OpenAI has made important updates to its macOS app's advanced voice mode, with the most notable change being the addition of image uploading capabilities. Previously, this feature was unavailable. After this update, users can upload images through the voice interface and display them directly on the interface, significantly improving the convenience and intuitiveness of users' interaction with ChatGPT. This update is not a simple addition of features, but an important step for OpenAI to improve user experience and improve product functions, laying the foundation for a more advanced AI interaction model in the future.
Recently, OpenAI made a major update on its macOS app's advanced voice mode interface, introducing a brand new image upload feature. In the past, this function could not be clicked, but now users can finally upload pictures through the voice UI, and the uploaded pictures will be displayed on the interface.
However, it is worth noting that ChatGPT does not prompt the user that it cannot see the uploaded images. Users can upload an image to discuss related to ChatGPT, and all of this becomes more intuitive.
In addition to uploading pictures, users can also directly use the laptop's camera to take photos, so that they can instantly share the captured content.不过，目前还没有屏幕共享、摄像头共享和视频共享的功能，这些功能仍在待定状态。
在这次更新中，还对某些消息的显示方式进行了小幅调整。例如，关于如何开始新聊天以使用高级语音模式的提示，界面也进行了改版。 It is worth mentioning that if the user is in a session involving video sharing, ChatGPT may prompt the user to explicitly share the video to help improve the model.这条提示出现在用户启用了分享视频数据以供模型训练的选项时。然而，目前这一选项在UI 中尚未上线，可能会在屏幕和视频共享功能推出时一起发布。
这次更新可能预示着未来Canvas 编辑器的推出，这个功能之前曾被发现在开发中。 Next, the full rollout of voice mode may take longer unless OpenAI decides to release first, regardless of the video and screen sharing capabilities.
Key points:
新增图片上传功能:用户现在可以通过语音UI 上传并讨论图片，提升互动体验。
直接拍照分享:用户可以使用笔记本摄像头拍照，但尚无视频共享功能。
未来展望:Canvas 编辑器即将推出，语音模式的全面发布或将推迟。
此次OpenAI的更新，展现了其在提升用户体验和拓展AI应用场景上的持续努力。未来，随着Canvas编辑器和更完善的语音模式的推出，用户将体验到更加便捷和智能的AI交互方式。 Looking forward to more surprises in OpenAI.