactly where GPT-Live wants to widen the gap. via AI News (author: AI Base)