US20260169767
2026-06-18
Physics
G06F9/451
The AI-powered UI design tool leverages a computer hardware system to generate graphical user interface (GUI) designs for applications. This system includes a neural network connected to both immutable and mutable memory. The process begins by receiving input that defines the application's requirements, which are then used to populate the immutable memory. The mutable memory is filled with descriptions of previous UI components, allowing the neural network to generate new component descriptions for the GUI in response to prompts.
Current AI tools, particularly large language models (LLMs), enhance software development by providing coding assistance, but similar tools for UI design are lacking. This absence poses challenges for developers who need efficient solutions to streamline UI design processes. Traditional methods require recreating designs from scratch, which is time-consuming and often inefficient. Additionally, developers might lack UI/UX design expertise, resulting in issues such as improper function division and design inconsistencies.
The system employs a neural Turing machine to generate new UI component descriptions, ensuring consistency with a UI component design library. It utilizes data from immutable memory, including historical and design system criteria, and application-related requirements. The process involves generating example data and a data model from the application context, which is then used to create prompts for the neural network. Feedback from previous designs can refine these prompts, addressing position conflicts, data model issues, and visual checks.
The hardware system consists of a neural network, immutable and mutable memory, and a processor that executes operations based on application requirements. The immutable memory is populated with these requirements and a UI design library, while mutable memory stores previous component descriptions. The neural network generates new component descriptions using this data, ensuring alignment with the design library. The system also uses feedback from past designs to improve new component descriptions.
The AI-powered design tool uses deep learning through a neural network to produce new UI designs. This network functions like biological neurons, using layers of nodes with associated weights and thresholds. It learns from training data to improve its output. The tool's methodology involves identifying prior component versions, using criteria from immutable memory, and constructing prompts to generate new descriptions. Components are described by position encoding, library identification, and textual description, ensuring they fit within the overall GUI design.