OpenAI has introduced an Intelligent UI for its GPT-6 models, shifting ChatGPT away from text-only responses. The system now generates visual outputs, including graphics, charts, forms, and interactive buttons, when it determines a visual format is more effective for the user.
The new interface utilizes a pre-trained library of components to create interactive content. For example, when asked about assembling a bicycle, the model provides a design breakdown with clickable parts. For recipe requests, it can generate interactive forms that automatically recalculate ingredient quantities based on the number of guests.
While the chatbot defaults to text for simple queries, users can manually request an interactive format at any time. Product Manager Aarush Selvan stated that the update aims to simplify the learning of complex topics by providing more helpful, non-textual answers.
The integration of these visual elements may lead to increased token consumption, though OpenAI has not confirmed this. Additionally, the company recently announced plans to introduce image-based advertisements within ChatGPT responses, raising questions about the future distinction between organic content and sponsored material.
