The digital economy, as a new form of economy, is flourishing. As a mapping of human economic activities in the digital space, the innovation of digital economy theory involves almost all fields of economics, including consumer behavior, enterprise production behavior, market operation mechanism, economic growth mode, income distribution mechanism, and the relationship between government and market. Among them, clarifying fundamental issues such as data elements, digital economic growth, and production relations in the digital economy is an important prerequisite for theoretical innovation in the digital economy. The digital economy is the sum of various economic activities driven by big data, and its basic characteristic is that data becomes a key factor of production. Data is essentially a carrier of information and knowledge, reflecting not only objective economic activities but also the psychological reactions of the public to government policies and major events. Data does not occupy natural physical space (but needs to be stored in the virtual storage space of computing devices), which gives it fast mobility. Therefore, data can efficiently connect the entire process of economic activities such as production, exchange, distribution, circulation, and consumption, optimizing the allocation and combination of various production factors. Traditional production factors, such as land, labor, and capital, all have obvious consumptive properties, and data will not decrease or depreciate due to use. This non consumptive nature breaks the limitations of limited supply of traditional factors on production, and also makes the use of data non exclusive, that is, the same data can be used multiple times by a single user or simultaneously by multiple users, and the value of data will not be damaged as a result. However, non exclusivity also brings ambiguity to the ownership of data and the definition of rights and responsibilities, making digital economic activities often accompanied by externalities. As a new type of production factor, data can only fully exert its empowering role by deeply integrating with traditional production factors. The combination of data and labor forms digital labor, such as online hosts, ride hailing drivers, etc; The combination of data and capital forms digital capital, such as some large digital platforms; The combination of data and technology forms digital technologies, such as unmanned driving technology, drones, big language models, etc; The combination of data and management forms digital management, such as human-computer interaction, intelligent manufacturing, smart cities, etc. Data empowerment has the following important characteristics: firstly, it has a nonlinear empowerment effect. The information and knowledge contained in data are not simply linearly related to the amount of data, and the aggregation of different data may reveal new related knowledge, thereby enhancing the empowering effect of data. This non-linear empowerment effect is an important reason for the increasing marginal returns in the digital economy. The second is the existence of scene dependency. The empowering effect of data is highly dependent on the application scenario. For example, customer waiting time or geographic location data is crucial for ride hailing drivers, but if they deviate from a specific time or location, the value of this data will rapidly decline. Thirdly, it has timeliness. The value of data is closely related to its generation time. Many data reflect economic activity information at the time of their generation, and if not applied in a timely manner, the empowering effect of the data may gradually decline over time. The fourth is the existence of value heterogeneity. Even in the same application scenario, due to differences in data utilization efficiency, algorithm optimization level, and business adaptability, there may still be significant differences in the empowerment of data elements among different enterprises. Treating data as a new type of production factor is an important theoretical innovation, but it is only the starting point of innovation in digital economy theory. To truly understand the inherent logic and operating laws of the digital economy, it is necessary to construct rigorous economic theories based on the uniqueness of data elements and form a systematic analytical framework. For example, traditional economic theory is based on the assumption of scarcity, viewing the contradiction between the finite nature of resources and the infinite nature of human demand as a fundamental problem in economics. However, the non exclusivity of data elements breaks through the scarcity constraint, posing new challenges to traditional economic theory. For example, the free products and services that often exist in the digital economy are also difficult to fully explain through traditional economic theory. Although users do not directly pay in currency, they create value for the platform by providing data. This implicit transaction model overturns the traditional logic of value exchange and profit realization. At the same time, the transaction of personal data often comes with externalities, such as a user's data sharing may inadvertently leak the privacy information of other users. Without appropriate government intervention, the operation of the data factor market may be difficult to effectively safeguard personal privacy rights and social welfare. The explanation for these phenomena cannot be separated from the innovation and development of economic theory. How to determine the value of data elements based on their empowerment and value? This is a fundamental issue in the innovation of digital economy theory. According to Marx's labor theory of value, the value creation of data elements can be understood as the human labor condensed in the process of data elementization, which is specifically reflected in three aspects: first, the labor condensation in the process of raw data formation, second, the physical labor transfer in the process of data collection and processing, and third, the active labor input of data workers. These active labor and materialized labor together constitute the value foundation of data elements, and the value transfer of data elements is to transfer the value of these labor to new digital products or services. The empowerment of data elements is not limited to the production process, but runs through various stages of the entire economic activity, including exchange, circulation, and consumption. Therefore, to examine the value of data elements, a comprehensive analysis must be conducted from the entire economic process. More importantly, in the process of empowering the value of digital products, data is not simply a simple model of proportional linear growth, but rather achieves more complex value enhancement by integrating its own value with the value of other elements, forming