AI-Generated Content (AIGC): A Survey

TL;DR

This survey inventories AI-generated content (AIGC), detailing its definitions, production modes, required data/hardware/algorithms, and the industry chain. It contrasts AI-assisted and AI-generated writing, analyzes the advantages of large-scale pre-trained and diffusion-based approaches, and discusses Metaverse integration. The paper also outlines central challenges—data quality, compute demands, privacy, NLP generalization, and governance—before proposing promising directions such as cross-modal generation, enhanced search, media and e-commerce applications, and film. By identifying ethical, trust, and policy considerations, the authors emphasize the need for transparency, human-in-the-loop validation, and responsible deployment to maximize AIGC’s societal and economic benefits.

Abstract

To address the challenges of digital intelligence in the digital economy, artificial intelligence-generated content (AIGC) has emerged. AIGC uses artificial intelligence to assist or replace manual content generation by generating content based on user-inputted keywords or requirements. The development of large model algorithms has significantly strengthened the capabilities of AIGC, which makes AIGC products a promising generative tool and adds convenience to our lives. As an upstream technology, AIGC has unlimited potential to support different downstream applications. It is important to analyze AIGC's current capabilities and shortcomings to understand how it can be best utilized in future applications. Therefore, this paper provides an extensive overview of AIGC, covering its definition, essential conditions, cutting-edge capabilities, and advanced features. Moreover, it discusses the benefits of large-scale pre-trained models and the industrial chain of AIGC. Furthermore, the article explores the distinctions between auxiliary generation and automatic generation within AIGC, providing examples of text generation. The paper also examines the potential integration of AIGC with the Metaverse. Lastly, the article highlights existing issues and suggests some future directions for application.

Figures (14)

• Figure 1: Three stages of AIGC.
• Figure 2: The outline of this survey.
• Figure 3: Three different modes of content production.
• Figure 4: Relations between hardware, algorithms, and data.
• Figure 5: The evolution of AI models.