ChatGPT Is All You Need: Untangling Its Underlying AI Models, Architecture, Training Procedure, Capabilities, Limitations And Applications

Abstract

ChatGPT has now become a global phenomenon that has revolutionized the manner in which machines interact with humans. It is a noteworthy enhancement in the field of generative AI, where generative AI is designed to create new contents similar to human-generated contents. ChatGPT has been developed to utilise a generative AI model as its underlying model. It is self-adaptive as it learns from its multimodal interactions with users and improves its responses based on the feedback received from users; thus, providing more accurate and personalised responses to users. ChatGPT is constantly evolving to incorporate new technologies and capabilities within it; therefore, a comprehensive study is required to understand its various aspects. This paper will present a comprehensive study of ChatGPT exploring its architecture, underlying AI models and techniques, training procedures, capabilities, limitations and applications. This comprehensive study will provide users a stepby-step guide with detailed information by analysing the various aspects of ChatGPT systematically.