a cumulative effect. The complexity and nonlinear characteristics empowered by data elements deepen the connotation of value creation and value transfer, and expand the research scope of labor value theory. We should take Marx's labor theory of value as a guide, combined with the empowering characteristics of data elements, to further enrich and develop the labor theory of value. The full play of the empowering role of data elements and the optimized allocation of data element resources require accurate identification or discovery of the value of data elements as a prerequisite, which cannot be achieved without a sound data element market. In theory, the value of data elements can be obtained from exchange relationships in the digital economy: under market economy conditions, the price of goods or services fluctuates around their value in a perfectly competitive market and is realized through exchange. Therefore, the price of a perfectly competitive market provides a feasible method for measuring the value of data elements. But to achieve this goal, it is necessary to establish a sound data factor market system, with a sufficient number of data factor suppliers and demanders, which are difficult to achieve in the short term. Therefore, how to accurately measure the value of data elements or formulate reasonable prices for them has become an urgent theoretical and practical problem that needs to be solved. In this regard, we can draw on market mechanism design theory, experimental economics methods, and simulation experiments of complex economic systems to explore the pricing mechanism of data elements in depth. Reasonable pricing of data elements can not only guide the flow of data resources to areas or industries with higher efficiency, but also provide a solid foundation for accurately evaluating the value of data assets. Therefore, economic theory research must be closely integrated with digital economy practice, adhere to evidence-based methodology, and scientifically evaluate the value and contribution of data elements. It should be emphasized that, in the absence of a clear theoretical logic for reasonable pricing of data elements and sufficient practical exploration, extra caution must be exercised in the evaluation and inclusion of data assets in the table. Unilaterally exaggerating or overestimating the value of data assets and rashly promoting the inclusion of data assets in the table may lead to a "data asset foam", which will have a negative impact on the healthy development of the digital economy. The contribution rate of data elements to economic growth was widely used in information technologies such as computers from the 1970s to the 1990s. However, in statistical data, productivity, especially in the service industry, seems to have remained stagnant. This phenomenon has attracted widespread attention, and the academic community refers to it as the "productivity paradox": "We can see computers everywhere, but we cannot see them in productivity statistics." The key reason for this phenomenon is that capital factors, especially intangible capital (such as research and development investment, software, and data assets), have not been accurately measured. In the 2008 United Nations System of National Accounts, research and development, software, and other intellectual property products were officially included in the capital category and effectively measured. Based on this improved capital measurement, the contribution rate of information technology to economic growth was effectively reflected in economic statistics. Similarly, scientific measurement of data elements is crucial for accurately assessing their contribution to economic growth. Here, a fundamental economic theoretical question that urgently needs to be addressed is: how much role do data elements play in economic growth. To study this issue, it is necessary to start from the underlying logic of economics, examine the ways in which data is combined with other production factors in various links of economic activities, especially how data reshapes economic structure and production processes, and how to optimize resource allocation to produce more and better products and services. This requires the construction of a production function that conforms to the inherent logic and important characteristics of the digital economy based on theoretical analysis. This means that in the context of the digital economy, we need to further innovate and develop economic growth theories, as well as innovate economic measurement and economic statistics to provide theoretical and methodological support for revealing the underlying mechanisms of data-driven economic growth. The data elements of production relations in the digital economy era and the increasingly widespread use of big data based artificial intelligence technology have not only reshaped productivity, but also profoundly changed production relations, mainly manifested in the following aspects: the substitution of labor by artificial intelligence. The application of artificial intelligence technology has significantly improved production efficiency and greatly shortened the socially necessary labor time required to produce the same amount of products, thus inevitably leading to the phenomenon of artificial intelligence replacing labor. Traditional machines mainly replace simple labor, especially physical labor, while artificial intelligence has to some extent replaced mental labor. For example, ChatGPT can not only learn and understand human language for dialogue, but also complete complex tasks such as writing emails, papers, video scripts, translating text, and writing code. Asymmetric relationships in platform economy. The widespread application of digital technology has given rise to new economic forms such as gig economy and platform economy. The gig economy provides workers with more flexible employment options, but this economic model heavily relies on massive data resources and digital technology, resulting in an asymmetric relationship between capital that masters technology platforms and gig workers. Without good economic governance, gig workers will be in a disadvantaged position in the game against platform capital. Similarly, large technology platforms rely on strong capital, digital technology, and data resources, and through network and scale effects, can form asymmetric relationships with small and medium-sized enterprises and consumers. The income gap caused by the digital divide. The digital divide refers to the differences between different social groups in Internet accessibility and use ability. The objective existence of the digital divide has exacerbated the uneven distribution of economic and social benefits brought by digital technology in different regions and social groups. Urban areas, economically developed areas, and highly educated and highly skilled people are more accessible and make