As early as 2019, the Ministry of Science and Technology issued the Guidelines for the Construction of National Pilot Zones for the Innovative Development of a New Generation of Artificial Intelligence to promote AI as an important leading force in regional development. The recent "14th Five-Year Plan for the Development of the Digital Economy" issued by the State Council also clearly puts forward the need to optimize and upgrade the digital infrastructure, promote the orderly development of intelligent computing centers, and create a new type of intelligent infrastructure that integrates intelligent arithmetic, universal algorithms, and development platforms.
As a strategic technology to drive and lead the new round of technological revolution and industrial change, AI is not only an important means to empower industrial transformation and innovation, but also an important driving force for the high-quality development of the digital economy, which is playing an increasingly important role in unleashing the potential of the digital economy and constructing a new pattern of digital economic development.
First, the call to privacy computing and public **** data openness, breaking the AI dilemma
Artificial Intelligence industry to data, arithmetic, algorithms as the core, and the cost of traditional factors of production measurements are very different. Currently, the development of AI enterprises are facing many challenges: high investment costs for self-built arithmetic, poor elasticity, high maintenance costs, high cost of cloud GPU scaling; long market paths, difficult to establish a sense of trust with the customer, the customer domain data can not provide a model for continuous optimization; the AI toolchain is difficult to deploy, the consumption of funds, and the cost of financing is high. Most importantly, for AI companies, training data is difficult to obtain and baseline models cannot be tuned.
These problems cannot be solved by AI enterprises alone.
First of all, in terms of realizing the technical aspects of improving AI progress, data, as an important support for AI progress, is limited by the characteristics of the data and can not be used for large-scale aggregation. In recent years, the hot privacy computing technology is believed to be able to help AI maximize data resources under certain conditions. Through the ability to make "data available and invisible", it can be aggregated under the premise of protecting data privacy and security, and realize the use of data***, so as to better play the supporting role of data in the development of the field of artificial intelligence. The construction of its industry or regional data network to help data in the controllable, measurable and subject to privacy and security protection of the premise to help the development of AI industry, this technology may become the "14th Five-Year Plan" in the development of the digital economy to enable the efficient layout of the AI infrastructure important hand.
Data as the core production capital of AI is released in the process, in addition to technical capabilities, early public data sets also play an important role in the context of the digital economy, the government has realized this. China is rich in data resources, research data shows that by 2025 China's data volume will reach 48.6ZB,In the central level data element documents, government data and social resource data have a clear distinction, in which government data is the data produced and collected by the government in the process of fulfilling its duties, and the opening of government data is an important dimension of the country's implementation of the big data strategy. Therefore, the government is taking the lead in improving the system of regulations and institutions in the field of data, public ****data*** enjoyment and open application, and promoting the standardization and standardization of public ****data resource transactions. Many places in our country have made positive exploration of this, but also has a certain effect.
In the case of two-way parallel technology and norms, privacy computing is very likely to become an indispensable infrastructure in the digital economy industry. Its importance is also reflected in the innovation of the mode of data element utilization.
Relying on privacy computing to construct a city data element platform and computing power base, optimize the development of AI industry through internal element resources, configure the data element market from the beginning of the demand for industrial development, and use data elements as market elements to attract digital economy enterprises. Open the door to help local guide artificial intelligence industry ecology, maximize the value of mining data, and seek incremental market innovation services. The exploration of the "data element investment" model, through the "data investment" in the emerging development of artificial intelligence enterprises, can meet the local digital economy to enhance the practical initiatives.
Two, privacy computing or become an indispensable infrastructure in the field of artificial intelligence
Privacy computing, usually refers to the data provider does not disclose the original data under the premise of the data processing and computing, to complete the data value mining technology system. It can ensure that data is "available and invisible" in the process of circulation and integration, and realize the separation between data ownership and data usage, thus avoiding the leakage of authorized data to the greatest extent.
In 2021, the first year of commercialization of privacy computing will usher in a phase of progress in technology and commercialization. In the field of technology, the head of privacy computing enterprises Huazhong Qingjiao is a typical "academic school"; commercialization direction, the wing side of the healthy number, rich number of technology is the representative of the "practice school". The core advantage of the former is privacy protection computing and data circulation technology, standards and infrastructure based on cryptographic theories such as multi-party secure computing; the latter is a provider and operator of "Internet of Data and Computing" based on privacy computing, which is committed to realizing the value of data and contributing to the circulation of data value, and focuses on the landing of practical application scenarios; and Fortune Technology is based on privacy computing and specializes in financial services, and has cooperated with more than 50 companies in the field. Fortune Tech, on the other hand, is based on privacy computing and specializes in the financial sector, and has cooperated with more than 50 enterprises so far, with headline customer cases in various industries. Representatives of the "practice school" are also shown in the privacy computing based on different industries, with model pioneering customer cases, to bring inspiration to the application of technology. As you can see, there is no lack of artificial intelligence in these cases.
It is worth mentioning that the academic attack and application landing as the basis of privacy computing, the two are driven by each other, and the emergence of representative enterprises in both directions also proves that the privacy computing industry has a certain foundation, and is one step closer to the scale of application.
And a number of privacy computing companies have mentioned that the entire data ecosystem is a multi-level, multi-disciplinary market, should actively develop the ecosystem, packaged with "data openness is the norm, not open is accidental" willingness to open. The opening of data resources such as government data, public **** data, social resource data, enterprise data, etc., will greatly affect the development of AI industry data elements market-oriented.
Three, privacy computing as the base of the development platform integration of new intelligent infrastructure or will become the future trend
By building an AI open platform to realize the real meaning of the AI ecological openness, including market demand, scene docking, technical collaboration, data **** enjoyment, etc., is the future of what we want to see. . And privacy computing, here has an irreplaceable value.
Fan Jianping, former chief economist of the State Information Center, has said that artificial intelligence and the digital economy will become the new engine of China's medium- and long-term economic development. It is foreseeable that the future of "privacy computing + AI" will change the logic of operation of thousands of industries and become an important "cornerstone" of the digital economy by allowing data to flow safely. I believe that with the continuous development of AI technology and the continuous breakthrough in the scale of AI industry applications, the integration of a new type of intelligent infrastructure with privacy computing as the technical base of the development platform will help AI enterprises to continuously seek market innovation services, and at the same time, guide the country's artificial intelligence industry ecosystem.
(info)