Although not a requirement for 5G, AI algorithms, especially those under the banner of neural networks and machine learning, will be integral to many of the new use cases and performance promises. Without such algorithms it is hard to see how 5G will achieve the desired levels of automated control and configuration necessary for network slicing and dynamic resource allocation to work effectively. Use cases such as the Industrial IoT and VR/AR (virtual reality/augmented reality) will also require AI algorithms to achieve the required levels of control and precision for the first and to organize the computationally intensive execution for the second. Machine learning will also figure increasingly prominently in 5G security (see separate item). A distinction can be drawn…