The New ChatGPT Is More Show Than Tell

OpenAI unveiled an “Intelligent UI” update for ChatGPT on Wednesday, introducing visual, custom elements to enhance the answering of user queries. This new experience is powered by the company’s latest GPT-6 model, which became available to paid users today and free users tomorrow.

During a test of the user interface prior to its launch, the visual elements felt like a welcome improvement. When asked for a slug diagram, ChatGPT generated an anatomy-focused visual with buttons to explore each part, from tentacles to the muscular foot. Similar interactive tools were demonstrated, such as a San Francisco apartment affordability calculator with movable sliders, and an airplane seat explorer to compare cabin layouts of different planes.

While these visual elements proved useful in some cases, OpenAI’s product manager, Aarush Selvan, noted that not every ChatGPT response will include a chart or graphic. "We have a design team that spent a lot of time figuring out when adding a diagram, chart, or buttons adds value versus when it's starting to feel cluttered," Selvan explained.

These visual results echo Google’s recent expansion of bespoke results in Search, called "generative UI." Google’s example involved adjustable graphics for understanding complex concepts like black holes using sliders. Chatbot experiences have increasingly become multimodal, blending text, images, and voice into an interactive user experience. OpenAI’s update for a more visually appealing and interactive ChatGPT signals that the multimodal era is upon us, reaching over a billion users.

This shift to generative user interfaces may pose a challenge for designers, as AI increasingly dictates software design for individual users. With AI tools capable of rapidly generating custom software, chatbots are becoming more personalized and attuned to individual preferences. The future of chatbots may see a greater emphasis on visual aids, such as slug diagrams, and designers must adapt to this evolving landscape.