This chapter introduces machine learning in contemporary artificial intelligence. The first section looks at an expert system developed in the early days of AI research – ID3, which employs a decision-tree-based algorithm. The second section looks at advances in deep learning, which has transformed modern machine learning. We introduce a deep learning model inspired by the mammalian visual system, illustrating how it can extract hierarchical information from the raw data. The third section addresses two examples of neural networks -- autoencoders and convolutional neural networks, which can feature in layers of deep learning networks. The last section looks at a distinct type of machine learning -- reinforcement learning. We explain how deep reinforcement learning has made possible the two most spectacular milestones in artificial intelligence - AlphaGo and AlphaGo Zero.

AI is a complex, multifaceted concept and is therefore hard to define because AI can refer to technological artifacts, certain methods or a scientific field that is split into many subfields and that is continuously changing and evolving AI systems can therefore be seen as digital artifacts that require hardware and software components and that contain at least one learning or learned component, i.e., a component that is able to change the system’s behavior based on presented data and the processing of this data.

China’s fuzzy logic system and government support for pilot petri dishes is perfectly suited to the current state of AI research. This has enabled the rapid development of world-class AI applications, particularly in image recognition. This is due, in part, to the regulatory environment facilitating the development of AI pilots. Yet it is further argued that this suitability is due to a combination of three factors: (1) the current state of AI research and its applicability to numerous real-world applications; (2) the open nature of AI research culture globally; and (3) the complex emerging role of public–private petri dishes in China for testing innovative applications. The chapter also explains how public–private connections are formed, including how top-down government signalling is important to the trajectory of private companies.