OpenAI has announced the long-awaited addition of video chat and screen sharing features to its advanced voice mode.
This new feature is now available to ChatGPT Teams, Plus, and Pro users on iOS and Android mobile applications, and is expected to roll out to ChatGPT Enterprise and Education subscribers in January next year. However, users from the EU, Switzerland, Iceland, Norway, and Liechtenstein will not have access to this advanced voice mode.
OpenAI first mentioned this feature in May of this year, showcasing how GPT-4o could "watch" games and explain the gameplay. Subsequently, the advanced voice mode was officially launched to users in September. Users can initiate video calls via a new button on the screen of the advanced voice mode, similar to Facetime, allowing ChatGPT to respond in real-time to the content shown by users in the video.
In OpenAI's demonstration, ChatGPT used the video feature to assist a user in brewing coffee. It could recognize coffee equipment, guide the user on when to insert a filter, and evaluate the brewing results. Additionally, ChatGPT can remember the names of those it interacts with, demonstrating a higher level of interactivity. This type of video interaction is similar to Google's recently launched Project Astra, which can also answer users' questions about objects seen during video chats, such as identifying sculptures on the streets of London.
The screen sharing feature allows ChatGPT to extend beyond the application itself into the browser environment. Users can simply enable screen sharing through a three-dot menu, opening applications on their phones and asking ChatGPT about what they see. In the demonstration, OpenAI researchers activated screen sharing and opened a messaging app, requesting ChatGPT's assistance in replying to a photo message.
However, ChatGPT's screen sharing functionality shares similarities with features recently introduced by Microsoft and Google. Last week, Microsoft launched a preview version of Copilot Vision, allowing Pro subscribers to open Copilot chat while browsing the web, enabling it to recognize photos on web pages or assist in map guessing games. Similarly, Google's Project Astra can read browser content in a comparable manner.
Additionally, OpenAI has introduced a fun and lighthearted "Santa Mode," where users can chat with ChatGPT mimicking Santa's voice.
Unlike the user restrictions of the new features, the "Santa Mode" is available across mobile applications, web versions, and Windows and MacOS applications until early January next year. It's important to note that conversations with Santa will not be saved in chat history and will not affect ChatGPT's memory function.
Key Points:
🎥 New video chat feature, ChatGPT can respond in real-time to what users see.
🖥️ Screen sharing feature launched, users can request help from ChatGPT on their phones.
🎅 "Santa Mode" launched, allowing users to interact with ChatGPT mimicking Santa's voice.
