Generative Interfaces for Language Models

2025-08-26

AI Summary

Current large language models (LLMs) are constrained by linear request–response interaction paradigms, limiting their effectiveness in multi-turn, high-information-density, and exploratory collaborative tasks. To address this, we propose Generative UI—a novel paradigm enabling LLMs to dynamically generate interactive, task-specific user interfaces. Methodologically, we introduce (i) a structured interface representation, (ii) an iterative optimization framework grounded in user feedback, and (iii) a task-adaptive UI generation mechanism. We further design a multidimensional evaluation framework assessing functionality, interactivity, and affective impact. Experimental results demonstrate that Generative UI outperforms conventional chat-based interfaces in over 70% of test cases, yielding statistically significant improvements in both user preference and task efficiency. Our analysis explicitly identifies high-benefit usage scenarios and key adoption drivers, underscoring the paradigm’s potential to advance human–LLM collaboration beyond static dialogue.

Technology Category

• Humans and AI: Intelligent User Interfaces
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Large language models (LLMs) are increasingly seen as assistants, copilots, and consultants, capable of supporting a wide range of tasks through natural conversation. However, most systems remain constrained by a linear request-response format that often makes interactions inefficient in multi-turn, information-dense, and exploratory tasks. To address these limitations, we propose Generative Interfaces for Language Models, a paradigm in which LLMs respond to user queries by proactively generating user interfaces (UIs) that enable more adaptive and interactive engagement. Our framework leverages structured interface-specific representations and iterative refinements to translate user queries into task-specific UIs. For systematic evaluation, we introduce a multidimensional assessment framework that compares generative interfaces with traditional chat-based ones across diverse tasks, interaction patterns, and query types, capturing functional, interactive, and emotional aspects of user experience. Results show that generative interfaces consistently outperform conversational ones, with humans preferring them in over 70% of cases. These findings clarify when and why users favor generative interfaces, paving the way for future advancements in human-AI interaction.

Problem

• Addressing inefficiency in linear request-response LLM interactions
• Proposing generative interfaces for adaptive UI-based engagement
• Evaluating generative versus chat-based interfaces across user experience dimensions

Innovation

• Proactive UI generation for adaptive engagement
• Structured interface-specific representations for task translation
• Iterative refinements for dynamic user interaction