ChatGPT Can Now Observe and Engage with You Live in Real Time

OpenAI's ChatGPT has introduced a new video interaction feature, allowing users to engage in real-time AI analysis through their smartphone cameras. This long-awaited capability enables face-to-face conversations with ChatGPT for activities like solving math problems, offering recipe suggestions, telling stories, and even playing games or guiding learning processes with children.

The "Advanced Voice Mode with vision" is available exclusively to Plus, Team, and Pro subscribers, with monthly fees of $20 and $200, respectively. Users can activate this feature by tapping the voice icon next to the chat bar and selecting the video button; screen sharing requires an additional tap on the three-dot (hamburger) menu.

During a livestream on Thursday, OpenAI’s Chief Product Officer Kevin Weil announced the update as part of the company’s "12 Days of OpenAI" series. Other notable releases include the public launch of the o1 model, the introduction of the ChatGPT Pro plan, enhanced reinforcement fine-tuning for customized models, and the debut of the generative video app Sora.

While Google and Meta are also developing similar AI assistants, OpenAI’s new features are currently limited to specific subscription tiers, with a broader rollout expected early next year. Notably, users in the EU will have to wait for these features to become available.

The release of this functionality has faced delays, initially promised within a few weeks, but postponed due to concerns over the unauthorized mimicry of Scarlett Johansson's voice in the advanced voice mode. Now, with the official launch of the video mode, OpenAI reaffirms its commitment to enhancing user experience and driving technological innovation.