Generative artificial intelligence (GAI) has become an innovative technology that is capable of generating new and creative content like text, images, programming code, audio, video, 3D models, and other types of media. Popular GAI models like ChatGPT, BERT, PaLM, Gemini, and DALL-E have acquired the attention of numerous users with a wide range of applications in education, the arts, and healthcare, among other fields. Various studies have explored different individual GAI tools, issues, and challenges; however, a comprehensive view on issues is not available in the literature. Considering this research gap, the authors examine a number of text-based GAI tools as well as the associated issues and challenges, classifying these according to various concerns. These GAI tools are capable of generating realistic content. Each tool has its own unique characteristics, which makes different GAI tools dissimilar. A summary of various text-based GAI tools along with their respective benefits and drawbacks has been presented. The issues and challenges associated with GAI have also been categorized. Since GAI continues to evolve, it raises various concerns; some of the significant ones are discussed in this study, including bias in data, security and privacy concerns, societal concerns, as well as ethical and legal concerns.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Comparative Study of Text-Based Generative Artificial Intelligence Models and Tools

  • Satwant Kaur,
  • Manu Sood

摘要

Generative artificial intelligence (GAI) has become an innovative technology that is capable of generating new and creative content like text, images, programming code, audio, video, 3D models, and other types of media. Popular GAI models like ChatGPT, BERT, PaLM, Gemini, and DALL-E have acquired the attention of numerous users with a wide range of applications in education, the arts, and healthcare, among other fields. Various studies have explored different individual GAI tools, issues, and challenges; however, a comprehensive view on issues is not available in the literature. Considering this research gap, the authors examine a number of text-based GAI tools as well as the associated issues and challenges, classifying these according to various concerns. These GAI tools are capable of generating realistic content. Each tool has its own unique characteristics, which makes different GAI tools dissimilar. A summary of various text-based GAI tools along with their respective benefits and drawbacks has been presented. The issues and challenges associated with GAI have also been categorized. Since GAI continues to evolve, it raises various concerns; some of the significant ones are discussed in this study, including bias in data, security and privacy concerns, societal concerns, as well as ethical and legal concerns.