In order to strengthen the basic support for artificial intelligence, what platforms will my country build?

Building and deploying AI innovation platforms is the basic support for strengthening the research and development and application of AI. In the future, the national level plans to vigorously promote the open source of various general software and technology platforms, and promote military-civilian sharing in accordance with the requirements and relevant regulations of deep military-civilian integration.

The open source software and hardware basic platform for artificial intelligence focuses on building a unified computing framework platform that supports artificial intelligence paradigms such as knowledge reasoning, probability statistics, and deep learning, forming an ecological chain that promotes mutual collaboration between artificial intelligence software, hardware, and intelligent cloud.

The group intelligence service platform focuses on building knowledge resource management and open sharing tools based on large-scale collaboration on the Internet, forming a group intelligence crowd creation platform and service environment for the innovation links of industry, academia, research and application.

The hybrid enhanced intelligent support platform focuses on building heterogeneous real-time computing engines and new computing clusters that support large-scale training, providing service-oriented, systematic platforms and solutions for complex intelligent computing.

The autonomous unmanned system support platform focuses on building a support system for common core technologies of artificial intelligence such as environmental perception, autonomous collaborative control, and intelligent decision-making in complex environments for autonomous unmanned systems, forming an open, modular, and reconfigurable autonomous unmanned system development and testing environment.

The AI ​​basic data and security testing platform focuses on the construction of public data resource libraries, standard test data sets, cloud service platforms, etc. for AI, and forms methods, technologies, specifications and tool sets for AI algorithm and platform security testing and evaluation (Figure 1).

Figure 1: Overall layout of AI innovation platform

Dr. Wang Xiwen, author of the first monograph on " Industry 4.0 " in China, pointed out in his new book "Intelligence +: Interpretation of the Development Plan of the New Generation of Artificial Intelligence" that the guarantee conditions at different levels do not exist in isolation. They interact with each other to form the infrastructure platform for the development of new social sciences, including the following levels: social resources and rules layer, involving various social environments and conditions; physical layer, involving physical environments such as data sensors, wearable computing devices, and data transmission networks, involving primary data processing facilities such as data storage and allocation; computing simulation layer, involving shareable computing and simulation systems; social application layer, involving the operating interface for accessing or feedback on social problems. Through the infrastructure platform, the research and use of big data can be carried out smoothly at a lower cost, and the knowledge spillover of big data can benefit the public at a wider level.

1. Open-source software and hardware platform for artificial intelligence

Build an open source software and hardware foundation platform for artificial intelligence. Establish a big data artificial intelligence open source software foundation platform, an artificial intelligence cloud service platform that collaborates between terminals and the cloud, a new multi-intelligence sensor device and integration platform, a new product design platform based on artificial intelligence hardware, and a big data intelligent service platform in the future network (Figure 2).

Figure 2: Building an open source AI software and hardware foundation platform

2. Group Intelligence Service Platform

Build a group intelligence service platform. Establish a group intelligence crowd innovation computing support platform, a technology crowd innovation service system, a group intelligence software development and verification automation system, a group intelligence software learning and innovation system, an open environment group intelligence decision-making system, and a group intelligence sharing economy service system (Figure 3).

Figure 3: Building a swarm intelligence service platform

Although my country has abundant human resources at this stage, it has not yet released rich and powerful group intelligence and fully played its supporting role in the national innovation system. The "New Generation Artificial Intelligence Development Plan" is based on national conditions and actual needs, focuses on platforms and applications, and proposes to build a group intelligence crowd innovation computing support platform, create a group intelligence technology crowd innovation service system for scientific and technological innovation, and promote the widespread application of group intelligence service platforms in important fields such as intelligent manufacturing, smart cities, smart agriculture, and smart medical care, forming an innovative application system and innovation ecology driven by group intelligence, occupying the high end of the global value chain.

Specifically, by creating a cross-disciplinary and cross-industry "crowd intelligence space" for basic research and high-tech research, effectively integrating various scientific and technological resources and intellectual resources, constructing an Internet-based crowd intelligence service platform, supporting the establishment of common application service systems such as scientific and technological crowd innovation, software innovation, and crowd intelligence decision-making, and solving major problems in the country's economic and social development and people's livelihood improvement. In particular, closely combining my country's development needs in the smart economy and smart society, forming a number of products and solutions for major group intelligence application needs, such as: building a crowd intelligence software learning and innovation system and a crowd intelligence software development and verification automation system to serve the country's major needs for independent software innovation; building a human-machine collaborative and interactively driven evolutionary crowd intelligence decision-making system to achieve complex problem solving and intelligent decision-making in an open environment; developing a crowd intelligence sharing economy service system for various livelihood service fields, improving the utilization rate and sharing of scarce and high-quality resources in the livelihood field, and improving the quality of life of the Chinese people. At the same time, promoting the formation of a scientific and technological innovation ecosystem based on collective intelligence in the country's major scientific and technological directions and fields, and cultivating a new ecology and new model for the development of emerging and prosperous collective intelligence industries will accelerate the transformation and upgrading of traditional industries and the development of emerging industries, making China's collective intelligence the core driving force of national scientific and technological innovation, and fully supporting the country's major strategy of "mass entrepreneurship and innovation".

3. Hybrid Enhanced Intelligent Support Platform

Build a hybrid enhanced intelligent support platform. Establish an artificial intelligence supercomputing center, a large-scale super intelligent computing support environment, an online intelligent education platform, a "human-in-the-loop" driving brain, an intelligent platform for industrial development complexity analysis and risk assessment, an intelligent guarantee platform to support the safe operation of nuclear power, and a human-machine co-driving technology research and development and testing platform (Figure 4).

Figure 4: Building a hybrid augmented intelligence support platform

4. Autonomous unmanned system support platform

Build a support platform for autonomous unmanned systems. Establish a common core technology support platform for autonomous unmanned systems, a support platform for autonomous control of drones and automatic driving of cars, ships and rail transit, a support platform for service robots, space robots, marine robots, polar robots, and a technical support platform for smart factories and smart control equipment (Figure 5).

Figure 5: Building an autonomous unmanned system support platform

(V) Artificial Intelligence Basic Data and Security Testing Platform

Build an AI basic data and security testing platform. Build a public data resource library, standard test data set, and cloud service platform for AI, establish AI algorithm and platform security testing and evaluation models, and develop an AI algorithm and platform security evaluation tool set (Figure 6).

Figure 6: Building AI basic data and security detection platform