Editorial Review

This entry in the Artificial Intelligence series (8 titles) opens with an engaging scenario: A silver humanoid robot successfully identifies a number of objects. When a nearby human asks for something to eat, the robot hands him an apple, explaining it’s the only edible option. This relatable example of generative AI—the ability to learn from data—is an example of a GAN (generative adversarial network). The text does a good job of explaining how a GAN is modeled after the human brain and profiles other applications, including chatbots, creating images, and solving problems for NASA. Equal space is devoted to drawbacks: misinterpretations, false-data input, bias (both conscious and unconscious), intellectual property rights, and job loss. Substantial support material helps make this an accessible introduction to some pretty sophisticated concepts.