Generating synthetic data has become an important technique in data science that provides solutions to many challenges such as private data, rare data, and rich information. This research explores the diversity of computing techniques, from artificial intelligence techniques such as GANs, generalized instruction tuning and variable auto-encoders (VAEs) to legal rendering, live cloning and data protection technology. A summary of each method is provided and its content, advantages, limitations, and practical applications in various fields are discussed. Through comparative analysis, this article assesses the advantages and disadvantages of the various methods and offers perspectives on their suitability for different applications. It also discusses the challenges and future directions in the development of synthetic materials and provides recommendations to researchers and professionals.