10:53Machine-translated from Azerbaijani by AIRead in Azerbaijani
OpenAI Unveils Interface That Creates Charts and Calculators for ChatGPT
On October 7, OpenAI launched a new visual interface for ChatGPT called “Intelligent UI.” Powered by GPT-6, the new feature adds task-specific interactive visuals to text responses. Pro, Plus, Business, and Enterprise users received the feature that same day. The rollout to Free plan and Go users begins on October 8. Access to the GPT-6 model will also become available to everyone this week.
The new interface can add buttons, adjustable charts, interactive diagrams, and task-specific calculators to responses. Users can interact directly with these elements and change their settings. The goal is to make complex topics easier to understand. However, not every response will be visual. Users will be able to reduce the number of such elements if they wish. During the demonstration, several diagrams were used to explain how an airplane wing generates lift. The system can also create illustrated recipe instructions, a diagram showing how a bicycle works, and a map for a multiday hike. Other examples included a personal savings calculator and a travel plan. All elements are generated on the spot based on the user’s request. In an example explaining soccer’s offside rule, users can change the players’ positions and see how the boundary lines are updated. In a dinner plan, a menu and shopping list are created based on the number of guests. The number of guests can be changed using a slider. Visual responses use information obtained from internet sources, and those sources must be cited in the response. In one test, an illustration showing the anatomy of a snail was created, with a separate button added for each part. In another example, a calculator was created to calculate housing costs in San Francisco based on income and expenses. Another test produced an interactive tool comparing the cabin layouts of 4 different airplanes. In the tool, which included Alaska and Delta planes, users could tap seats and view exit rows. Using deep thinking, image generation, and AI agents that perform tasks autonomously generally consumes more of the usage allowance. OpenAI believes the new visual interface will not significantly affect usage limits because the system operates efficiently. For now, the feature accepts only text prompts. However, users can state their request by voice and have it converted to text. Together with voice models and “ChatGPT Images 2.5,” the new feature turns the service, which has more than 1 billion users, into a platform that combines text, images, and audio.
The new interface can add buttons, adjustable charts, interactive diagrams, and task-specific calculators to responses. Users can interact directly with these elements and change their settings. The goal is to make complex topics easier to understand. However, not every response will be visual. Users will be able to reduce the number of such elements if they wish. During the demonstration, several diagrams were used to explain how an airplane wing generates lift. The system can also create illustrated recipe instructions, a diagram showing how a bicycle works, and a map for a multiday hike. Other examples included a personal savings calculator and a travel plan. All elements are generated on the spot based on the user’s request. In an example explaining soccer’s offside rule, users can change the players’ positions and see how the boundary lines are updated. In a dinner plan, a menu and shopping list are created based on the number of guests. The number of guests can be changed using a slider. Visual responses use information obtained from internet sources, and those sources must be cited in the response. In one test, an illustration showing the anatomy of a snail was created, with a separate button added for each part. In another example, a calculator was created to calculate housing costs in San Francisco based on income and expenses. Another test produced an interactive tool comparing the cabin layouts of 4 different airplanes. In the tool, which included Alaska and Delta planes, users could tap seats and view exit rows. Using deep thinking, image generation, and AI agents that perform tasks autonomously generally consumes more of the usage allowance. OpenAI believes the new visual interface will not significantly affect usage limits because the system operates efficiently. For now, the feature accepts only text prompts. However, users can state their request by voice and have it converted to text. Together with voice models and “ChatGPT Images 2.5,” the new feature turns the service, which has more than 1 billion users, into a platform that combines text, images, and audio.