One example is an interactive anatomy diagram of a slug, where users can select individual body parts to learn more about them. ChatGPT can also build calculators with adjustable sliders, such as a tool that compares apartment costs with income and other expenses.
Travel-related prompts can produce similarly customized results. In one test, ChatGPT generated an airplane seat comparison tool covering multiple Alaska and Delta aircraft. Users could select individual seats for more information while exit rows were visually highlighted.
Other examples show how the system can turn explanations into interactive simulations. A question about the offside rule in English football could produce a field graphic that allows users to reposition players and see how the relevant boundaries change. ChatGPT can also create tools such as a dinner-party planner with a guest-count slider that adjusts a menu and grocery list.
The visual components are generated specifically for each request rather than pulled from a fixed library of interfaces. Information used to create them may come from online sources, which are expected to be cited in the response.
OpenAI does not intend to add interactive elements to every answer. Aarush Selvan, a product manager at OpenAI, said the company has focused on determining when visuals improve a response rather than simply making it busier.
“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 said.
Users who prefer simpler responses can also tell ChatGPT to reduce the number of visual elements it generates.
OpenAI told CNET that the richer interface is not expected to substantially increase users’ consumption of usage limits. Although the new outputs can contain more information than a conventional text response, the company described the system as efficient enough that it does not expect a “meaningful” change in usage.
For now, Intelligent UI works with text prompts, though users can dictate those prompts through voice-to-text.
The release adds another visual and interactive layer to ChatGPT as OpenAI expands the product beyond conventional text conversations. The company has also invested in voice interactions and image generation, giving users more ways to communicate with the chatbot and receive information in different formats.
The approach also resembles Google’s generative UI work in Search, where answers can include interactive graphics built around a user’s question. OpenAI’s version brings that concept directly into ChatGPT, allowing the structure of an answer to change depending on what the user is trying to understand or accomplish.
With Intelligent UI, ChatGPT is no longer limited to explaining something through paragraphs alone. When OpenAI determines that interaction would make the answer more useful, the chatbot can instead build a small, purpose-specific interface around the request.
This analysis is based on reporting from CNET & Wired.
Image courtesy of OpenAI.
This article was generated with AI assistance and reviewed for accuracy and quality.
