Generative AI systems are transforming content creation, but their usability remains a key challenge. This paper examines usability factors such as user experience, transparency, control, and cognitive load. Common challenges include unpredictability and difficulties in fine-tuning outputs. We review evaluation metrics like efficiency, learnability, and satisfaction, highlighting best practices from various domains. Improving interpretability, intuitive interfaces, and user feedback can enhance usability, making generative AI more accessible and effective.

翻译：生成式人工智能系统正在变革内容创作领域，但其可用性仍是关键挑战。本文探讨了用户体验、透明度、控制度和认知负荷等可用性因素。常见挑战包括输出结果的不可预测性和微调困难。我们回顾了效率、可学习性和满意度等评估指标，并重点介绍了跨领域的最佳实践。通过提升可解释性、优化直观界面和完善用户反馈机制，能够有效增强生成式AI的可用性，使其更具可及性和实用性。