Discussion on the Foundation of Logic in Artificial Intelligence

In recent years, generative AI models such as ChatGPT, based on Transformer architectures, have developed rapidly and become a hot topic in society. They are widely used in programming, audio, and video processing fields, significantly enhancing productivity in these areas. However, various problems have also arisen during their application. At its core, generative AI still relies on symbolic formal logic and lacks the ability to simulate human cognitive patterns of understanding the world through prior logic or even dialectical logic. As such, it is impossible for these systems to achieve strong artificial intelligence as we know it. Currently, computer systems are founded on discrete mathematics and symbolic logic. This paper proposes an idea that combines prior logic theory with object-oriented programming techniques, using categorical judgment methods as a foundation to mimic human cognition patterns and discusses a possibility of approaching the realization of strong artificial intelligence.