The essence of computation: from code to cognition Calculation, as the core of information processing, has already gone beyond simple mathematical operations and permeated into every aspect of our lives. From scientific computing to artificial intelligence, the connotation of computing continues to expand, and its essence is becoming increasingly thought-provoking. This article attempts to break through traditional concepts and connect computing with human cognition, artificial intelligence, and complex systems theory, exploring its essence and future. The two core dimensions of computation Traditional concepts often limit computation to algorithm and code execution, ignoring the information processing and cognition behind it. We believe that the essence of computation consists of two core dimensions: The implementation of computational logic: This is the fundamental level of computation, covering programming languages, algorithm design, hardware architecture, and more. Its goal is to efficiently and accurately perform computational tasks, converting inputs into expected outputs. Currently, Moore's Law is gradually becoming invalid, and traditional computing models are facing performance bottlenecks. The emergence of emerging models such as quantum computing and neural morphological computing provides the possibility to break through limitations. They utilize quantum mechanical phenomena or simulate the structure of human brain neural networks to achieve more efficient and low-energy calculations, providing new avenues for solving complex problems. Dealing with computational complexity: Real world problems are often highly complex, involving massive amounts of data, nonlinear relationships, and emergent behaviors. Traditional computing models are difficult to effectively address these challenges. The development of technologies such as distributed computing and cloud computing, as well as the application of complex system theory, helps us better understand and cope with computational complexity. Distributed computing decomposes tasks into multiple nodes for parallel execution, improving efficiency and fault tolerance. Cloud computing provides the ability to access computing resources on demand, making it possible to process massive amounts of data. The theory of complex systems studies the interactions and emergent behaviors between individuals in complex systems, providing a theoretical framework for understanding and predicting the behavior of complex systems. Human cognition and computation The human brain is a highly complex distributed computing system, where each individual uses their own neural network to uniquely 'map' and 'compute' the world. This "personalized computing" model is a core feature of human cognition and an important inspiration for the development of artificial intelligence. Each individual's brain has a unique structure and experience, and their way of processing external information is also different. This individual diversity enables humans to understand the world from different perspectives and generate creative thinking. The Development Path of Artificial Intelligence Artificial intelligence aims to build computer systems that can simulate human intelligence. At present, there are two main paths for the development of artificial intelligence: Connectivism: Represented by deep learning, it learns patterns and patterns from massive amounts of data by simulating neural network structures. The success of large-scale language models such as GPT demonstrates the enormous potential of connectionism in fields such as natural language processing. However, connectionist models often lack interpretability and generalization ability. Symbolism: emphasizes the symbolic representation and logical reasoning of knowledge, such as knowledge graphs, expert systems, etc. In recent years, neural symbolic artificial intelligence has attempted to combine connectionism and symbolism to build more powerful and interpretable artificial intelligence systems. This hybrid approach aims to leverage the advantages of both paradigms, combining the learning ability of connectionism with the reasoning and interpretability of symbolism. Computational World and Observer Calculation does not exist independently, it requires a reference object/subject to reflect its meaning. Like the observer effect in quantum mechanics, the presence of an observer can affect the results of calculations. Building a subject centered 'computing world' will help us better understand the essence of computing and the future development of artificial intelligence. In this' computational world ', complex system characteristics such as individuality, emergence, and irreducibility will be fully reflected. Each individual is an independent computing unit that forms a complex and dynamic network through interaction and information exchange with other individuals. The behavior of the entire system emerges from the interaction of individual agents and cannot be simply attributed to the attributes of individual agents. conclusion This article reexamines the essence of computation, expanding it from simple code execution to information processing and cognitive processes. We analyzed the two core dimensions of computation: the implementation of computational logic and the response to computational complexity, and linked them to the development of human cognition and artificial intelligence. By introducing the concepts of "personalized computing" and "computational world", we attempt to construct a more comprehensive and profound framework for understanding computing, and provide insights for the future development of artificial intelligence. In the future, with the continuous development of computing technology, we expect to see more innovative computing paradigms and applications emerge, driving progress in human society.