With the incremental commercial use of 5G wireless technology, academic research on the sixth-generation (6G) mobile communication technology has also begun. 6G wireless networks will enhance and expand 5G applications by achieving a higher data rate per user/device (10–100 times greater than 5G). Besides, it will support wider coverage and a larger number of connected devices, massive Internet of Things (IoT), distributed massive multiple-input multiple-output (DM-MIMO), and high and reliable connectivity. Machine learning (ML) techniques, as part of artificial intelligence (Al), have proven to be extremely useful in a wide variety of applications and now their applications in 6G wireless communication systems have been the subject that attracts incredible interest in recent years. ML involves teaching the machines to perform tasks independently based on making data-driven decisions and can accurately estimate various parameters and promote interactive decision-making. Therefore, one of the main and key components of 6G systems will be the use of AI for mobile communication networks and we believe that 5G/6G and AI will be the two core technologies of the future intelligent system. In this tutorial, we first focus on the most applicable ML approaches and algorithms and present the requirements and the best context of use for each of those algorithms. A series of simple examples are given as introductory concepts. Then, in the second part, we introduce the most current applications of AI in 5G/6G wireless technologies which consist of body-area networks, smart homes and industry, connected healthcare, remote surgery, mission-critical, autonomous driving, and vehicle-to-vehicle (V2V) communications, etc. Finally, we present a related project which is about localization in 5G wireless technology using ML methods. In this project, a fingerprint-based positioning method in M-MIMO systems was proposed to provide an accurate and reliable localization system in a 5G wireless network. For this purpose, signal features in DM-MIMO systems and instantaneous channel state information (CSI) in collocated M-MIMO (CM-MIMO) systems were extracted from the received signals as fingerprints. Then, an optimal clustering scheme was presented to split up the target area into several small regions, which minimized the searching space of reference points and decreased the computational complexity. Finally, a regression model using algorithms such as Gaussian process regression (GPR) and neural network was created for each region based on the data distribution within each region to provide further positioning accuracy. Through this approach, the average error was improved to a few meters, which was expected in 5G networks and the computational complexity of utilizing the ML method was also reduced.